Hans W. Gottinger, Essays on Technology, Markets and Development
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
ESSAYS ON<br />
TECHNOLOGY,<br />
MARKETS AND<br />
DEVELOPMENT
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
ESSAYS ON<br />
TECHNOLOGY,<br />
MARKETS AND<br />
DEVELOPMENT<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
STRATEC Munich, Germany<br />
www.stratec-c<strong>on</strong>.net<br />
stratec_c@yahoo.com<br />
gottingerhans@gmail.com
Aggregate Keywords (10): Investment, Technological <strong>Development</strong>, Stochastic Models,<br />
Network Ec<strong>on</strong>omics, Supply Chain Competiti<strong>on</strong>/Coopetiti<strong>on</strong>, Strategic Alliances,<br />
Microec<strong>on</strong>omics of Ec<strong>on</strong>omic Growth, Ec<strong>on</strong>omic Growth <strong>and</strong> Catching-Up, <strong>Development</strong> Paths,<br />
Ec<strong>on</strong>omic Growth under Envir<strong>on</strong>mental C<strong>on</strong>straints
CONTENTS<br />
HANS W. GOTTINGER<br />
CONTENTS<br />
ESSAYS ON TECHNOLOGY, MARKETS<br />
AND DEVELOPMENT<br />
C<strong>on</strong>tents ................................................................................................ I<br />
Foreword. .............................................................................................. III<br />
Preface <strong>and</strong> Introducti<strong>on</strong> ............................................................................... V<br />
1. Innovati<strong>on</strong> <strong>and</strong> <strong>Technology</strong> ............................................................ 1<br />
1.1 Stochastics of Innovati<strong>on</strong> Processes .............................................................. 3<br />
1.2 Stochastic Models for Capital Growth ........................................................... 17<br />
1.3 Modeling stochastic innovati<strong>on</strong> races ........................................................... 33<br />
1.4 Global technological races. ..................................................................... 53<br />
1.5 Stochastic racing <strong>and</strong> competiti<strong>on</strong> in network markets .......................................... 67<br />
1.6 High speed technology competiti<strong>on</strong>. ............................................................ 85<br />
1.7 Sequential <strong>Technology</strong> Choice <strong>and</strong> R&D Racing. ................................................ 103<br />
1.8 Internet of Things Ec<strong>on</strong>omic <strong>and</strong> Industrial Dimensi<strong>on</strong>s ......................................... 123<br />
2. <strong>Markets</strong> <strong>and</strong> Competiti<strong>on</strong> ........................................................... 133<br />
2.1 Innovati<strong>on</strong>, Dynamics of Competiti<strong>on</strong> <strong>and</strong> Market Dynamics ................................... 135<br />
2.2 Network Ec<strong>on</strong>omics for Open Source Technologies ............................................. 151<br />
2.3 Vertical Competiti<strong>on</strong> <strong>and</strong> Outsourcing ......................................................... 169<br />
2.4 Supply-Chain Coopetiti<strong>on</strong> ..................................................................... 177<br />
2.5 A tale <strong>on</strong> Chinese Innovati<strong>on</strong> <strong>and</strong> Competiti<strong>on</strong> ................................................. 183<br />
2.6 Chinese Innovati<strong>on</strong> <strong>and</strong> Competiti<strong>on</strong> .......................................................... 189<br />
2.7 Competitive Positi<strong>on</strong>ing through Strategic Alliances ............................................ 199<br />
3. Growth <strong>and</strong> <strong>Development</strong> ........................................................... 217<br />
3.1 Microec<strong>on</strong>omic Foundati<strong>on</strong>s of Ec<strong>on</strong>omic Growth (Review) ..................................... 219<br />
3.2 Ec<strong>on</strong>omic Growth, Catching-Up, Falling Behind <strong>and</strong> Getting Ahead ............................. 301<br />
3.3 Catch-Up <strong>and</strong> C<strong>on</strong>vergence. ................................................................... 315<br />
3.4 Escaping from <strong>Development</strong> Traps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345<br />
3.5 Some Applicati<strong>on</strong>s of a Result in C<strong>on</strong>trol Theory to Ec<strong>on</strong>omic Planning Models .................. 355<br />
3.6 Optimal Ec<strong>on</strong>omic Growth when CO2 C<strong>on</strong>straints are Critical ................................... 377<br />
3.7 Ec<strong>on</strong>omic Choice <strong>and</strong> <strong>Technology</strong> Diffusi<strong>on</strong> in New Product <strong>Markets</strong> ............................ 385<br />
I
HANS W. GOTTINGER<br />
CONTENTS<br />
II
SOURCES OF ARTICLES CORRESPONDING TO SECTIONS<br />
HANS W. GOTTINGER<br />
Sources of articles corresp<strong>on</strong>ding to Secti<strong>on</strong>s 1.1 – 3.7<br />
1 <strong>Technology</strong><br />
1.1 Journal of Ec<strong>on</strong>omics 49(2), 1989, 123-138<br />
1.2 Journal of Ec<strong>on</strong>omics 54(3), 1991, 267-281<br />
1.3 Technological Forecasting & Social Change 69, 2002, 607-624<br />
1.4 Japan <strong>and</strong> the World Ec<strong>on</strong>omy 18, 2006, 181-193<br />
1.5 Int. J. <strong>Technology</strong>, Policy <strong>and</strong> Management 4(3), 2004, 240-256<br />
1.6 Int. J. Business <strong>and</strong> Systems Research 10(1), 85-102<br />
1.7 Journal of Policy Studies (Japan) 15(9), 2003, 43-62<br />
1.8 Int. J. Bus. <strong>and</strong> Ec<strong>on</strong>omics Research 6(5), 2017, 115-123<br />
2 <strong>Markets</strong><br />
2.1 Archives of Bus.Research 4(1), 2015, 1-16<br />
2.2 Adv. Soc.Sci. Research 4(7), 2017, 164-181<br />
2.3 Int. J. Bus. <strong>and</strong> Ec<strong>on</strong>omics Research 4(6), 2015, 315-322<br />
2.4 Int. J. Bus. <strong>and</strong> Ec<strong>on</strong>omics Research 4(2), 2015, 67-71<br />
2.5 Research Memo. Abstract, STRATEC Munich , 2012<br />
2.6 Int. J. Bus. <strong>and</strong> Ec<strong>on</strong>omics Research 2(2), 2013, 22-32<br />
2.7 Int. J. Revenue Management 8(1), 2008, 1-17<br />
3 Growth <strong>and</strong> <strong>Development</strong><br />
3.1 Unpublished Supplement to: <str<strong>on</strong>g>Hans</str<strong>on</strong>g> W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g> <strong>and</strong> Mattheus F.A. Goosen , eds<br />
Strategies of Ec<strong>on</strong>omic Growth <strong>and</strong> Catch-Up, NovaScience, New York 2012<br />
3.2 Journal of Policy Studies (Japan) 21(11), 2005, 1-14<br />
3.3 J.Appl.Ec<strong>on</strong>omics <strong>and</strong> Business 2(5), 2014, 67-95<br />
3.4 (with C. Umali), J.Appl.Ec<strong>on</strong>omics <strong>and</strong> Business 3(1), 2015, 5-13<br />
3.5 Weltwirtschaftliches Archiv (Review of Ec<strong>on</strong>omics) 111. 1975, 728-751<br />
3.6 Energy Ec<strong>on</strong>omics, July 1992, 192-199<br />
3.7 Weltwirtschaftliches Archiv (Review of Ec<strong>on</strong>omics) 123, 1987, 93-120<br />
III
HANS W. GOTTINGER<br />
SOURCES OF ARTICLES CORRESPONDING TO SECTIONS<br />
IV
FOREWORD<br />
HANS W. GOTTINGER<br />
FOREWORD<br />
Yesterday (or was it Thursday) Apple was crowned as the first “trilli<strong>on</strong> dollar company”, willing a supposed<br />
race with Alphabet (Google), Amaz<strong>on</strong> <strong>and</strong> Microsoft. Many guru hours am<strong>on</strong>g financial journalists have<br />
been, <strong>and</strong> more will be spent <strong>on</strong> recounting the high (<strong>and</strong> a few low) points of Apple’s rise. This book has<br />
little to say <strong>on</strong> the dynamics of that particular competiti<strong>on</strong>, except perhaps to note that insights from the<br />
theory of games may play a role in the story.<br />
Speaking for myself, I find it more fascinating that Apple’s path to success, is littered with the corpses<br />
or dying remains of former corporate super-stars. Why, <strong>on</strong>e w<strong>on</strong>ders, did the inventors <strong>and</strong> innovators<br />
behind the key inventi<strong>on</strong>s <strong>and</strong> innovati<strong>on</strong>s, gain so little of the wealth they created? For that matter, why<br />
did formerly dominant tech firms, such as GE, IBM, Xerox, Kodak <strong>and</strong> HP, lose their way <strong>on</strong> the yellow<br />
brick road? Why did the current crop of winners get so powerful <strong>and</strong> why did the losers lose out? Moreover,<br />
can we expect Amaz<strong>on</strong> <strong>and</strong> Alphabet to fall by the wayside in their turn?<br />
These questi<strong>on</strong>s are of interest to inventors <strong>and</strong> venture capitalists, of course. They are also of interest<br />
to l<strong>on</strong>g-term instituti<strong>on</strong>al investors, such as pensi<strong>on</strong> funds, endowment funds <strong>and</strong> insurance companies.<br />
At another level, they are potentially important to our underst<strong>and</strong>ing of the ec<strong>on</strong>omic system as a whole.<br />
Ec<strong>on</strong>omists have l<strong>on</strong>g been at a loss to underst<strong>and</strong> the underlying drivers of booms <strong>and</strong> busts, al<strong>on</strong>g with<br />
their accompanying changes in wealth distributi<strong>on</strong> <strong>and</strong> c<strong>on</strong>sequences for political power.<br />
Ec<strong>on</strong>omic theory relies heavily <strong>on</strong> linear models. But the “boom-bust” pattern is evidently n<strong>on</strong>-linear.<br />
Let me indulge in a bit of theorizing of my own. I postulate that both booms <strong>and</strong> busts are likely to reflect<br />
two sides of the same phemomen<strong>on</strong>, the phenomen<strong>on</strong> of “lock-in/lock-out”. What this means is that a first<br />
choice am<strong>on</strong>g competing bidders or technologies or business models can become effectively irreversible<br />
<strong>on</strong>ce initial advantages (e.g. learning or experience) kick in. For example, the “experience curve” is a<br />
particularly important comp<strong>on</strong>ent of increasing productivity Returns to scale create (temporary) advantages<br />
for incumbents (booms) <strong>and</strong>, c<strong>on</strong>sequently, disadvantages to challengers. This phenomen<strong>on</strong> can (<strong>and</strong> does)<br />
often result in sub-optimal (or even bad) l<strong>on</strong>g-term outcomes (busts).<br />
The obvious example today is the fact that our whole industrial system (<strong>and</strong> urban development, in particular)<br />
is “locked in” to dependence <strong>on</strong> a road-based transport system requiring private automobiles (<strong>and</strong> trucks)<br />
using liquid hydrocarb<strong>on</strong> fuels. Electric <strong>and</strong> steam-powered vehicles were still competitive in 1900, but they<br />
had a short range, due to storage limitati<strong>on</strong>s. Steam power overcame that limit by introducing a c<strong>on</strong>denser<br />
in the 1920s (with the Doble), but it was much too late to compete. Electrificati<strong>on</strong> of cars <strong>on</strong>ly began again<br />
after the year 2000, thanks to the inventi<strong>on</strong> of lithium-i<strong>on</strong> batteries. Range <strong>and</strong> battery re-charge rate are<br />
still limitati<strong>on</strong>s, but performance in those domains is improving fast. What if the lithium-i<strong>on</strong> battery had<br />
been invented a century earlier?<br />
The “lock-in” of uranium-based nuclear power technology (<strong>and</strong> the “lock-out” of a potentially much superior<br />
thorium-based technology) is another example. It seems that uranium (specifically the isotope U-235) became<br />
the fuel of choice precisely because of the military uses of its by-products, neptunium <strong>and</strong> plut<strong>on</strong>ium. Now,<br />
in stati<strong>on</strong>ary power applicati<strong>on</strong>s plut<strong>on</strong>ium is an unwanted (except by terrorists) pollutant, not a valuable<br />
by-product. Yet uranium technology still gets all the investment.<br />
Positive returns-to-scale create a short-term advantage – hence a barrier to challenge by competitors –<br />
resulting from positive returns to scale. The assumpti<strong>on</strong> of c<strong>on</strong>stant or declining returns to scale is wr<strong>on</strong>g<br />
V
HANS W. GOTTINGER<br />
FOREWORD<br />
because “network” service industries (like the teleph<strong>on</strong>e system) dem<strong>on</strong>strably exhibit increasing returns<br />
to scale. (See pp 151-160 of this book). The simple explanati<strong>on</strong> for positive returns in network sectors is that<br />
the larger the network the greater the number of people c<strong>on</strong>nected to each node, resulting in increased value<br />
to each customer. This phenomen<strong>on</strong>, in more general terms, is also resp<strong>on</strong>sible for populati<strong>on</strong> clustering<br />
(cities), the success of department stores, browsers, the success of search engines (like Google) <strong>and</strong> a variety<br />
of other ec<strong>on</strong>omically important activities. All of these phenomena are impossible to explain in terms of<br />
ec<strong>on</strong>omic theories based <strong>on</strong> c<strong>on</strong>stant or decreasing returns to scale.<br />
Positive feedbacks in a deterministic system do not necessarily lead the system to a unique equilibrium<br />
state. The dynamic equati<strong>on</strong>s for the ec<strong>on</strong>omy in disequilibrium are n<strong>on</strong>-linear, <strong>and</strong> for n<strong>on</strong>-linear systems<br />
there can be multiple alternative equilibria. At a bifurcati<strong>on</strong> point the system then “chooses” <strong>on</strong>e path,<br />
often for short-term reas<strong>on</strong>s. There is no guarantee whatsoever that the path adopted at a bifurcati<strong>on</strong> point<br />
leads to the “best” l<strong>on</strong>g-term soluti<strong>on</strong>. Indeed, as, the path selected is often sub-optimal, <strong>and</strong> may lead<br />
the ec<strong>on</strong>omy “over a cliff” if there is no scope for remedial acti<strong>on</strong>. It is a priori unlikely that the optimum<br />
short-term choice is also the optimum l<strong>on</strong>g-term choice.<br />
This book will, hopefully, provide some clues to these mysteries.<br />
Robert Ayres, Novartis Professor of Ec<strong>on</strong>omics <strong>and</strong> <strong>Technology</strong>, INSEAD, Paris<br />
VI
PREFACE AND INTRODUCTION<br />
HANS W. GOTTINGER<br />
<str<strong>on</strong>g>Essays</str<strong>on</strong>g> <strong>on</strong> Decisi<strong>on</strong>s, <strong>Technology</strong>, <strong>Markets</strong> <strong>and</strong> <strong>Development</strong><br />
PREFACE AND INTRODUCTION<br />
This collecti<strong>on</strong> of essays highlights the interface topics of innovati<strong>on</strong> <strong>and</strong> technology, markets <strong>and</strong> competiti<strong>on</strong>,<br />
growth <strong>and</strong> development in an evoluti<strong>on</strong>ary dynamic ec<strong>on</strong>omy of a dominantely capitalistic market ec<strong>on</strong>omy<br />
in a globalized c<strong>on</strong>text. They span a time frame from the 1980s to current developments.<br />
Previews<br />
In Part 1 the core of Secs. 1.1. to 1.8 emphasize the linkage between high speed technological competiti<strong>on</strong><br />
through technology races in firms, industries <strong>and</strong> nati<strong>on</strong>al ec<strong>on</strong>omies. We start with observati<strong>on</strong>s how<br />
stochastic innovati<strong>on</strong> processes emerge in a firm <strong>and</strong> are carried through by capital investments <strong>and</strong><br />
how the capital allocati<strong>on</strong> activity is governed by a stochastic process c<strong>on</strong>strained by optimal stopping<br />
of productivity enhancing investments (Secs.1.1, 1.2). Further <strong>on</strong>, the remaining Secs. 1.3 to 1.7 rely <strong>on</strong><br />
(stochastic) differential game analysis in symmetric <strong>and</strong> asymmetric rivalrous situati<strong>on</strong>s am<strong>on</strong>g firms that<br />
extends the modeling of R&D based competiti<strong>on</strong> bey<strong>on</strong>d c<strong>on</strong>venti<strong>on</strong>al models found in the ec<strong>on</strong>omics <strong>and</strong><br />
technology literature. Many high technology industries are characterized by positive network externalities.<br />
Firms essentially compete <strong>and</strong> cooperate <strong>on</strong> R&D <strong>and</strong> the producti<strong>on</strong> of goods <strong>and</strong> services that share a<br />
network. Some models of competiti<strong>on</strong> c<strong>on</strong>tain special features that apply equally well to network markets.<br />
One is the uncertainty in technological development or uncertainty in the realizati<strong>on</strong> of a firm’s R&D effort.<br />
The other is the dynamic nature of price competiti<strong>on</strong> between firms in the presence of network effects.<br />
Firms compete with each other over an extended period of time <strong>and</strong> must therefore strategically choose<br />
prices as the market shares of the firms evolve.<br />
Most existing models of network markets focus <strong>on</strong> <strong>on</strong>ly <strong>on</strong>e of the two features. Further, almost all the<br />
models have the comm<strong>on</strong>ality that <strong>on</strong>e firm captures the market instantaneously <strong>and</strong> sells to all c<strong>on</strong>sumers<br />
from then <strong>on</strong> (the “winner-takes-all” market) . However, the history of high technology industry abounds<br />
with instances where rival firms (<strong>and</strong> technologies) have had extended battles for providing the industry<br />
st<strong>and</strong>ard.<br />
In network markets evoluti<strong>on</strong> of market share is a very interesting phenomen<strong>on</strong> especially in the face of<br />
uncertainty about future product quality. Sec.1.5 attempts to provide a framework that could capture the<br />
richness of market share evoluti<strong>on</strong> in the presence of network externalities. In a specific situati<strong>on</strong> of a high<br />
-tech market if you are not the indisputable leader you have <strong>on</strong>ly little hope to achieve leadership unless<br />
you come up with a major innovati<strong>on</strong> replacing the leading firm. Incremental innovati<strong>on</strong> – making slight<br />
improvements in the leaders’ products – will not enable a small firm to overtake a leader that enjoys the<br />
benefits of network ec<strong>on</strong>omies. Therefore, it is not atypical for a fringe firm that invests heavily to displace<br />
the leader by leapfrogging the leader’s technology.<br />
We are encountering the problem of tradeoff between “network dominance” <strong>and</strong> “radical innovati<strong>on</strong>”<br />
that could tip the market the other way, with a significant caveat added that breakthrough R&D is highly<br />
uncertain. From a strategic perspective, in this envir<strong>on</strong>ment, for any two firms of asymmetric size, both<br />
compete dynamically over prices to win market share. In this dynamic process there are two ways to achieve<br />
(temporary) m<strong>on</strong>opolistic status. The ‘smaller’ firm can use dynamic pricing competiti<strong>on</strong> to delay the time<br />
in which the ‘larger’ firm wins a critical market share in the hope to hit the innovati<strong>on</strong> first <strong>and</strong> displace<br />
it. If the innovator is “patient”, the probability of innovati<strong>on</strong> <strong>and</strong> the discount factor are sufficiently high,<br />
there is an equilibrium in which duopoly persists (no firm achieves a critical market share) until <strong>on</strong>e of the<br />
two firms wins the race for innovati<strong>on</strong>.<br />
In Sec.1.7 we model the R&D process as <strong>on</strong>e that is composed of a series of stages, at each of which some<br />
set of product characteristics are determined. At each stage, the firm exercises discreti<strong>on</strong> as to the extent<br />
VII
HANS W. GOTTINGER<br />
PREFACE AND INTRODUCTION<br />
to which the particular stage characteristics will be developed. Further, when the firm’s R&D resource<br />
endowment is fixed <strong>and</strong> stage-specific, the firm faces a tradeoff between the extent to which it will develop<br />
the stage characteristics <strong>and</strong> the time that it will spend in doing so. We will refer to the degree to which<br />
the stage characteristic is developed as the characteristic’s (targeted) level of strategic positi<strong>on</strong>ing. In this<br />
sense, the more “forward-looking” the product characteristics, the greater the profits the firm can expect<br />
to get from the product but the l<strong>on</strong>ger the expected time to completi<strong>on</strong> of R&D. We will speak of this as<br />
the speed–leadership tradeoff in each stage. When the completi<strong>on</strong> of R&D requires many such tradeoffs to<br />
be resolved sequentially, firms will, at the beginning of each stage, want to make their choices in light of<br />
the choices made by all firms in previous stages <strong>and</strong> their stage completi<strong>on</strong> records to that time. To carry<br />
technological innovati<strong>on</strong> through networks across firm/industry boundaries creates a new paradigm of<br />
network centered innovati<strong>on</strong> that integrate through intelligent software tools as outlined <strong>and</strong> reviewed <strong>on</strong><br />
the present state of affairs in Sec. 1.8.<br />
In highly competitive technological industries new challenges <strong>and</strong> opportunities are arising in the new<br />
product development arena. Driven by global markets, global competiti<strong>on</strong>, the global dispersi<strong>on</strong> of scientific/<br />
engineering talent, <strong>and</strong> the advent of new informati<strong>on</strong> <strong>and</strong> communicati<strong>on</strong> technologies (ICT) a new visi<strong>on</strong><br />
of product development is that of a highly disaggregated , distributive process with people <strong>and</strong> organizati<strong>on</strong>s<br />
spread throughout the world. At the same time products are becoming increasingly complex requiring<br />
numerous engineering decisi<strong>on</strong>s to bring them to market. Competitive pressures mean that “time to market”<br />
has become a key to new product success. However, at the same time it is important to keep the innovati<strong>on</strong><br />
<strong>and</strong> quality dimensi<strong>on</strong>s of the new product at their optimal level.<br />
A central questi<strong>on</strong> to address is how should firms invest in innovati<strong>on</strong> <strong>and</strong> what are the implicati<strong>on</strong>s of such<br />
investments for competitive advantage? Underst<strong>and</strong>ing why a firm benefits from investments in innovati<strong>on</strong><br />
<strong>and</strong> quality illuminates issues of competitive strategy <strong>and</strong> industrial organizati<strong>on</strong>. In the field of competitive<br />
strategy, much attenti<strong>on</strong> has been devoted to the c<strong>on</strong>cept of core capabilities. Underst<strong>and</strong>ing how firms make<br />
optimal investments in the face of competiti<strong>on</strong> reveals the nature of competiti<strong>on</strong> <strong>and</strong> provides theoretical<br />
<strong>and</strong> managerial implicati<strong>on</strong>s for developing core competence <strong>and</strong> dynamic capabilities.<br />
In major parts of competitive analysis involving R&D decisi<strong>on</strong>s the focus is <strong>on</strong> breakthrough innovati<strong>on</strong>s<br />
which could create entirely new markets, for example, in studies featuring patent racing between competing<br />
firms. In more comm<strong>on</strong> competitive situati<strong>on</strong>s we observe firms, however, competing by investing in<br />
incremental improvements of products. It is an important aspect when innovati<strong>on</strong> is c<strong>on</strong>sidered to be<br />
manifested in product quality, process improvements <strong>and</strong> in the overall quality culture of an organizati<strong>on</strong><br />
. For example, after product launch, incremental improvement of different aspects of product quality,<br />
improvements in various business processes <strong>and</strong> an incremental adopti<strong>on</strong> of a quality culture are quite realworld<br />
phenomena. Thus innovati<strong>on</strong> is c<strong>on</strong>ceptualized as being manifested in terms of quality. Moreover,<br />
the quality dimensi<strong>on</strong> extends to pursue c<strong>on</strong>tinuous total quality improvement for the entire product life<br />
cycle. Some firms operate in a simultaneous product launch situati<strong>on</strong> while others compete sequentially<br />
by adopting the role of leader or follower. The strategic implicati<strong>on</strong>s in these diverse circumstances can be<br />
treated within a unified framework of dynamic stochastic differential games.<br />
Part 2 , Competiti<strong>on</strong> <strong>and</strong> <strong>Markets</strong>, is devoted to look into the interacti<strong>on</strong> of innovati<strong>on</strong> culture, competitive<br />
dynamics <strong>and</strong> shaping of markets with c<strong>on</strong>tinuous emergence <strong>and</strong> proliferati<strong>on</strong> of open source technologies<br />
dominating in ICTs (Secs. 2.1, 2.2). In Sec. 2.2 we focus <strong>on</strong> a c<strong>on</strong>testable market with network externalities<br />
engaging an incumbent <strong>and</strong> an entrant. The incumbent, unlike the entrant, already has an installed base<br />
of c<strong>on</strong>sumers. We look at decisi<strong>on</strong> situati<strong>on</strong>s of firms regarding how proprietary they want to make their<br />
technology , either through patent protecti<strong>on</strong> or through development in open source systems (OSS).<br />
We explicitly model the direct <strong>and</strong> indirect effects of network externalities. For example, more software<br />
companies are willing to produce programs for an operating system (OS) if it has a larger c<strong>on</strong>sumer base.<br />
This increased competiti<strong>on</strong> could lead to an improvement of the quality of the OS. The model predicts that<br />
using open source technologies is likely to enhance the rate of R&D , <strong>and</strong> c<strong>on</strong>sequently the quality of the<br />
product.. An incumbent that would choose this strategy is likely to deter entrance of a newcomer because<br />
it can play out its advantage of a larger network.<br />
Competitive <strong>and</strong> cooperative forces, termed “coopetiti<strong>on</strong>”, are likely to synchr<strong>on</strong>ize in extending to supply<br />
VIII
PREFACE AND INTRODUCTION<br />
HANS W. GOTTINGER<br />
chain networks (Secs. 2.3, 2.4) . This opens the door for strategic cooperati<strong>on</strong> as focussed <strong>on</strong> a timely case<br />
of a Chinese industry study (Secs. 2.5, 2.6). The degree of competitive positi<strong>on</strong>ing in view of forming<br />
strategic alliances (Sec. 2.7) is explored. This preview c<strong>on</strong>cludes by emphasizing the analytical , strategic<br />
<strong>and</strong> practical implicati<strong>on</strong>s <strong>and</strong> providing directi<strong>on</strong>s for future industry analysis.<br />
In recent years with the emergence of e-business <strong>and</strong> a supply chain view for product development processes,<br />
multiple firms with varying <strong>and</strong> at times c<strong>on</strong>flicting objectives enter into collaborative arrangements. In<br />
such situati<strong>on</strong>s, the competitive strategy based <strong>on</strong> quality <strong>and</strong> innovati<strong>on</strong> could potentially permeate in<br />
those collaborative setups. A recent example is the collaborative venture of S<strong>on</strong>y <strong>and</strong> Samsung to build<br />
a cutting edge plant for LCD flat screen TV though displaying <strong>on</strong>going fierce rivalry in new product<br />
launches in exactly the same product categories. When innovati<strong>on</strong> <strong>and</strong> quality levels form the core of a<br />
firm’s capabilities, each member in the supply chain would have an incentive to invest <strong>and</strong> improve their<br />
dynamic capabilities. This leads to tacit competiti<strong>on</strong> am<strong>on</strong>g collaborative product development partners<br />
by means of active investment in innovati<strong>on</strong> <strong>and</strong> quality.<br />
Although the forces of innovati<strong>on</strong> are central to competiti<strong>on</strong> in young, technically dynamic industries, they<br />
also affect mature industries where life cycles historically were relatively str<strong>on</strong>g, technologies mature, <strong>and</strong><br />
dem<strong>and</strong>s stable.<br />
A strategy for technology must c<strong>on</strong>fr<strong>on</strong>t primarily what the focus of technical development will be. The<br />
questi<strong>on</strong> is what technologies are critical to the firm’s competitive advantage. In this c<strong>on</strong>text, technology<br />
must include the know-how the firm needs to create, produce <strong>and</strong> market its products <strong>and</strong> deliver them<br />
to customers. As a major step in creating a technology strategy it has to define those capabilities where<br />
the firm seeks to achieve a distinctive advantage relative to competitors. For most firms, there are a large<br />
number of important areas of technological know-how but <strong>on</strong>ly a h<strong>and</strong>ful where the firm will seek to create<br />
truly superior capability.<br />
Having determined the focus of technical development <strong>and</strong> the source of capability, the firm must establish<br />
the timing <strong>and</strong> frequency for innovati<strong>on</strong> efforts. Part of the timing issue involves developing technical<br />
capabilities, <strong>and</strong> the rest involves introducing technology into the market. The frequency of implementati<strong>on</strong><br />
<strong>and</strong> associated risks will depend in part <strong>on</strong> the nature of the technology <strong>and</strong> the markets involved (e.g. disk<br />
drive vs. automotive technology), but in part <strong>on</strong> strategic choice. At the extreme, a firm may adopt a rapid<br />
incremental strategy, that is, frequent, small changes in technology that cumulatively lead to c<strong>on</strong>tinuous<br />
performance improvement. The polar opposite might be termed the great leap forward strategy. In this<br />
approach, a firm chooses to make infrequent but large-scale changes in technology that substantially<br />
advance the state of the art .<br />
Sec. 2.3 uncovers the importance of asymmetries by investigating the difference between firms in terms of<br />
key parameters. Moreover, for the examinati<strong>on</strong> of inter-firm competiti<strong>on</strong> through product development, the<br />
c<strong>on</strong>text of simultaneous entry <strong>and</strong> sequential entry are treated separately. This allows a deeper underst<strong>and</strong>ing<br />
of the implicati<strong>on</strong>s of informati<strong>on</strong> asymmetry <strong>and</strong> commitment which have been regarded as important<br />
determinants in the c<strong>on</strong>text of game-theoretic studies.<br />
The study of innovati<strong>on</strong> based competiti<strong>on</strong> has often c<strong>on</strong>sidered aspects related to patent races <strong>and</strong> incremental<br />
product-process innovati<strong>on</strong> to achieve distinctive advantage. However, recently innovati<strong>on</strong>-based competiti<strong>on</strong><br />
has become an aspect of buyer-supplier relati<strong>on</strong>ships. There are many instances in manufacturing where <strong>on</strong>e<br />
finds situati<strong>on</strong>s of lock-ins created by innovative suppliers. For example, in the computer industry Intel <strong>and</strong><br />
Microsoft as suppliers of microprocessor <strong>and</strong> operating systems, respectively, to desktop manufacturers like<br />
IBM , Hewlett-Packard <strong>and</strong> Dell illustrate such innovati<strong>on</strong>-based lock-ins. Indeed, there exists an evolving<br />
power structure (dubbed “channel power”) in a supply chain driven by innovati<strong>on</strong> competence of its members.<br />
A business c<strong>on</strong>text is envisaged in which, at any given point in time of the relati<strong>on</strong>ship, both the buyer <strong>and</strong><br />
the supplier could be pursuing innovati<strong>on</strong> simultaneously. We recognize that the primary motivati<strong>on</strong> for<br />
such investments in innovati<strong>on</strong> by members of the supply chain is to increase their differential or relative<br />
IX
HANS W. GOTTINGER<br />
PREFACE AND INTRODUCTION<br />
channel power in the supply chain. Also in a situati<strong>on</strong> where the buyer is locked in by a supplier, the<br />
buyer may actively pursue the creati<strong>on</strong> of a substitute technology by investing in innovati<strong>on</strong>. The primary<br />
motivati<strong>on</strong> for the buyer in this case would be to eliminate the technology lock-in <strong>and</strong> become independent.<br />
Another c<strong>on</strong>tributi<strong>on</strong>, Sec.2.4, presents the competitive role of innovati<strong>on</strong> am<strong>on</strong>g collaborating firms.<br />
The models provide reas<strong>on</strong>ing for inter-firm incentives in forming collaborative arrangements for product<br />
development. The dynamics of relati<strong>on</strong>ship am<strong>on</strong>g supply chain partners is viewed in terms of their<br />
respective innovati<strong>on</strong> competence. It is emphasized that varying power arrangements in a supply chain leads<br />
to different implicati<strong>on</strong>s for investments in innovati<strong>on</strong> by buyer <strong>and</strong> supplier. It highlights the incentive for<br />
buyer <strong>and</strong> supplier to strategically maneuver their overall innovati<strong>on</strong> levels by appropriate investments.<br />
The purpose is to provide a strategic framework <strong>and</strong> insights regarding power <strong>and</strong> competiti<strong>on</strong> in a<br />
collaborative supply chain setup.<br />
Managerial dimensi<strong>on</strong>s<br />
The analytical results provide insights into managerial implicati<strong>on</strong>s regarding strategic issues c<strong>on</strong>cerned<br />
with product development <strong>and</strong> supply chain planning. The c<strong>on</strong>ceptual framework provides a foundati<strong>on</strong> for<br />
managers to view competitive advantage emanating from product development in a more integrated manner.<br />
The results of Sec.1.6 point towards a time variant innovati<strong>on</strong> investment strategy. It suggests that firms<br />
competing by means of new product development must choose instantaneous investment in innovati<strong>on</strong><br />
such that it increases with time. The quality manifestati<strong>on</strong>s exhibit an increasing trajectory. In a symmetric<br />
competiti<strong>on</strong>, the investment profile of the leader is sigmoidal while that of the follower is c<strong>on</strong>vex increasing.<br />
‘T’he impact of such an investment profile is aptly reflected in the quality improvement trajectory.<br />
The results provide implicati<strong>on</strong>s for a firm for innovati<strong>on</strong> investments by c<strong>on</strong>sidering their relative strengths<br />
<strong>and</strong> weaknesses. The explorati<strong>on</strong> of firm asymmetries provide some interesting implicati<strong>on</strong> that a firm<br />
should c<strong>on</strong>sider when competing via new product development. The results can be translated into executable<br />
decisi<strong>on</strong>s for investments. These results provide insights to evaluate different scenarios for pursuing strategies<br />
which would lead to best market outcome.<br />
Secs. 2.3 <strong>and</strong> 2.4 provide a way of thinking about business strategy <strong>and</strong> operati<strong>on</strong>al alignment in a<br />
collaborative network. The results provide directi<strong>on</strong>s for strategies to be adopted in a supply chain c<strong>on</strong>text.<br />
These are becoming increasingly important in the present business envir<strong>on</strong>ment, where many firms are<br />
joined together in a collaborative network. In industries driven by innovati<strong>on</strong> ,e,g. semi-c<strong>on</strong>ductors or<br />
biotechnology, having the c<strong>on</strong>trol <strong>on</strong> the overall innovati<strong>on</strong> levels that drive the technology l<strong>and</strong>scape could<br />
mean a str<strong>on</strong>g strategic advantage for <strong>on</strong>e partner over other c<strong>on</strong>tributing firms.<br />
Innovati<strong>on</strong> competence plays an important role in this approach <strong>on</strong> a supply chain relati<strong>on</strong>ship. In Sec. 2.4<br />
the specific scenario modeled is representative of practical situati<strong>on</strong>s. As an example, it can be observed<br />
that IBM is actively involved in creating quantum computers. One reas<strong>on</strong>ing for investments in quantum<br />
computers could be an extensi<strong>on</strong> of a technology l<strong>and</strong>scape. However, equally important is the realizati<strong>on</strong><br />
that a success of such an endeavour could lead to a lock-out of Intel from this newly formed market for<br />
quantum courputers. The simplistic model used in Sec. 2.4 allows for an explorati<strong>on</strong> of such a situati<strong>on</strong>.<br />
The stati<strong>on</strong>ary Markov perfect Nash equilibrium investment strategy of the buyer <strong>and</strong> supplier is found to<br />
be time invariant <strong>and</strong> is characterized by the parameters in the model. An interesting insight obtained as<br />
a result of analysis of the model is that the supplier can influence the motivati<strong>on</strong> of the buyer to invest in<br />
substitute technology. The underlying mechanism can be translated into pricing strategy that results in a<br />
l<strong>on</strong>g-term buyer-supplier relati<strong>on</strong>ship.<br />
The analysis of a model of buyer-supplier relati<strong>on</strong>ship allows to investigate a more generalized setup of<br />
buyer-supplier innovati<strong>on</strong>-based competiti<strong>on</strong>. The aspect of channel power is tightly integrated in the<br />
analysis to underst<strong>and</strong> the complex behavioral aspects using a simple framework. These results indicate<br />
that structural risk plays an important role in innovati<strong>on</strong> investment decisi<strong>on</strong>s. Moreover, the variance in<br />
wealth formati<strong>on</strong> is an important indicator of how much to invest. These results are quite intuitive <strong>and</strong><br />
X
PREFACE AND INTRODUCTION<br />
HANS W. GOTTINGER<br />
enable managers to objectively resolve some of these strategic decisi<strong>on</strong>s.<br />
The propositi<strong>on</strong>s provide insights into the dynamics of relati<strong>on</strong>ship between a buyer <strong>and</strong> a supplier firm.<br />
These propositi<strong>on</strong>s present the role <strong>and</strong> resp<strong>on</strong>sibility of buyer in creating motivati<strong>on</strong> for the supplier to<br />
collaborate. The importance of critical assets is aptly amplified in the results. The possessi<strong>on</strong> of critical<br />
assets, often observed with the upstream partner gives the supplier a potential to achieve relative market<br />
closure through a positi<strong>on</strong> of dominance over competitors. It is likely that a firm in possessi<strong>on</strong> of such a<br />
critical asset also has the potential to achieve effective leverage over collaborating partner-, <strong>and</strong> suppliers.<br />
The resp<strong>on</strong>sibility of the buyer is to create a delicate balance between managing supplier <strong>and</strong> customers.<br />
Similar to the c<strong>on</strong>ceptual development the results suggest that for a buyer that wishes to be successful an<br />
underst<strong>and</strong>ing must be developed about how to own <strong>and</strong> c<strong>on</strong>trol critical assets that provide opportunities to<br />
create customer dependency <strong>and</strong> “lock-in” .A competence in procurement management forms the cornerst<strong>on</strong>e<br />
for success of l<strong>on</strong>g term business strategy.<br />
Extensi<strong>on</strong>s<br />
Several extensi<strong>on</strong>s can be c<strong>on</strong>sidered for supply chain modeling <strong>and</strong> they establish behavioral features for<br />
the Microfoundati<strong>on</strong>s of Ec<strong>on</strong>omic Growth, Sec. 3.1.<br />
First; within the framework of Sec.1.6 the model can be enriched by evaluating the improvement in product<br />
quality with learning effect. The learning effect is very well documented in the literature <strong>and</strong> the extensi<strong>on</strong><br />
to incorporate learning effects is straightforward. Specifically, learning can be used to characterize the<br />
dynamics of evoluti<strong>on</strong> of product quality <strong>and</strong> in the characterizati<strong>on</strong> of costs associated with innovati<strong>on</strong><br />
investments.<br />
For analytical simplicity the revenue functi<strong>on</strong> is treated as a salvage value. As an extensi<strong>on</strong> the incorporati<strong>on</strong><br />
of the dynamics of evoluti<strong>on</strong> of revenue in the analysis of the model can be explored. In the sequential play<br />
game, the state dynamics of the follower can be c<strong>on</strong>sidered to be dependent <strong>on</strong> the state of the leader at<br />
previous time instant. This modificati<strong>on</strong> would allow for analysis of a model in which the leader uses openloop<br />
Nash equilibrium strategy for innovati<strong>on</strong> investments whereas the follower would adopt a Markovian<br />
Nash equilibrium strategy. Additi<strong>on</strong>al insights can be gained by this modificati<strong>on</strong> for leader-follower<br />
competitive dynamics in new product development.<br />
Firm asymmetries can be explored by c<strong>on</strong>sidering multiple parameters at the same time. This would enable<br />
a richer underst<strong>and</strong>ing of the strategies that a leader <strong>and</strong> a follower should adopt based <strong>on</strong> their strengths<br />
<strong>and</strong> weaknesses.<br />
The competitive dynamics of buyers <strong>and</strong> suppliers are explored. The essays present the competitive role<br />
of innovati<strong>on</strong> am<strong>on</strong>g collaborating firms. The research asserts that the buyer <strong>and</strong> the supplier could be<br />
in a competitive relati<strong>on</strong>ship due to efforts in innovati<strong>on</strong>. Existing literature in operati<strong>on</strong>s management<br />
doesn’t explicitly c<strong>on</strong>sider the implicati<strong>on</strong>s of innovati<strong>on</strong> based competitive strategies in a collaborative<br />
supply chain c<strong>on</strong>text. Models are presented that provide motivati<strong>on</strong> to explicitly c<strong>on</strong>sider such competitive<br />
situati<strong>on</strong>s. The noti<strong>on</strong> of lock-in <strong>and</strong> channel power is implicit in the model <strong>and</strong> presents an approach for<br />
theory-building activities in this area.<br />
We know that the model presented derives from a simplified versi<strong>on</strong> of the dynamics between a buyer <strong>and</strong><br />
a supplier. The c<strong>on</strong>straints placed <strong>on</strong> different parameters <strong>and</strong> the c<strong>on</strong>text of a m<strong>on</strong>opoly supplier <strong>and</strong> a<br />
buyer facing perfect competiti<strong>on</strong> is an approximati<strong>on</strong> for analytical c<strong>on</strong>venience. However, this enables a<br />
theoretical investigati<strong>on</strong> of some important aspects underlying buyer-supplier relati<strong>on</strong>ships. Specifically,<br />
the research provides a way of thinking about business strategy in collaborative networks. These are<br />
becoming increasingly important in the present business envir<strong>on</strong>ment, where many firms are part of a<br />
collaborative network. In industries driven by innovati<strong>on</strong> , the industrial internet, advanced manufacturing,<br />
semic<strong>on</strong>ductors <strong>and</strong> biotechnology, having c<strong>on</strong>trol over the overall innovati<strong>on</strong> levels could mean a str<strong>on</strong>g<br />
strategic advantage for <strong>on</strong>e partner over other c<strong>on</strong>tributing firms.<br />
At the same time, the firm already having the channel power can take some acti<strong>on</strong>s that would enable a<br />
XI
HANS W. GOTTINGER<br />
PREFACE AND INTRODUCTION<br />
l<strong>on</strong>g-term relati<strong>on</strong>ship built <strong>on</strong> trust based governance.<br />
Several extensi<strong>on</strong>s could be c<strong>on</strong>sidered to better represent such “coopetitive” relati<strong>on</strong>ships between<br />
collaborating firm. Given the objective interests, it is necessary for collaborating partners to know how far<br />
the other side is prepared to c<strong>on</strong>cede before it is no l<strong>on</strong>ger profitable to be in a relati<strong>on</strong>ship. Informati<strong>on</strong> about<br />
the cost structures of the collaborating partners <strong>and</strong> their relative utility from the exchange relati<strong>on</strong>ship is<br />
critical to an underst<strong>and</strong>ing of power in exchange relati<strong>on</strong>ship. Hence, <strong>on</strong>e of the extensi<strong>on</strong> of the problem<br />
is to c<strong>on</strong>sider aspects of informati<strong>on</strong> asymmetry. The two key problems generated by private informati<strong>on</strong><br />
<strong>and</strong> imperfect observability are – adverse selecti<strong>on</strong> <strong>and</strong> moral hazard. The adverse selecti<strong>on</strong> is a c<strong>on</strong>diti<strong>on</strong><br />
of supplier opportunism that occurs prior to a signing of a c<strong>on</strong>tract. Moral hazard <strong>on</strong> the other h<strong>and</strong> refers<br />
to supplier opportunism that occurs <strong>on</strong>ce the buyer has signed a c<strong>on</strong>tract.. A rati<strong>on</strong>al buyer tries to possess<br />
informati<strong>on</strong> that counters the above motives <strong>on</strong> the part of suppliers. Aspects related to detailed price<br />
comparis<strong>on</strong> between the supplier; its competitors <strong>and</strong> the range of substitutes provides means to investigate<br />
buyer strategies. An investigati<strong>on</strong> of the implicati<strong>on</strong>s of threat by buyer of placing the c<strong>on</strong>tract elsewhere<br />
to get additi<strong>on</strong>al price c<strong>on</strong>cessi<strong>on</strong>s from its preferred supplier can be carried out.<br />
The assumpti<strong>on</strong>s made in the models <strong>and</strong> the c<strong>on</strong>straints put <strong>on</strong> variables can be relaxed to obtain a better<br />
representati<strong>on</strong>. A scenario with multiple suppliers/buyers enables enrichment of the findings . The result<br />
regarding guidelines for price b<strong>and</strong>width to create l<strong>on</strong>g-term relati<strong>on</strong>ships can be explored in further detail.<br />
The implicati<strong>on</strong>s <strong>on</strong> profits by adopting a strategy for l<strong>on</strong>g-term relati<strong>on</strong>ship can be compared <strong>and</strong> c<strong>on</strong>trasted<br />
with those of adopting the Markov-perfect Nash equilibrium strategies. The insights obtained from this<br />
approach also provide a background for future studies in supply chain c<strong>on</strong>tracts. The problem c<strong>on</strong>text <strong>and</strong><br />
results clearly dem<strong>on</strong>strate the competitive role of innovati<strong>on</strong> between the buyer <strong>and</strong> the supplier. It provides<br />
a rati<strong>on</strong>ale for investigati<strong>on</strong> of supply chain c<strong>on</strong>tracts by explicitly c<strong>on</strong>sidering the evoluti<strong>on</strong> of innovati<strong>on</strong><br />
competence <strong>and</strong> critical assets of collaborating partners.<br />
In Sec.2.4 the stochastic differential. game model c<strong>on</strong>siders the Brownian moti<strong>on</strong> for the evoluti<strong>on</strong> of<br />
wealth for supplier <strong>and</strong> buyer. These Brownian moti<strong>on</strong>s can be treated to be autocorrelated to investigate<br />
the associated implicati<strong>on</strong>s of investment strategy. The model can be extended to examine a sequential<br />
game <strong>and</strong> the associated investment strategy for buyer <strong>and</strong> supplier. Another extensi<strong>on</strong> to the model is<br />
c<strong>on</strong>siderati<strong>on</strong> of spill-over effects. The spill-over of knowledge created by investments in innovati<strong>on</strong> is<br />
comm<strong>on</strong> in literature pertaining to ec<strong>on</strong>omics <strong>and</strong> management science. Such a spill-over of knowledge<br />
may provide some interesting results regarding buyer-supplier power structure <strong>and</strong> investment strategy.<br />
In Part 3, Ec<strong>on</strong>omic Growth <strong>and</strong> <strong>Development</strong>, Secs. 3.1 – 3.7 we start with a broad survey in Sec. 3.1 <strong>on</strong><br />
the microfoundati<strong>on</strong>s of ec<strong>on</strong>omic growth where many model results <strong>on</strong> firm <strong>and</strong> industrial racing behavior<br />
in Parts 1 <strong>and</strong> 2 are sourced to built the structural basis for industrial growth <strong>and</strong> in the aggregate for<br />
macroec<strong>on</strong>omic growth. In Sec.3.2 the link micro-macro is the focus of a modified classical growth model<br />
(the Solow model) which at the center is built up<strong>on</strong> catch-up behavior in the aggregate. Secs. 3.3 <strong>and</strong> 3.4<br />
turn the page <strong>on</strong> development , Sec. 3.3 <strong>on</strong> catch-up behavior <strong>and</strong> development paths between nati<strong>on</strong>s in<br />
a global <strong>and</strong> historical c<strong>on</strong>text from the first industrial revoluti<strong>on</strong> <strong>on</strong> c<strong>on</strong>vergence <strong>and</strong> divergence of paths<br />
<strong>and</strong> changing positi<strong>on</strong>s <strong>on</strong> various performance criteria in ec<strong>on</strong>omic growth <strong>and</strong> prosperity. While Sec. 3.3<br />
is more c<strong>on</strong>cerned with “racing to the top” in a more closely related field Sec. 3.4 emphasizes development<br />
paths from the bottom to overcome or break out of development traps or poverty traps. Secs. 3.5 <strong>and</strong><br />
3.6 partially depart from the main theme, Sec. 3.5 portrays a a c<strong>on</strong>trol theoretic model of development<br />
planning, more a traditi<strong>on</strong>al macro model either to be superimposed <strong>on</strong> or interacting cooperatively with<br />
more decentralized market <strong>and</strong> trade oriented industrial structures. Sec. 3.6 emphasizes another aspect<br />
of opimal ec<strong>on</strong>omic growth under possible fixed/variable c<strong>on</strong>staints trough climate change restricti<strong>on</strong>s.<br />
In what follows we restrict comments <strong>on</strong> the main theme of Part 3 .<br />
On ec<strong>on</strong>omic growth <strong>and</strong> development we assume that a key driver of ec<strong>on</strong>omic growth am<strong>on</strong>g nati<strong>on</strong>s<br />
or regi<strong>on</strong>al ec<strong>on</strong>omic entities is an intrinsic, <strong>on</strong>going, perpetual <strong>and</strong> historically observable rivalry to<br />
propel a state‘s st<strong>and</strong>ing, prestige, power <strong>and</strong> ec<strong>on</strong>omic performance through getting ahead of or not to<br />
fall too much behind its rivals in a pecking order. As we observe in ec<strong>on</strong>omic history through the age of<br />
industrializati<strong>on</strong> from the mid 18th century, the cumulative process of industrial development in a dynamic<br />
time c<strong>on</strong>text can be correlated with some countries emerging as early leaders being challenged by others,<br />
XII
PREFACE AND INTRODUCTION<br />
HANS W. GOTTINGER<br />
later adopters that strived for new leadership positi<strong>on</strong>s. This is in c<strong>on</strong>trast to those that were not part of the<br />
techno-ec<strong>on</strong>omic race, falling behind in ec<strong>on</strong>omic prosperity, <strong>and</strong> inevitably not being able to catch-up<br />
over decades or even centuries.<br />
The design mechanism for strategies of ec<strong>on</strong>omic growth essentially embodies a science <strong>and</strong> technology<br />
competiti<strong>on</strong> between nati<strong>on</strong>s <strong>and</strong> its industries as the creator of value added products <strong>and</strong> processes through<br />
competiti<strong>on</strong> as reflected in the level of gross domestic product (GDP). This is in the Schumpeterian traditi<strong>on</strong>.<br />
It follows from exploring the varying implicati<strong>on</strong>s of the noti<strong>on</strong>, clearly advanced by Schumpeter (1942) that<br />
it is the expectati<strong>on</strong> of supernormal profits from a temporary m<strong>on</strong>opoly positi<strong>on</strong> following an innovati<strong>on</strong><br />
that is the chief driver of R&D investment.<br />
The success of ec<strong>on</strong>omic growth due to diffusi<strong>on</strong> of advanced technology or the possibility of leapfrogging<br />
is mainly attributable to how social capability evolves, (i.e., which effects become more influential, growing<br />
resp<strong>on</strong>siveness to competiti<strong>on</strong> or growing obstacles to it <strong>on</strong> account of vested interests <strong>and</strong> established<br />
positi<strong>on</strong>s). In a time of overcoming a major financial crisis with a slowing pace of deregulati<strong>on</strong>, privatizati<strong>on</strong>,<br />
liberalizati<strong>on</strong> <strong>and</strong> lifting of trade barriers it remains an interesting questi<strong>on</strong> regarding an internati<strong>on</strong>al<br />
competitive order whether industry racing patterns are sufficiently c<strong>on</strong>trolled by open world-wide markets<br />
or whether complementary internati<strong>on</strong>al agreements (e.g. regulati<strong>on</strong>s, c<strong>on</strong>trols) are needed to eliminate or<br />
mitigate negative externalities without compromising the positive externalities that come with industry racing.<br />
We investigate multiple kinds of races, fr<strong>on</strong>tier races am<strong>on</strong>g leaders <strong>and</strong> would be’ leaders <strong>on</strong> a global,<br />
regi<strong>on</strong>al <strong>and</strong> local scale corresp<strong>on</strong>ding to different leagues like in major sports events, as there are in each<br />
category catchup races am<strong>on</strong>g laggards <strong>and</strong> imitators. Technological fr<strong>on</strong>tiers at the firm <strong>and</strong> industry race<br />
levels offer a powerful tool through which to view evolving technologies within an industry. By providing<br />
benchmarking roadmaps that show where an individual firm is relative to other firms in the industry, they<br />
highlight the importance of strategic interacti<strong>on</strong>s in the firm’s technology decisi<strong>on</strong>s.<br />
In an ec<strong>on</strong>omic history perspective getting ahead, catching up <strong>and</strong> falling behind processes are taking<br />
place toward leaders <strong>and</strong> followers, trailers <strong>and</strong> laggards within a group of industrialized, industrializing<br />
<strong>and</strong> developing countries in pursuit of higher levels of power, welfare <strong>and</strong> productivity. Moses Abramovitz<br />
(1986) explains the central idea of the catch-up hypothesis as the trailing countries adopting behaviour<br />
of a backlog of unexploited technology. Supposing that the level of labor productivity were governed<br />
entirely by the level of technology embodied in capital stock, <strong>on</strong>e may c<strong>on</strong>sider that the differentials in<br />
productivities am<strong>on</strong>g countries are caused by the “technological age” of the stock used by a country relative<br />
to its “chr<strong>on</strong>ological age”.<br />
The technological age of capital is an age of technology at the time of investment plus years elapsing from<br />
that time. Since a leading country may be supposed to be furnished with the capital stock embodying, in<br />
each vintage, technology which was at the very fr<strong>on</strong>tier at the time of investment, the technological age<br />
of the stock is, so to speak, the same as its chr<strong>on</strong>ological age. While a leader is restricted in increasing<br />
its productivity by the advance of new technology, trailing countries, al<strong>on</strong>g Abramovitz’ line, have the<br />
potential to make a larger leap as they are provided with the privilege of exploiting the backlog in additi<strong>on</strong><br />
to the newly developed technology. Hence, followers being behind with a larger gap in technology will<br />
have a str<strong>on</strong>ger potential for growth in productivity. The potential, however, will be reduced as the catchup<br />
process goes <strong>on</strong> because the unexploited stock of technology becomes smaller <strong>and</strong> smaller.<br />
This hypothesis explains the diffusi<strong>on</strong> process of best practice technology <strong>and</strong> gives the same sort of<br />
S-curve change in productivity rise of catching-up countries am<strong>on</strong>g a group of industrialized countries as<br />
that of followers to the leader in an industry. Although this view can explain the tendency to c<strong>on</strong>vergence<br />
of productivity levels of follower countries, it fails to answer the historical puzzles why a country, such as<br />
the United States, has preserved the st<strong>and</strong>ing of the technological leader (taking excepti<strong>on</strong>s for two world<br />
wars) for a l<strong>on</strong>g time since taking over leadership from Britain <strong>and</strong> Germany before the First World War<br />
<strong>and</strong> why the shifts have taken place in the ranks of follower countries in their relative levels of productivity,<br />
i.e., technological gaps between them <strong>and</strong> the leader.<br />
Abramovitz poses some extensi<strong>on</strong>s <strong>and</strong> qualificatios <strong>on</strong> this simple catchup hypothesis in an attempt to<br />
XIII
HANS W. GOTTINGER<br />
PREFACE AND INTRODUCTION<br />
explain these facts. Am<strong>on</strong>g other factors than technological backwardness he lays stress <strong>on</strong> a country’s<br />
social capability, i.e., years of educati<strong>on</strong> as a proxy of technical competence in its political, commercial,<br />
industrial, <strong>and</strong> financial instituti<strong>on</strong>s. The social capability of a country may become str<strong>on</strong>ger or weaker<br />
as technological gaps close <strong>and</strong> thus, he states, the actual catch-up process does not lend itself to simple<br />
formulati<strong>on</strong>. This view has a comm<strong>on</strong> underst<strong>and</strong>ing to what Mancur Ols<strong>on</strong> (1982) expresses to be public<br />
policies <strong>and</strong> instituti<strong>on</strong>s’ as his explanati<strong>on</strong> of the great differences in per capita income across countries,<br />
stating that any poorer countries that adopt relatively good ec<strong>on</strong>omic policies <strong>and</strong> instituti<strong>on</strong>s enjoy rapid<br />
catchup growth. This view should be taken seriously when we wish to underst<strong>and</strong> the technological catching<br />
up to American leadership in some key industries by Japan, in particular, during the post-war period <strong>and</strong><br />
explore the possibility of a shift in st<strong>and</strong>ing between these two countries. This c<strong>on</strong>siderati<strong>on</strong> will directly<br />
bear <strong>on</strong> the future trend of the state of the art which exerts a crucial influence <strong>on</strong> the development of the<br />
world ec<strong>on</strong>omy.<br />
On a macro scale this racing paradigm would suggest that the U.S. would mobilize all its technology/<br />
science/entrepreneurial resources to increase or keep its distance to other large emerging ec<strong>on</strong>omies (such<br />
as China or India). With science/ technology being an evoluti<strong>on</strong>ary cumulative enterprise, for a dominating<br />
country with a portfolio of major increasing returns industries, the odds of leapfrogging <strong>on</strong>eself are higher<br />
than being leapfrogged by close followers, thus this asymmetry could play a distinctive role. Abramovitz’<br />
advantage of backwardness may hold <strong>on</strong> up to a certain limit but with decreasing returns.<br />
In an interesting paper, under some seemingly reas<strong>on</strong>able assumpti<strong>on</strong>s <strong>on</strong> linear technology investment<br />
<strong>and</strong> dynamic equilibrium path of capital accumulati<strong>on</strong> in a neoclassical type model Lau <strong>and</strong> Wan (1993)<br />
obtain the following results which they argue are fully c<strong>on</strong>sistent with empirical growth ec<strong>on</strong>omics: These<br />
are: (a) Not all ec<strong>on</strong>omies c<strong>on</strong>verge in growth with each other. (b) Ec<strong>on</strong>omies with an initial technical<br />
capability will c<strong>on</strong>verge in growth with the advanced ec<strong>on</strong>omies. The difference in per capita output grows<br />
exp<strong>on</strong>entially, if the developing ec<strong>on</strong>omy engages <strong>on</strong>ly in imitati<strong>on</strong> (<strong>and</strong> not innovati<strong>on</strong>). (c) With an initial<br />
technical capability there is a ‘high growth’ period, preceded (followed) by a phase of ‘trend accelerati<strong>on</strong>’<br />
(‘trend decelerati<strong>on</strong>’). (d) With an initial technical capability the the technology gap widens forever. The<br />
previous analysis indicates that the possibilities opened through competitive industrial racing are far richer<br />
<strong>and</strong> more surprising than they would emerge from the macro-scale models.<br />
A dynamic competitive nati<strong>on</strong>al ec<strong>on</strong>omy with its underlying instituti<strong>on</strong>al factors is fed by a c<strong>on</strong>sistent<br />
industrial policy intrinsically supported by a strategic innovati<strong>on</strong> system. Although this view can explain the<br />
tendency to c<strong>on</strong>vergence of productivity levels of follower countries, it fails to answer the historical puzzle<br />
why a country, the United States, has preserved the st<strong>and</strong>ing of the technological leader for a l<strong>on</strong>g time this<br />
century, <strong>and</strong> why it has intermittently been threatened by Germany, Japan <strong>and</strong> lately by emerging China <strong>and</strong><br />
to a lesser degree by India. We show the empirical proliferati<strong>on</strong> of comparative growth <strong>and</strong> development<br />
over the main period of industrializati<strong>on</strong>. Why have the shifts taken place in the ranks of follower countries<br />
in their relative levels of productivity, i.e., technological gaps between them <strong>and</strong> the leader?<br />
XIV
1.<br />
INNOVATION<br />
AND<br />
TECHNOLOGY
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
HANS W. GOTTINGER<br />
3
HANS W. GOTTINGER<br />
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
4
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
HANS W. GOTTINGER<br />
5
HANS W. GOTTINGER<br />
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
6
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
HANS W. GOTTINGER<br />
7
HANS W. GOTTINGER<br />
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
8
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
HANS W. GOTTINGER<br />
9
HANS W. GOTTINGER<br />
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
10
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
HANS W. GOTTINGER<br />
11
HANS W. GOTTINGER<br />
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
12
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
HANS W. GOTTINGER<br />
13
HANS W. GOTTINGER<br />
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
14
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
HANS W. GOTTINGER<br />
15
HANS W. GOTTINGER<br />
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
16
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
HANS W. GOTTINGER<br />
17
HANS W. GOTTINGER<br />
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
18
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
HANS W. GOTTINGER<br />
19
HANS W. GOTTINGER<br />
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
20
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
HANS W. GOTTINGER<br />
21
HANS W. GOTTINGER<br />
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
22
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
HANS W. GOTTINGER<br />
23
HANS W. GOTTINGER<br />
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
24
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
HANS W. GOTTINGER<br />
25
HANS W. GOTTINGER<br />
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
26
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
HANS W. GOTTINGER<br />
27
HANS W. GOTTINGER<br />
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
28
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
HANS W. GOTTINGER<br />
29
HANS W. GOTTINGER<br />
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
30
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
HANS W. GOTTINGER<br />
31
HANS W. GOTTINGER<br />
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
32
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
HANS W. GOTTINGER<br />
33
HANS W. GOTTINGER<br />
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
34
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
HANS W. GOTTINGER<br />
35
HANS W. GOTTINGER<br />
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
36
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
HANS W. GOTTINGER<br />
37
HANS W. GOTTINGER<br />
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
38
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
HANS W. GOTTINGER<br />
39
HANS W. GOTTINGER<br />
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
40
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
HANS W. GOTTINGER<br />
41
HANS W. GOTTINGER<br />
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
42
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
HANS W. GOTTINGER<br />
43
HANS W. GOTTINGER<br />
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
44
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
HANS W. GOTTINGER<br />
45
HANS W. GOTTINGER<br />
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
46
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
HANS W. GOTTINGER<br />
47
HANS W. GOTTINGER<br />
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
48
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
HANS W. GOTTINGER<br />
49
HANS W. GOTTINGER<br />
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
50
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
HANS W. GOTTINGER<br />
51
HANS W. GOTTINGER<br />
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
52
1.4 GLOBAL TECHNOLOGICAL RACES<br />
HANS W. GOTTINGER<br />
53
HANS W. GOTTINGER<br />
1.4 GLOBAL TECHNOLOGICAL RACES<br />
54
1.4 GLOBAL TECHNOLOGICAL RACES<br />
HANS W. GOTTINGER<br />
55
HANS W. GOTTINGER<br />
1.4 GLOBAL TECHNOLOGICAL RACES<br />
56
1.4 GLOBAL TECHNOLOGICAL RACES<br />
HANS W. GOTTINGER<br />
57
HANS W. GOTTINGER<br />
1.4 GLOBAL TECHNOLOGICAL RACES<br />
58
1.4 GLOBAL TECHNOLOGICAL RACES<br />
HANS W. GOTTINGER<br />
59
HANS W. GOTTINGER<br />
1.4 GLOBAL TECHNOLOGICAL RACES<br />
60
1.4 GLOBAL TECHNOLOGICAL RACES<br />
HANS W. GOTTINGER<br />
61
HANS W. GOTTINGER<br />
1.4 GLOBAL TECHNOLOGICAL RACES<br />
62
1.4 GLOBAL TECHNOLOGICAL RACES<br />
HANS W. GOTTINGER<br />
63
HANS W. GOTTINGER<br />
1.4 GLOBAL TECHNOLOGICAL RACES<br />
64
1.4 GLOBAL TECHNOLOGICAL RACES<br />
HANS W. GOTTINGER<br />
65
HANS W. GOTTINGER<br />
1.4 GLOBAL TECHNOLOGICAL RACES<br />
66
240 Int. J. <strong>Technology</strong>, Policy <strong>and</strong> Management, Vol. 4, No. 3, 2004<br />
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
Stochastic racing <strong>and</strong> competiti<strong>on</strong> in network<br />
markets<br />
HANS W. GOTTINGER<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g>-Werner <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Institute of Management Sciences, Maastricht,<br />
Kwansei Gakuin University (KGU),<br />
Kobe-S<strong>and</strong>a Campus, Japan<br />
E-mail: bvz28020@ksc.kwansei.ac.jp<br />
Abstract: We look at a competitive situati<strong>on</strong> in network markets where there is<br />
uncertain technological development in product/process technologies. Firms<br />
‘price’ compete in those markets to gain market share before any of them<br />
succeeds in getting an innovati<strong>on</strong> to move ahead of its rival(s). If the firms are<br />
in a technology race <strong>and</strong> the probability of innovati<strong>on</strong> success is small, then a<br />
firm with a bigger network advantage is likely to attract more customers in the<br />
absence of innovati<strong>on</strong>. If, however, any of those firms expect innovati<strong>on</strong> with a<br />
high probability, <strong>and</strong> n<strong>on</strong>e of them have a big advantage, then they keep <strong>on</strong><br />
sharing the market until the innovati<strong>on</strong> occurs.<br />
Keywords: network ec<strong>on</strong>omics; market structure <strong>and</strong> pricing; firm strategy;<br />
market performance.<br />
Reference to this paper should be made as follows: <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H-W. (2004)<br />
‘Stochastic racing <strong>and</strong> competiti<strong>on</strong> in network markets’, Int. J. <strong>Technology</strong>,<br />
Policy <strong>and</strong> Management, Vol. 4, No. 3, pp.240–256.<br />
Biographical notes: Dr <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g> is Professor of Ec<strong>on</strong>omics at the University<br />
of Osaka, <strong>and</strong> with the Institute of Management Science, Maastricht. His fields<br />
of interest include network ec<strong>on</strong>omics, strategy <strong>and</strong> competiti<strong>on</strong> issues, energy<br />
<strong>and</strong> envir<strong>on</strong>mental ec<strong>on</strong>omics.<br />
1 Introducti<strong>on</strong><br />
Many high technology industries are characterised by positive network externalities.<br />
Firms essentially compete against each other for goods that share a network.<br />
Our model of competiti<strong>on</strong> c<strong>on</strong>tains two special features that apply equally well to<br />
network markets. One is the uncertainty in technological development or uncertainty in<br />
the realisati<strong>on</strong> of a firm’s R&D effort. The other is the dynamic nature of price<br />
competiti<strong>on</strong> between firms in the presence of network effects.<br />
Firms compete with each other over an extended period of time <strong>and</strong> must, therefore,<br />
strategically choose prices as the market shares of the firms evolve.<br />
Most existing models of network markets focus <strong>on</strong> <strong>on</strong>ly <strong>on</strong>e of the two features.<br />
Further, almost all the models have the feature that <strong>on</strong>e firm captures the market<br />
instantaneously <strong>and</strong> sells to all c<strong>on</strong>sumers from then <strong>on</strong> (the ‘winner-takes-all’ market).<br />
However, the history of the high technology industry abounds with instances where rival<br />
Copyright © 2004 Inderscience Enterprises Ltd.<br />
67
HANS W. GOTTINGER<br />
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
Stochastic racing <strong>and</strong> competiti<strong>on</strong> in network markets 241<br />
firms (<strong>and</strong> technologies) have had extended battles for providing the industry st<strong>and</strong>ard<br />
(Uttenback, 1994).<br />
In network markets, evoluti<strong>on</strong> of market share is a very interesting phenomen<strong>on</strong><br />
especially in the face of uncertainty about the future product quality. This paper also<br />
attempts to provide a framework that could capture the richness of the market share<br />
evoluti<strong>on</strong> in the presence of network externalities.<br />
This paper is a follow-up to a stochastic racing model (<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, 2002a): it also takes<br />
some cues <strong>on</strong> racing from the seminal c<strong>on</strong>tributi<strong>on</strong> by C. Futia (1980).<br />
In its strategic dimensi<strong>on</strong> it is closely related to Katz <strong>and</strong> Shapiro (1992) <strong>and</strong> Harris<br />
<strong>and</strong> Vickers (1987), particularly, in their ‘tug-of-war’ presentati<strong>on</strong> <strong>and</strong> multi-stage racing.<br />
Overall our approach is emerging from new ideas <strong>on</strong> anti-trust analysis in<br />
dynamically competitive industries (Evans <strong>and</strong> Schmalensee, 2001).<br />
It is illuminating to take a full quote out of their comprehensive analysis:<br />
“Firms that are not leaders in network industries generally have little hope of<br />
reaching that status unless they come up with a major innovati<strong>on</strong> – <strong>on</strong>e that can<br />
defeat the natural advantage that network effects bestow <strong>on</strong> the industry<br />
leaders. Incremental innovati<strong>on</strong> – making slight improvements in the leaders’<br />
products – will not enable a small firm to overtake a leader that enjoys the<br />
benefits of network ec<strong>on</strong>omies. Similarly, the possibility of being displaced by<br />
a major innovati<strong>on</strong> will shape leaders’ research agendas. If there is a chance<br />
that today’s products will be replaced by a major innovati<strong>on</strong>, a leader’s survival<br />
depends <strong>on</strong> bringing this innovati<strong>on</strong> to market <strong>and</strong> thereby replacing itself<br />
before others do. As a result, competiti<strong>on</strong> in network industries often involves<br />
intense R&D efforts aimed at capturing or retaining market leadership. It is not<br />
atypical for a fringe firm that invests heavily to displace the leader by<br />
leapfrogging the leader’s technology…” (Evans <strong>and</strong> Schmalensee, 2001,<br />
pp.10–12)<br />
We are exploring this problem of trade-off between ‘network dominance’ <strong>and</strong> ‘radical<br />
innovati<strong>on</strong>’ that could tip the market the other way, with the significant complexity added<br />
that breakthrough R&D is highly uncertain. From a strategic perspective, in this<br />
envir<strong>on</strong>ment, for any two firms of asymmetric size, both compete dynamically over<br />
prices to win market share. In this dynamic process there are two ways to achieve<br />
(temporary) m<strong>on</strong>opolistic status. The ‘smaller’ firm can use dynamic pricing competiti<strong>on</strong><br />
to delay the time in which the ‘larger’ firm wins a critical market share in the hope to hit<br />
the innovati<strong>on</strong> first <strong>and</strong> displace it. If the probability of innovati<strong>on</strong> <strong>and</strong> the discount factor<br />
are sufficiently high, there is an equilibrium in which duopoly persists (no firm achieves<br />
a critical market share) until <strong>on</strong>e of the two firms wins the race for innovati<strong>on</strong>.<br />
We take a historical example from markets for IT products. In the pers<strong>on</strong>al computer<br />
(PC) operating system market. Microsoft products (MS DOS <strong>and</strong> Windows) have been<br />
dominating the market since the mid 1980s. An operating system is the fundamental<br />
programme that c<strong>on</strong>trols the allocati<strong>on</strong> <strong>and</strong> use of computer resources. Thus, the utility<br />
that operating systems provide to c<strong>on</strong>sumers depends <strong>on</strong> the number of compatible<br />
applicati<strong>on</strong>s. As a general rule, an applicati<strong>on</strong> that relies <strong>on</strong> a specific operating system<br />
will not functi<strong>on</strong> <strong>on</strong> another operating system unless it is ported to that specific operating<br />
system. Therefore, because of its dominance, the majority of applicati<strong>on</strong>s have been<br />
written to run <strong>on</strong> Microsoft operating systems (MS DOS). The dominati<strong>on</strong> of MS DOS<br />
has become even str<strong>on</strong>ger since the arrival of Windows 95 in the PC operating system<br />
market. This, in turn, has provided a great indirect positive network externality to PC<br />
owners who adopted Windows 95 as an operating system. Many other firms, such as IBM<br />
68
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
HANS W. GOTTINGER<br />
242 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
<strong>and</strong> Be Inc., introduced their own operating systems <strong>and</strong> tried to compete with Windows.<br />
These products, however, lacked sufficient compatible applicati<strong>on</strong>s to efficiently compete<br />
with Microsoft products. The lack of compatible applicati<strong>on</strong>s prevented enough<br />
applicati<strong>on</strong> developers <strong>and</strong> c<strong>on</strong>sumers from regarding OS/2 Warp or BeOS as a viable<br />
alternative to the dominant incumbent, Windows. This obstacle prevented these potential<br />
entrants from obtaining a sizeable market share. Their failure to enter the market<br />
successfully, however, was not due to the inferior quality of their operating systems.<br />
In fact, OS/2 Warp is reported to be at least as good as Windows, <strong>and</strong> BeOS offers<br />
superior support for multimedia applicati<strong>on</strong>s. If c<strong>on</strong>sumers who use multimedia<br />
applicati<strong>on</strong>s frequently adopt Windows at the expense of BeOS, they have to give up the<br />
c<strong>on</strong>venience that is provided by BeOS. Thus, for multimedia specific users, adopting<br />
BeOS as their operating system at the expense of another operating system might provide<br />
the highest utility. Nevertheless, the lack of compatible applicati<strong>on</strong>s, which in turn<br />
implies the lack of positive network externality, has prevented c<strong>on</strong>sumers from adopting<br />
OS/2 Warp or BeOS. As a c<strong>on</strong>clusi<strong>on</strong>, the positive network externality for the dominant<br />
incumbent (Windows) has worked as an entry barrier against entrants (OS/2 Warp or<br />
BeOS), which do not have network externality. Such an entry barrier can <strong>on</strong>ly be<br />
overcome by a radical innovati<strong>on</strong>, virtually leapfrogging the dominant incumbent.<br />
2 Market histories<br />
We substantiate our claim <strong>on</strong> a natural tradeoff between network size <strong>and</strong> quality change<br />
(through significant technological change) by looking independently <strong>on</strong> four market<br />
histories for the IT industry.<br />
2.1 Internet explorer<br />
The first versi<strong>on</strong> of Internet Explorer was introduced in the web browser market in July<br />
1995. This versi<strong>on</strong> of Internet Explorer was reportedly inferior to Netscape’s Navigator.<br />
Microsoft made many improvements in its next versi<strong>on</strong>, Internet Explorer 3.0, in August<br />
1996. Internet Explorer 3.0 matched nearly all the features in Navigator. However,<br />
Microsoft was not able to draw significant market share at the expense of Navigator with<br />
this enhanced versi<strong>on</strong>. Internet Explorer 4.0, released in late 1997, finally managed to<br />
obtain roughly the same number of reviewers who preferred it to Navigator as the number<br />
of those who preferred Navigator. In 1996, most c<strong>on</strong>sumers were much more familiar<br />
with Navigator, the dominant web browser than with Internet Explorer. Thus although the<br />
quality of Internet Explorer was much improved since its release into the market, more<br />
than 80% of c<strong>on</strong>sumers were still using Navigator until late 1996. It was evident that<br />
Internet Explorer could not draw a significant percentage of c<strong>on</strong>sumers away from<br />
Navigator by simply improving the quality of its web browser. Senior executives within<br />
Microsoft c<strong>on</strong>cluded that Internet Explorer could not catch up with Navigator’s market<br />
share by itself. J. Allchin, senior executive within Microsoft, wrote to P. Martiz, in late<br />
December 1996:<br />
“I d<strong>on</strong>’t underst<strong>and</strong> how IE (Internet Explorer) is going to win. The current<br />
path is simply to copy everything that Netscape does, packaging <strong>and</strong> product<br />
wise. Let’s {suppose} IE is as good as Navigator/Communicator. Who wins?<br />
The <strong>on</strong>e with 80% market share. Maybe being free helps us, but <strong>on</strong>ce people<br />
are used to a product it is hard to change them. C<strong>on</strong>sider Office. We are most<br />
expensive today, but we’re still winning.”<br />
69
HANS W. GOTTINGER<br />
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
Stochastic racing <strong>and</strong> competiti<strong>on</strong> in network markets 243<br />
From 1996 to 1998, Internet Explorer’s market share increased dramatically.<br />
The estimated market share of Navigator fell from 80% in January 1996 to 55% in<br />
November 1997, while that of Internet Explorer increased from 5% to 36% over the same<br />
period. These changes in market shares, however, were mainly due to the fact that<br />
Microsoft leveraged Windows to increase the usage of Internet Explorer, rather than<br />
c<strong>on</strong>stantly improving the quality of its browser.<br />
2.2 Server operating system<br />
The envir<strong>on</strong>ment of the server operating system market is different from that of the<br />
PC-operating system in the sense that a server is maintained by a group of computer<br />
experts (not by novices as in the case of a PC) <strong>and</strong> is intended for specific purposes,<br />
rather than for general uses. Moreover, a choice of server operating systems requires<br />
many technical features, such as functi<strong>on</strong>ality, system management, performance,<br />
reliability (stability) <strong>and</strong> security. Functi<strong>on</strong>ality <strong>and</strong> performance are important technical<br />
features for both PCs <strong>and</strong> servers, but other aspects are not very important for PCs. This<br />
difference comes from the fact that while a PC is designed for use by <strong>on</strong>e pers<strong>on</strong> at a<br />
time, a server is designed to provide data, services, <strong>and</strong> functi<strong>on</strong>ality through a digital<br />
network to multiple users. Most individual PC owners do not have an incentive to buy a<br />
server-operating system, since server operating systems generally cannot functi<strong>on</strong><br />
efficiently <strong>on</strong> PC hardware. In other words, the c<strong>on</strong>sumers in the two markets are<br />
basically separated. According to Data Quest, the Unix System is currently the most<br />
widely used operating system platform. It exp<strong>and</strong>ed its market share from 36% in 1996 to<br />
43% in 1998. The newcomer in the server operating system market, Windows NT, also<br />
achieved great success during that period; its market share rose from 10% in 1996 to 16%<br />
in 1998. Linux licence shipments grew at a rate of 212.5%, accounting for more than<br />
17% of all server operating system units shipped, Other than Unix servers <strong>and</strong> Windows<br />
NT, Open VMS <strong>and</strong> Novel.<br />
NetWare <strong>and</strong> OS/2 are the major competitors in the server-operating system market.<br />
A small product differentiati<strong>on</strong> in a server-operating system enables it to create <strong>and</strong> to<br />
maintain its own dem<strong>and</strong>. In sum, a dominant product cannot eliminate other products<br />
with its prevailing positive network externalities when substitutability for products are<br />
low.<br />
2.3 OS/2 warp<br />
In late 1994, IBM introduced its Intel-<strong>on</strong>ly compatible operating system, OS/2 Warp, as<br />
an alternative to Windows. OS/2 was originally designed to be optimised for Intel<br />
microprocessors. This resulted in a lack of portability. Because the Intel compatible PC<br />
dominated the PC market, the lack of portability did not prevent c<strong>on</strong>sumers from<br />
adopting OS/2 as an operating system. Furthermore, OS/2 was not reported to be inferior<br />
to Windows. The greatest disadvantage of adopting OS/2 as an operating sytem was the<br />
lack of compatible software; approximately 2,500 applicati<strong>on</strong>s were available. In 1995,<br />
IBM still attempted to compete with Windows 95 in the PC-operating system market<br />
with its release of OS/2 2.0. However, IBM figured that it would lose between 70% <strong>and</strong><br />
90% of its sales volume if it installed OS/2 to the exclusi<strong>on</strong> of Windows 95. In 1996,<br />
IBM withdrew from the PC-operating system market <strong>and</strong>, instead, has been repositi<strong>on</strong>ing<br />
OS/2 for the server-operating system market. The failure of OS/2 Warp to gain enough<br />
70
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
HANS W. GOTTINGER<br />
244 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
market share to survive is mainly due to its lack of applicati<strong>on</strong>s rather than inferiority to<br />
Windows. By the time OS/2 Warp was released, Windows was already the predominant<br />
operating system <strong>and</strong> ran an overwhelming number of applicati<strong>on</strong>s. Most c<strong>on</strong>sumers used<br />
Windows not because it was superior in quality to its competitors, but rather because it<br />
provided the greatest positive network externality am<strong>on</strong>g the available PC-operating<br />
systems.<br />
In other words, Microsoft successfully deterred entry into the PC-operating system<br />
market by using its predominating network externality.<br />
2.4 BeOS<br />
Be, Inc., founded in 1990 by a former executive of Apple’s product divisi<strong>on</strong>, released its<br />
cross-platform operating system, BeOS. BeOS functi<strong>on</strong>s virtually identically <strong>on</strong> both<br />
Intel-compatible PCs <strong>and</strong> Macintosh systems. In additi<strong>on</strong>, BeOS is optimised for<br />
multiprocessor media-based systems. These two features provide great c<strong>on</strong>venience for<br />
multimedia-specific users. Be has stated from the beginning that it intends to coexist<br />
within the Windows <strong>and</strong> Apple World as a specialised operating system. Even taking this<br />
specific purpose into account, the market share of BeOS is too small compared to the<br />
number of Windows users. The most outst<strong>and</strong>ing disadvantage is the lack of compatible<br />
applicati<strong>on</strong>s (approx. 1,000). Be, Inc. tried to overcome this problem by maintaining a<br />
website which provided a list of freeware <strong>and</strong> shareware applicati<strong>on</strong>s. Be, Inc. failed to<br />
gain sizeable market share, <strong>and</strong> was eventually sold to Palm. However, Palm c<strong>on</strong>tinues to<br />
produce BeOS <strong>and</strong> to maintain its BeWare website. BeOS still exists as a niche operating<br />
system, in spite of its superior support for multimedia applicati<strong>on</strong>s. The failure of BeOS<br />
to obtain a significant market share is evidence in support of the propositi<strong>on</strong> that the<br />
incumbent can deter entry by using its network externality.<br />
3 Some issues of technological racing<br />
Katz <strong>and</strong> Shapiro (1992) look at whether there is too much or too little technological<br />
innovati<strong>on</strong> in a market with network externalities, compared to the social optimum. In<br />
particular, they explore whether a new product, which embodies technological progress,<br />
is introduced too early or too late compared to the social optimum. In their model,<br />
technological progress is deterministic. Only the entrant gets the benefit of improving<br />
technology, but <strong>on</strong>ce it enters the market its product is fixed. Therefore, in any<br />
equilibrium of their model, either the new firm captures the market as so<strong>on</strong> as it enters or<br />
it does not enter at all. Their model is thus not able to capture the richness of pricing<br />
strategies <strong>and</strong> market evoluti<strong>on</strong> as a functi<strong>on</strong> of market shares. For instance, it is unable<br />
to explain the phenomen<strong>on</strong> of two competing technologies fighting for superiority in a<br />
market when both are making sales until <strong>on</strong>e firm finally leaves the market. Harris <strong>and</strong><br />
Vickers (1987) present two models of races in which there is both technological<br />
uncertainty <strong>and</strong> strategic interacti<strong>on</strong> between competitors as the race unfolds.<br />
Their aim is to see how the equilibrium effort levels of competitors vary with the<br />
intensity of rivalry between them. They show that under certain c<strong>on</strong>diti<strong>on</strong>s the leader in<br />
the race makes greater efforts than the follower, <strong>and</strong> efforts increase as the gap between<br />
competitors in their progress decreases. The model presented here can be seen al<strong>on</strong>g<br />
71
HANS W. GOTTINGER<br />
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
Stochastic racing <strong>and</strong> competiti<strong>on</strong> in network markets 245<br />
similar lines to that of Harris <strong>and</strong> Vickers, but more realistically, with the network<br />
advantage as the state variable instead of progress, <strong>and</strong> prices providing strategic<br />
interacti<strong>on</strong> between firms instead of the amount of effort put into R&D. While there is<br />
<strong>on</strong>e ultimate goal for each firm in the multistage race of Harris <strong>and</strong> Vickers to reach the<br />
finishing line of the race first, we leave it open for firms to decide which network they<br />
want to build. It is this feature that can lend profit maximisati<strong>on</strong> by firms to a complex<br />
network evoluti<strong>on</strong> pattern. Also, the amount of effort put into R&D stochastically moves<br />
firms to a new stage in the race, whereas in our model prices – the variables of strategic<br />
interacti<strong>on</strong> – deterministically move firms to a new market share.<br />
While a lot of work has been d<strong>on</strong>e in the area of R&D, very little of it specifically<br />
addresses markets with network externalities. Katz <strong>and</strong> Shapiro look at the choice of<br />
product compatibility in a market with technological progress. They look at both the<br />
cases of deterministic <strong>and</strong> uncertain technological progress. However, technological<br />
progress in their model is exogenous <strong>and</strong> they look at a two-period game, which does not<br />
provide a very suitable framework for looking at the dynamic pricing strategy of firms<br />
<strong>and</strong> the resulting evoluti<strong>on</strong> of market shares. Choi (1994) looks at a two-period model of<br />
a m<strong>on</strong>opoly in a market with network externalities. He studies the incentive of the<br />
m<strong>on</strong>opolist to introduce an incompatible improved product in the presence of network<br />
externalities. Kristiansen (1998) studies the c<strong>on</strong>sequences of network externalities <strong>on</strong> the<br />
riskiness of R&D projects chosen by an entrant <strong>and</strong> an incumbent. He shows that the<br />
incumbent chooses a risky project that very often lets a new firm with an incompatible<br />
technology enter. In additi<strong>on</strong>, the entrant has an incentive to choose more certain projects<br />
than are socially optimal <strong>and</strong> these strengthen the possibility of adopti<strong>on</strong> of an<br />
incompatible technology.<br />
This paper develops <strong>and</strong> analyses a model of technological uncertainty where firms<br />
compete strategically in prices in the presence of network externalities. Technological<br />
uncertainty is modelled as an improvement in the quality of the firm’s product.<br />
Whether the firm gets the innovati<strong>on</strong>, however, is uncertain. In the simplest model,<br />
<strong>on</strong>ce a firm gets the innovati<strong>on</strong>, the race ends <strong>and</strong> there is no further innovati<strong>on</strong>.<br />
Once the innovati<strong>on</strong> occurs, when there is no uncertainty, the model looks like that of<br />
Katz <strong>and</strong> Shapiro – the firm with the quality plus network advantage sells to all the<br />
c<strong>on</strong>sumers subsequently. We attempt to analyse how firms share the market by<br />
strategically choosing their prices before they succeed in getting an innovati<strong>on</strong>.<br />
The advantage of a firm is measured by the sum of its network size <strong>and</strong> the quality of<br />
its product. If the firms are in a technological race but the probability of innovati<strong>on</strong> is<br />
small, then again, in the absence of innovati<strong>on</strong>, the firm with the bigger advantage sells to<br />
all the c<strong>on</strong>sumers. If, however, those firms expect innovati<strong>on</strong> with a high probability then<br />
so l<strong>on</strong>g as neither firm has a very big advantage both firms take turns in selling to the<br />
c<strong>on</strong>sumers in the absence of innovati<strong>on</strong>. Thus, if the firms are <strong>on</strong> an equal base, neither<br />
firm gets pushed out unless <strong>on</strong>e of the firms gets an insurmountable advantage in the<br />
process. That is because with firms that are sufficiently close in terms of their advantage,<br />
neither firm finds it worthwhile to push the other firm out of the market since the ‘losing’<br />
firm prices more aggressively than the ‘leader’ as it has more to gain by staying in the<br />
market.<br />
72
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
HANS W. GOTTINGER<br />
246 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
4 Asymmetric R&D races<br />
We provide some anecdotal observati<strong>on</strong>s across various industries that show a str<strong>on</strong>g<br />
rivalry between a leader in the field <strong>and</strong> the nearest rival(s) <strong>on</strong> the technological fr<strong>on</strong>t <strong>and</strong><br />
its efforts to gain market share.<br />
We identify those interactive patterns as technological races. We can even extend<br />
those observati<strong>on</strong>s to megamergers <strong>and</strong> alliances as products of network ec<strong>on</strong>omies that<br />
did not protect those networks from relentless technological attacks by upstarts, new<br />
entrants or Schumpeterian entrepreneurs hailing ‘creative destructi<strong>on</strong>’. This asymmetric<br />
view reflects the match between the ‘Old Guard vs the Vanguard’ (Hamel <strong>and</strong> Switzer,<br />
2004) in the sense that an incumbent’s positi<strong>on</strong> endows her with a natural network<br />
advantage which could <strong>on</strong>ly be overcome by a breakthrough innovati<strong>on</strong> of a new entrant<br />
depending <strong>on</strong> some probability of innovati<strong>on</strong> success. We have manifestati<strong>on</strong>s of intense<br />
rivalry <strong>and</strong> ‘head-to-head’ competiti<strong>on</strong> across industries, be it advanced microprocessors<br />
between Intel <strong>and</strong> American Micro Devices (AMD), (Markoff <strong>and</strong> Lohr 2003), specialty<br />
pharmaceuticals against chr<strong>on</strong>ic diseases between Merck, Glaxo <strong>and</strong> Pfizer, <strong>and</strong> between<br />
biotech companies Amgen <strong>and</strong> Biogen. In the exp<strong>and</strong>ing market of software related web<br />
services it is Microsoft against IBM, Sun Microsystems <strong>and</strong> Oracle (Lohr, 2003), in<br />
servers it is IBM vs Sun <strong>and</strong> H-P (Clark, 2003) in c<strong>on</strong>sumer electr<strong>on</strong>ics <strong>and</strong> design it is<br />
S<strong>on</strong>y against Matsushita or Sanyo (Belsen, 2003).<br />
5 Modelling a race in a network<br />
We start out with a simple model in which two firms, Y <strong>and</strong> Z, are manufacturing a<br />
product using incompatible technologies: the product is infinitely durable. The firms have<br />
products of identical qualities that we normalise to zero. For simplicity the marginal cost<br />
of producti<strong>on</strong> is assumed to be zero for both firms.<br />
We define: The network base of firm i at time t, b it , is the total number of c<strong>on</strong>sumers<br />
it has sold to until time t, i ∈ {Y, Z}.<br />
The network advantage of firm i at time t, n it ≡ b it – b jt , i, j∈ {Y, Z}.<br />
A new c<strong>on</strong>sumer arrives in the market periodically. Each c<strong>on</strong>sumer has an inelastic<br />
unit dem<strong>and</strong> for the product <strong>and</strong> must decide the firm from which to buy the product as<br />
so<strong>on</strong> as he/she enters the market. C<strong>on</strong>sumers care about how many other c<strong>on</strong>sumers buy<br />
the same product because of network externalities. That is, they are likely to choose<br />
products of the larger network over those of the smaller. They also care about the quality<br />
of the product <strong>and</strong> its price. For simplicity, all ‘representative’ c<strong>on</strong>sumers have identical<br />
preferences. We may state their utility functi<strong>on</strong>s as u (n t , q t, p t ) where q t is the quality of<br />
the product as a perfect substitute for the network advantage <strong>and</strong> p t is its price.<br />
Throughout the phase of adding uncertainty to the innovati<strong>on</strong> process the c<strong>on</strong>sumers are<br />
expected to be risk-neutral. We assume that there is a threshold difference, D, in network<br />
sizes of the two firms such that if neither firm’s network is D (critically bigger than the<br />
other’s) then c<strong>on</strong>sumers get the same network benefits from the two firms. More<br />
realistically, <strong>on</strong>e could think of D not just as a number but also as an ‘order of<br />
magnitude’. Therefore, if neither firm has a network advantage of D then c<strong>on</strong>sumers buy<br />
from the firm which charges a lower price if the quality is the same. C<strong>on</strong>sumers are<br />
indifferent about the two firms if they charge the same price. Once a firm establishes a<br />
73
HANS W. GOTTINGER<br />
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
Stochastic racing <strong>and</strong> competiti<strong>on</strong> in network markets 247<br />
network advantage of D over the other firm, c<strong>on</strong>sumers get a much bigger benefit from it.<br />
Thus, if both firms charge the same price then c<strong>on</strong>sumers prefer the firm that has the<br />
network advantage of D (or more). It follows that <strong>on</strong>ce a firm gets a network advantage<br />
of D it becomes a (temporary) m<strong>on</strong>opolist of the market.<br />
If both firms did not have identical qualities to start out with, then we could think of<br />
the network advantage as comprising the network difference plus the quality difference,<br />
that is, treat network <strong>and</strong> quality as perfect substitutes in the c<strong>on</strong>sumer’s utility.<br />
The proposed rule for c<strong>on</strong>sumer decisi<strong>on</strong>s implies that c<strong>on</strong>sumers are myopic, that is,<br />
the current base of c<strong>on</strong>sumers with the firms is what matters to them. While in practice<br />
we would expect c<strong>on</strong>sumers to form expectati<strong>on</strong>s about the future networks of firms, <strong>and</strong><br />
thus introduce the possibility of multiple fulfilled-expectati<strong>on</strong>s equilibria, our assumpti<strong>on</strong><br />
of myopic c<strong>on</strong>sumers would help select the most efficient equilibrium outcome am<strong>on</strong>g<br />
them. Both assumpti<strong>on</strong>s – of myopic c<strong>on</strong>sumers <strong>and</strong> c<strong>on</strong>sumers with fully rati<strong>on</strong>al<br />
expectati<strong>on</strong>s – are extreme <strong>on</strong>es, <strong>and</strong> what is more realistic is something in between the<br />
two. Having myopic c<strong>on</strong>sumers has the added advantage of making the analysis more<br />
tractable.<br />
Let n Yt represent the state of the system (we could equivalently choose n Zt to be the<br />
state variable). Thus state 1 denotes the situati<strong>on</strong> where firm Y has <strong>on</strong>e (order of<br />
magnitude) more c<strong>on</strong>sumer than firm Z, state – 1 denotes the situati<strong>on</strong> where firm Z has<br />
<strong>on</strong>e (order of magnitude) more c<strong>on</strong>sumer than firm Y. We define 3 to be the absorbing<br />
state for firm Y <strong>and</strong> – 3 the absorbing state for firm Z, that is we let D equal 3. The net<br />
discounted payoff of a firm in its absorbing state is M, for the other firm likewise. The<br />
absorbing state can be thought of as the threshold advantage in quality or network, which<br />
is big enough to deter the other firm from ever trying to catch up. Thus the firm that has<br />
this advantage becomes a m<strong>on</strong>opolist in the market. It is the sole seller in the market<br />
making m<strong>on</strong>opoly profits of M whereas the other firm leaves the market <strong>and</strong> thus gets<br />
zero profits.<br />
The product is subject to technological innovati<strong>on</strong>. We assume that an innovati<strong>on</strong><br />
pushes a firm to its absorbing state. Thus, we assume that the innovati<strong>on</strong> is equivalent to<br />
a big enough network advantage that gives a firm an insurmountable lead over the other<br />
firm, in many given cases of asymmetrical network advantage the innovati<strong>on</strong>s need to be<br />
‘radical’. The innovati<strong>on</strong>, however, is uncertain, either a firm gets the required quality<br />
improvement or there is no improvement at all. The probability that a firm gets the<br />
innovati<strong>on</strong> in any time period is α ∈ [0,1/2] that is the same for both firms. We could<br />
think of α > 0 as representing the technological envir<strong>on</strong>ment in which the firms operate.<br />
We assume that <strong>on</strong>ce <strong>on</strong>e firm succeeds, there are no further innovati<strong>on</strong>s in the market.<br />
We also assume for notati<strong>on</strong>al simplicity that at most <strong>on</strong>e firm gets the innovati<strong>on</strong>. Thus,<br />
the probability that neither firm gets the innovati<strong>on</strong> in any <strong>on</strong>e time is 1 – 2α. In this<br />
situati<strong>on</strong>, we look at the innovati<strong>on</strong> process as ‘exogenous’ (though we could make it<br />
‘endogenous’ by letting the probability of innovati<strong>on</strong> success depend <strong>on</strong> the network size<br />
as a proxy for the degree of accumulated knowledge (learning) in determining innovati<strong>on</strong><br />
success).<br />
We could think of the technological innovati<strong>on</strong> as a new feature that the firms try to<br />
introduce in their respective products. Once <strong>on</strong>e firm gets its innovati<strong>on</strong>, it can get a<br />
patent even though the firms produce incompatible products. We assume that the new<br />
improved product that a firm might introduce is compatible with its older product.<br />
74
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
HANS W. GOTTINGER<br />
248 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Compatibility, in this model, translates into network sharing – compatible products share<br />
a network.<br />
Firms compete in prices <strong>and</strong> thus must decide what price to charge at each of the<br />
states in S = {–2, –1,0, 1,2}, in order to maximise their respective profits given by<br />
π Y (s 0 ,p) = ∑ t δ t (1–2α) t–1 [(1–2α)p Yt x Yt + αM] + δ τ+1 I(s τ+1 = 3)M, t = 0,1,2, …. , τ<br />
π Z (s 0 ,p) = ∑ t δ t (1–2α) t–1 [(1–2α)p Zt x Zt + αM] + δ τ+1 I(s τ+1 = 3)M, t = 0,1,2, …. , τ<br />
x<br />
it<br />
s<br />
t+<br />
1<br />
⎧1<br />
⎪<br />
= ⎨0<br />
⎪<br />
⎩<br />
0 or 1<br />
= s<br />
t<br />
+ x<br />
if p<br />
if p<br />
if p<br />
Yt<br />
it<br />
it<br />
it<br />
,<br />
< p<br />
> p<br />
= p<br />
jt<br />
jt<br />
jt<br />
where p it is the price charged by firm i in period t, x it ∈ {0, 1} is the number of period t<br />
c<strong>on</strong>sumers that the firm i sells to: τ + 1 is the time period in which a firm reaches its<br />
absorbing state even without getting the innovati<strong>on</strong>: s 0 ∈ S is the initial state; <strong>and</strong><br />
p = (p 1 , p 2 ), where<br />
p i = { p it }, t = 0,1,2,...,τ.<br />
We thus have a game where the firms are identical except in the sizes of their networks.<br />
There is uncertain technological innovati<strong>on</strong> but the size of the innovati<strong>on</strong> <strong>and</strong> the<br />
probability of innovati<strong>on</strong> are again identical for both firms. Therefore, by comparing the<br />
soluti<strong>on</strong> of a game with no uncertainty with the soluti<strong>on</strong> of our game, we should be able<br />
to gain valuable insights into the effects uncertainty has <strong>on</strong> the competiti<strong>on</strong> between two<br />
firms in the presence of network externalities.<br />
We look for Markov equilibria (Fudenberg <strong>and</strong> Tirole 1993), which are<br />
n<strong>on</strong>-collusive <strong>and</strong> symmetric. In such an equilibrium we c<strong>on</strong>sider a firm that is indifferent<br />
about selling or not selling. Thus if firm Y is the firm not selling at state s, then<br />
p Y (s) + δ ∏ Y (s+1) = δ ∏ Y (s–1)<br />
where ∏(s) represents the c<strong>on</strong>tinuati<strong>on</strong> payoff starting from state s. If Z is the firm not<br />
selling at state s then<br />
p Z (s) + δ ∏ Z (s–1) = δ ∏ Z (s+1).<br />
In such an equilibrium the n<strong>on</strong>-selling firm ensures that even if against all expectati<strong>on</strong>s<br />
the c<strong>on</strong>sumer buys its product then it will not make negative profits in equilibrium.<br />
Definiti<strong>on</strong> 1 An equilibrium is n<strong>on</strong>-collusive if neither firm charges a positive price<br />
at a n<strong>on</strong>-absorbing state. Alternatively, if either firm charges a positive price at a<br />
n<strong>on</strong>-absorbing state, equilibrium is collusive.<br />
Definiti<strong>on</strong> 2 An equilibrium is symmetric if p Y (s) = p Z (–s) for all permissible s <strong>and</strong><br />
x Y (s) = x Z (–s) for all permissible s ≠ 0.<br />
If instead of having an infinite horiz<strong>on</strong> game, we look at a game with a l<strong>on</strong>g but finite<br />
horiz<strong>on</strong>, the n<strong>on</strong>-collusive equilibria are the <strong>on</strong>es we expect would most likely be<br />
sustainable.<br />
We first c<strong>on</strong>sider the case in absence of uncertainty, that is, α = 0.<br />
75
HANS W. GOTTINGER<br />
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
Stochastic racing <strong>and</strong> competiti<strong>on</strong> in network markets 249<br />
Propositi<strong>on</strong> 1 For α = 0, the symmetric n<strong>on</strong>-collusive Markov equilibrium prices are<br />
unique: the firm with a positive network advantage makes all the sales in equilibrium.<br />
Proof: Suppose firm Z sells at s = 2. If firm Y were to sell to the c<strong>on</strong>sumer at s = 2,<br />
then in the next period it would get a profit of M. Therefore, if in a n<strong>on</strong>-collusive<br />
equilibrium firm Y does not sell at s = 2 as we have assumed, it must be getting a<br />
c<strong>on</strong>tinuati<strong>on</strong> payoff at least as big as M next period. There are two possibilities for s = 1:<br />
either firm Y sells or firm Z does.<br />
Suppose firm Y sells at s = 1. Then <strong>on</strong>ce the state reaches 1, it must alternate between<br />
1 <strong>and</strong> 2 forever more from then <strong>on</strong>. The c<strong>on</strong>tinuati<strong>on</strong> payoff to firm Y at state 1 then is<br />
just the discounted sum of the price it charges to the c<strong>on</strong>sumer at s = 1 every other period.<br />
For this to be at least M, the price at s = 1 must be positive <strong>and</strong> equilibrium is then<br />
collusive. Suppose firm Z sells at s = 1 also. Then <strong>on</strong>ce s = 1 is reached, it must remain<br />
within the states –1, 0, 1 forever more. But then again for firm Y’s c<strong>on</strong>tinuati<strong>on</strong> payoff at<br />
s = 1 to be at least M, it must get a positive payoff at s = 0 or –1, a c<strong>on</strong>tradicti<strong>on</strong>. Thus, Y<br />
must sell at s = 2. Suppose firm Z sells at s = 1 <strong>and</strong> Y sells at s = 2. Then <strong>on</strong>ce state 1 is<br />
reached, the state must remain within the states –1, 0, 1. In a n<strong>on</strong>-collusive equilibrium<br />
where neither firm charges a positive price, Y’s c<strong>on</strong>tinuati<strong>on</strong> payoff at s = 0 cannot be<br />
positive. Z’s c<strong>on</strong>tinuati<strong>on</strong> payoff at s = 2 is zero. Similarly, the smallest price Z is willing<br />
to charge at s = 1 cannot be less than zero. However, firm Y’s c<strong>on</strong>tinuati<strong>on</strong> payoff at s =<br />
2 is then positive <strong>and</strong> at most zero at s = 0, it is therefore willing to charge a negative<br />
price at s = 1 in order to lure the c<strong>on</strong>sumer into buying its product. But that means the<br />
c<strong>on</strong>sumer should buy from firm Y at s = 1 since its price is lower than Z’s, a<br />
c<strong>on</strong>tradicti<strong>on</strong>. Thus firm Y must sell at s = 1 also.<br />
We have so far shown that if a symmetric, n<strong>on</strong>-collusive Markov equilibrium exists,<br />
then it must have the firm with the network advantage selling. We now show that such an<br />
equilibrium does indeed exist <strong>and</strong> that it is (almost) unique.<br />
If in an equilibrium the firm with the network advantage always wins then the<br />
c<strong>on</strong>tinuati<strong>on</strong> payoffs of both firms at s = 0 must be zero in a symmetric equilibrium since<br />
both firms are willing to give up exactly the same surplus in order to win the c<strong>on</strong>sumer.<br />
This surplus is equal to the c<strong>on</strong>tinuati<strong>on</strong> payoff from winning the current, which has to be<br />
at least zero.<br />
At s = 1, the lowest firm Z is willing to charge is zero since whether it sells to the<br />
current c<strong>on</strong>sumer or not, it is going to get a payoff of zero from then <strong>on</strong>. But that means it<br />
must charge a price of zero at s = 2 also. The c<strong>on</strong>tinuati<strong>on</strong> payoff at s = 0 from winning is<br />
then δ 2 M.<br />
Thus, both firms charge –δ 3 M at s = 0 <strong>and</strong> get zero discounted payoffs. It can be<br />
checked that these are equilibrium prices.<br />
Thus, the unique equilibrium prices are as follows:<br />
p(2) = p(–2) = 0<br />
p(1) = p(–1) = 0<br />
p(0) = – δ 3 M.<br />
Both firms charge the same prices. Firm Y sells to the c<strong>on</strong>sumer at s = 1, 2,... firm Z sells<br />
to the c<strong>on</strong>sumer at s = –1, –2, either firm could sell at s = 0.<br />
Thus, we see that in the absence of uncertainty, the firm that is already ahead in terms<br />
of network size is the ‘winning’ firm. The other firm that has a network disadvantage is<br />
76
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
HANS W. GOTTINGER<br />
250 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
unable to overtake the firm that is ahead. This result is similar to that obtained by Katz<br />
<strong>and</strong> Shapiro (1992). Once a firm acquires a bigger network, the other firm is unable to<br />
overcome its disadvantage. Since in our model, quality <strong>and</strong> network are perfect<br />
substitutes in the c<strong>on</strong>sumer’s utility, we could replace network by quality to get the same<br />
result. The firm with the higher quality sells to all c<strong>on</strong>sumers, the lower quality firm is<br />
unable to push out the better quality firm from the market, given that both firms start out<br />
with the same network size.<br />
We now introduce uncertainty into the game <strong>and</strong> see how firms behave. As in the<br />
previous case, the <strong>on</strong>ly difference between the firms is in their networks. The uncertainty<br />
faced by firms is symmetric: both firms have the same probability of getting an<br />
innovati<strong>on</strong> <strong>and</strong> the effect of an innovati<strong>on</strong> is symmetric for both firms.<br />
Propositi<strong>on</strong> 2 There exist δ * <strong>and</strong> α * such that for δ * < δ < 1 <strong>and</strong> 0 < α < α * the<br />
n<strong>on</strong>-collusive symmetric equilibrium prices are unique. The firm with the bigger network<br />
sells to the c<strong>on</strong>sumer, either firm could sell when s = 0.<br />
Proof:<br />
The proof proceeds al<strong>on</strong>g a sequence of Lemmas.<br />
Since we are looking for symmetric equilibria, at s = 0 the lowest price both firms are<br />
willing to charge must be the same. Thus, in equilibrium both firms must be indifferent<br />
between selling or not at s = 0. In other words, at s = 0 both firms have the expected<br />
payoffs. Again, since we are looking at symmetric equilibria we need to <strong>on</strong>ly focus <strong>on</strong><br />
behaviour at s = 0, 1, 2. The different possibilities at s = 1 <strong>and</strong> s = 2 determine the<br />
different c<strong>and</strong>idates for equilibrium outcomes. There are four possible c<strong>and</strong>idates for<br />
equilibrium network advantage evoluti<strong>on</strong>, given that no innovati<strong>on</strong> occurs in the<br />
meantime.<br />
(i) x Y (1) = 1, x Y (2) = 1,<br />
(ii) x Z (1) = 1, x Y (2) = 1,<br />
(iii) x Z (1) = 1, x Z (2) = 1,<br />
(iv) x Y (1) = 1, x Z (2) = 1<br />
where x i (s) = 1 means that at state s firm i sells to <strong>on</strong>e c<strong>on</strong>sumer (<strong>and</strong> thus j ≠ i sells to no<br />
<strong>on</strong>e since there is <strong>on</strong>ly <strong>on</strong>e c<strong>on</strong>sumer in the market at any time period) in the absence of<br />
innovati<strong>on</strong> having taken place.<br />
Define p i * (s) to be the lowest price firm i is willing to charge <strong>and</strong> p(s) to be the<br />
equilibrium price at s. The lowest a firm is willing to charge at any state is just the<br />
difference between the profits if it lost the current c<strong>on</strong>sumer <strong>and</strong> if it w<strong>on</strong> it. In other<br />
words, the maximum a firm is willing to give up in order to attract a c<strong>on</strong>sumer is exactly<br />
the amount of extra profit the firm can expect to make by selling to that c<strong>on</strong>sumer. Define<br />
π i W (s), π i L (s) to be the expected profits of firm i next period <strong>on</strong>wards from respectively<br />
winning <strong>and</strong> losing the current c<strong>on</strong>sumer at state s, let π i (s) be the ex ante expected<br />
payoff of i at state s.<br />
We prove the propositi<strong>on</strong> using the following Lemmas:<br />
Lemma 1<br />
Proof:<br />
Case (iii) cannot occur in equilibrium.<br />
Suppose (iii) described the equilibrium network evoluti<strong>on</strong>. Then,<br />
p Y * (1) = p(1) = δ(1 – 2α)p(0) (1)<br />
77
HANS W. GOTTINGER<br />
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
Stochastic racing <strong>and</strong> competiti<strong>on</strong> in network markets 251<br />
This is just the difference in payoffs to firm Y from losing versus winning the current<br />
c<strong>on</strong>sumer.<br />
Similarly,<br />
π Y W (1) = 0 + δαM + δ (1 – 2α) π Y (1)<br />
π Y L (1) = (1 – 2α)p(0) + δαM + δ (1 – 2α) π Y (1)<br />
p Y * (1) = π Y L (1) – π Y W (1).<br />
p Z * (0) = p(0) = δ(1 – 2α)p(1) (2)<br />
Equati<strong>on</strong>s 1 <strong>and</strong> 2 together imply that<br />
p(1) = δ 2 (1 – 2α) 2 p(1).<br />
But that, in turn, implies that either p(1) = 0 or δ = 1 <strong>and</strong> α = 0. Since we are looking at<br />
the case where α ≠ 0 , p(1) must be zero, which means from (2), p(0) = 0. Thus<br />
π<br />
Y<br />
αM<br />
(0) = π<br />
Z<br />
(0)<br />
≡ π(0)<br />
1 − δ(<br />
1 − 2α)<br />
.<br />
We can calculate p Z * (2) <strong>and</strong> p Y * (2) (= p(2) since Y is the losing firm at s = 2)<br />
p Z * (2) = – δαM – δ 2 (1 – 2α)π(0)<br />
p Y * (2) = δαM + δ 2 (1 – 2α)π(0) – δM.<br />
But for δ < 1, p Y * (2) < p Z * (2) which means firm Y must be the <strong>on</strong>e selling to the<br />
c<strong>on</strong>sumer at s = 2 , a c<strong>on</strong>tradicti<strong>on</strong>.<br />
Lemma 2 Case (iv) describes the network evoluti<strong>on</strong> in equilibrium <strong>on</strong>ly when<br />
δ 2 (1 – 2α) 2 > 1/2 .<br />
Proof: π Z W (1) = 0 + δαM + δ (1 – 2α) π Z (1)<br />
π Z L (1) = (1 – 2α)p(2) + δαM + δ (1 – 2α) π Z (1)<br />
p Z * (1) = p(1) = π Z L (1) – π Z W (1)<br />
⇒ p(1) = δ(1 – 2α)p(2) (3)<br />
We can calculate p(2) = p Y * (2) using the same method:<br />
δαM δ(1−2 α)<br />
p(2) =<br />
1−δ( 1−2α )<br />
p(1)<br />
−δM<br />
−<br />
1 −δ 2<br />
( 1 − 2 α ) 2<br />
)<br />
*<br />
z<br />
2<br />
2<br />
δαM δ (1−<br />
2α)<br />
(2) =<br />
−<br />
2<br />
1−<br />
δ(<br />
1−<br />
2α)<br />
1−<br />
δ ( 1−<br />
2α)<br />
p<br />
2<br />
p(2<br />
(4)<br />
Substituting the expressi<strong>on</strong> for p(2) in (3), we get,<br />
2<br />
2<br />
2<br />
2<br />
δ (1−<br />
2α)<br />
αM<br />
δ (1−<br />
2α)<br />
2<br />
p(1) = +<br />
p(1) − δ (1−<br />
2α)M<br />
(5)<br />
2<br />
2<br />
1−<br />
δ(1−<br />
2α)<br />
1−<br />
δ (1−<br />
2α)<br />
Similarly, we calculate p(0) using the relati<strong>on</strong> of (3) to get p(0) = 0 since<br />
78
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
HANS W. GOTTINGER<br />
252 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
2 2<br />
δ (1−2 α) δ(1−2 α)<br />
p(0) = p(2) −<br />
1 − (1−2 ) 1 − (1−2 )<br />
2<br />
2 2 2<br />
δ α δ α<br />
p(1) .<br />
This in turn implies that p Y * (1) = 0.<br />
For case (iv) to occur in equilibrium, p(1) must be n<strong>on</strong>-negative. That leads to p Y * (1)<br />
≤ p(1) ; it also implies that p(2) is n<strong>on</strong>-negative, therefore p Z * (2) ≤ 0 ≤ p(2) from Equati<strong>on</strong><br />
(4). From (5) we see that p(1) is n<strong>on</strong>-negative <strong>on</strong>ly if δ 2 (1 – 2α) 2 ≥ 1/2.<br />
Lemma 3 As δ → 1 <strong>and</strong> α → 0, the pattern of network evoluti<strong>on</strong> described in (iv)<br />
can <strong>on</strong>ly occur in a collusive equilibrium.<br />
Proof: As δ → 1 , α → 0 , δ 2 (1 – 2α) 2 ≥ 1/2. Thus from Lemma 2 case (iv) can occur<br />
in an equilibrium. But when δ 2 (1 – 2α) 2 ≥ 1/2 , we see from (5) that p(1) > 0, that is, the<br />
equilibrium is collusive.<br />
Lemma 4 Case (i) describes the pre-innovati<strong>on</strong> network evoluti<strong>on</strong> in a n<strong>on</strong>-collusive<br />
equilibrium <strong>on</strong>ly when 3α [1 + δ (1 – 2α) ] ≤ 1.<br />
Proof:<br />
We calculate the lowest prices firms are willing to charge as described earlier <strong>on</strong><br />
p Y * (2) = δαM + δ 2 (1 – 2α) αM + δ 3 (1 – 2α) 2 M – δM + δ (1 – 2α) p(1) +<br />
δ 2 (1 – 2α) 2 p(2)<br />
p Z * (2) = p(2) = − δ α M − δ 2 (1 – 2α) α M<br />
p Y * (1) = δ 2 (1 – 2α) αM + δ 3 (1 – 2α) 2 αM – δ 2 (1 – 2α) M – δ (1 – 2() p(2)<br />
p Z *(1) = p(1) = δ 2 (1 – 2α) αM – δ 3 (1 – 2α) 2 αM<br />
p(0) = – δ (1 – 2α) p(1) – δ 2 (1 – 2α) 2 p(2) – δ 3 (1 – 2α) 2 M<br />
= δ 3 (1 – 2α) 2 M [2α + 2δα (1 – 2α) – 1].<br />
We can use the above equati<strong>on</strong>s to solve for p(0), p (1), <strong>and</strong> p(2), simultaneously. It can<br />
be checked that p Y *(2) ≤ p(2) always holds, but that p Y *(1) ≤ p(1) holds <strong>on</strong>ly if<br />
3α [1 + δ (1 – 2α) ] ≤ 1 (6)<br />
It can be checked that when inequality (6) is satisfied, then p(0), p(1), <strong>and</strong> p(2) are all<br />
negative. Note that (6) holds when δ → 1 <strong>and</strong> α → 0. Thus, when inequality (6) is<br />
satisfied the network evoluti<strong>on</strong> described in case (i) occurs as a result of the n<strong>on</strong>-collusive<br />
equilibrium described by the prices given above. Both firms charge p(s), s ∈ {−2, −1, 0,<br />
1, 2}, where p(0), p(1) <strong>and</strong> p(2) are as given above <strong>and</strong> p(−2) = p(2), p(−1) = p(1). At s =<br />
1, 2 firm Y sells to the c<strong>on</strong>sumer in the market, at s = − 1, − 2 firm Z sells to the<br />
c<strong>on</strong>sumer, <strong>and</strong> at s = 0 either firm could sell.<br />
We see that the equilibrium when there is no uncertainty is just the limit of the above<br />
equilibrium as α → 0.<br />
Lemma 5 Case (ii) describes the pre-innovati<strong>on</strong> network evoluti<strong>on</strong> in a n<strong>on</strong>collusive<br />
equilibrium <strong>on</strong>ly when 1 − 3δ 2 (1 – 2α) 2 ≤ 0 or when<br />
δ 4 (1 – 2α) 4 + 4 δ 2 (1 – 2α) 2 – 1 ≤ 0 <strong>and</strong> α − 3δ 2 (1 – 2α) 2 − (1 – 2α) (1 – δ) ≥ 0.<br />
Proof: Suppose (ii) describes the pattern of equilibrium network evoluti<strong>on</strong> in the<br />
absence of innovati<strong>on</strong>. Then, p Z * (2) = p(2) <strong>and</strong> p Y * (1) = p(1):<br />
79
HANS W. GOTTINGER<br />
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
Stochastic racing <strong>and</strong> competiti<strong>on</strong> in network markets 253<br />
M (1 2 )<br />
* (2) = δα δ − α<br />
−<br />
p(1)<br />
2<br />
1 − δ(1<br />
− 2α)<br />
1 − δ (1 − 2α)<br />
(7)<br />
p<br />
Z<br />
2<br />
2<br />
2<br />
M (1 2 )<br />
*(2) = δα δ − α<br />
+<br />
p(0) − δM<br />
2<br />
1−<br />
δ(1−<br />
2α)<br />
1−<br />
δ (1−<br />
2α)<br />
(8)<br />
pY<br />
2<br />
p<br />
Y<br />
2<br />
M (1 2 )<br />
*(1) = δα δ − α<br />
2<br />
+<br />
p(0) − δαM<br />
− δ(1−<br />
2α)p(2)<br />
δ (1−<br />
2α)M<br />
2<br />
2<br />
1−<br />
δ(1−<br />
2α)<br />
1−<br />
δ (1−<br />
2α)<br />
(9)<br />
2<br />
2<br />
M (1 2 )<br />
* (1) = δα δ − α<br />
−<br />
p(1) δαM<br />
2<br />
1 − δ(1<br />
− 2α)<br />
1 − δ (1 − 2α)<br />
(10)<br />
p<br />
Z<br />
+<br />
2<br />
p(0) = δ (1 – 2α) p(1) (11)<br />
Using Equati<strong>on</strong>s (7) <strong>and</strong> (11) to substitute in the values of p(2)<strong>and</strong> p(0) in Equati<strong>on</strong> (9),<br />
we get<br />
M<br />
p (1) =−<br />
(12)<br />
⎡<br />
2 2⎤<br />
2 2<br />
⎢ ⎛ ⎞<br />
1 12<br />
⎥ ⎛ ⎞ ⎛ ⎞<br />
⎢ −δ 12 1<br />
⎜<br />
− α⎟ ⎥δ ⎜<br />
− α⎟ ⎜<br />
−δ⎟<br />
⎢ ⎝ ⎠ ⎥ ⎝ ⎠ ⎝ ⎠<br />
⎣<br />
⎦<br />
⎡<br />
⎤<br />
⎡<br />
2⎤<br />
⎛ ⎞<br />
1 12<br />
⎢13 2⎛ ⎞<br />
⎢ −δ−α ⎥ 12<br />
⎥<br />
⎢ ⎜ ⎟ ⎥ ⎢ −δ ⎜<br />
−α ⎟<br />
⎣ ⎝ ⎠⎦⎢ ⎝ ⎠<br />
⎣<br />
⎥<br />
⎦<br />
For (ii) to occur in equilibrium, it must be true that p Z * (1) ≤ p(1) <strong>and</strong> p Y * (2) ≤ p(2). It can<br />
be checked that p Y * (2) ≤ p(2) iff 1 − 3δ 2 (1 – 2α) 2 ≤ 0, or δ 4 (1 – 2α) 4 + 4 δ 2 (1– 2α) 2 –<br />
1 ≤ 0. It can also be checked that p Z * (1) ≤ p(1) iff 1 − 3δ 2 (1 – 2α) 2 ≤ 0 or α − 3δ 2 α<br />
(1 – 2α) 2 − (1 – 2α) ( 1 – δ) ≥ 0, thus proving the lemma.<br />
Lemma 6<br />
For δ → 1 <strong>and</strong> α → 0 the equilibrium described in Lemma 5 is collusive.<br />
Proof: For δ → 1 <strong>and</strong> α → 0, 1 − 3δ 2 (1 – 2α) 2 < 0, thus case (ii) can occur in equilibrium<br />
from Lemma 5. But it can be checked from (12) that for 1 − 3δ 2 (1 – 2α) 2 ≤ 0 <strong>and</strong> δ < 1,<br />
p(1) is positive.<br />
Thus from Lemmas 1 to 6 we can infer that as δ → 1 <strong>and</strong> α → 0, the (almost) unique<br />
n<strong>on</strong>-collusive equilibrium shows the following properties. In the absence of innovati<strong>on</strong><br />
both firms charge the prices as below: As δ → 1 <strong>and</strong> α → 0, the equilibrium prices in the<br />
absence of innovati<strong>on</strong> are:<br />
p(2) = p(−2) = − δαM − δ 2 (1 − 2α) αM<br />
p(1) = p(−1) = − δ 2 (1 – 2α) αM − δ 3 (1 – 2α) 2 αM<br />
p(0) = δ 3 (1 – 2α) 2 M [2α + 2δα (1 – 2α) – 1].<br />
As α → 0, the equilibrium prices in the absence of innovati<strong>on</strong> are:<br />
p(2) = p(−2) → 0<br />
p(1) = p(−1) → 0<br />
p(0) → −δ 3 M.<br />
Firm Y sells to the c<strong>on</strong>sumer in the market at s = 1, 2 <strong>and</strong> firm Z sells when s = − 1, − 2.<br />
When s = 0, either firm could sell but both earn the same expected profits when s = 0,<br />
that is, x Y (1) = x Y (2) = x Z (−1) = x Z (−2) = 1 <strong>and</strong> x Y (0) = 1 or x Z (0) = 1.<br />
80
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
HANS W. GOTTINGER<br />
254 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Thus, in the limit as the probability of innovati<strong>on</strong> approaches zero, the equilibrium<br />
approaches the n<strong>on</strong>-collusive equilibrium without uncertainty.<br />
Propositi<strong>on</strong> 3 There exist δ * <strong>and</strong> α * such that for δ * < δ < 1 <strong>and</strong> α * < α < 1/2, the<br />
n<strong>on</strong>-collusive symmetric equilibrium prices are unique. The firm with network advantage<br />
of 2 <strong>and</strong> the firm with network advantage of 1 sell.<br />
Proof: For Lemmas 1, 2, 4 <strong>and</strong> 5, we infer that the <strong>on</strong>ly possible outcome in<br />
equilibrium has the firm with the advantage of 2 <strong>and</strong> the firm with the disadvantage of 1<br />
sell prior to an innovati<strong>on</strong> occurring. From Lemma 5 we see that for δ close to 1 <strong>and</strong> α<br />
close to 1/2, the equilibrium prices are negative. The equilibrium is, therefore, n<strong>on</strong>collusive.<br />
As δ → 1 <strong>and</strong> α → 1/2, the equilibrium prices are as follows:<br />
p(2) = p(−2) → − M/2<br />
p(1) = p(−1) → 0<br />
p(0) → 0.<br />
Propositi<strong>on</strong>s 2 <strong>and</strong> 3 illustrate the effect of uncertainty <strong>on</strong> equilibrium behaviour in a<br />
market with network externalities. In Propositi<strong>on</strong> 1 we saw that when the <strong>on</strong>ly difference<br />
between two firms is in the sizes of their networks of c<strong>on</strong>sumers, the firm with the bigger<br />
network sells to all c<strong>on</strong>sumers that subsequently enter the market.<br />
However, <strong>on</strong>ce we introduce uncertainty about the quality of a firm’s product this no<br />
l<strong>on</strong>ger is true. Even if the uncertainty faced by the firms is fully symmetric <strong>and</strong> the <strong>on</strong>ly<br />
difference between the firms is in the size of their respective networks, the firm with the<br />
bigger network does not always sell in equilibrium. For a sufficiently high success rate in<br />
innovati<strong>on</strong>, it is the firm with the smaller network that sells provided the network<br />
distance is not too large.<br />
The reas<strong>on</strong> for this strikingly different outcome compared to the situati<strong>on</strong> without<br />
uncertainty is as follows: When there is no uncertainty, the <strong>on</strong>ly way either firm can<br />
realise positive profits is by reaching its absorbing state. Outside of their respective<br />
absorbing states both firms make the same profits that is zero. Thus the way for either<br />
firm to make positive profits is to sell to all the c<strong>on</strong>sumers <strong>and</strong> reach its absorbing state.<br />
However, since the firm with the larger network needs a smaller additi<strong>on</strong> to its network to<br />
reach its absorbing state, it needs a shorter time span to reach its absorbing state, hence its<br />
discounted profits from selling to all c<strong>on</strong>sumers is larger. Not selling to any c<strong>on</strong>sumers,<br />
<strong>on</strong> the other h<strong>and</strong>, gives both firms zero profits. Thus, since the firm with the bigger<br />
network has more to gain from selling to all subsequent c<strong>on</strong>sumers while not selling<br />
gives it the same profits as the firm that is behind, it is willing to price more aggressively<br />
in order to lure the c<strong>on</strong>sumers. Thus, in equilibrium the firm that is ahead always sells.<br />
The introducti<strong>on</strong> of uncertainty, even when it is uniform across the firms, alters the<br />
incentives of the firms to fight for the c<strong>on</strong>sumers. When the difference in network (size)<br />
between the two firms is sufficiently small <strong>and</strong> the probability of success is high the<br />
presence of uncertainty reduces the incentive of the firm that is ahead to fight for the<br />
current c<strong>on</strong>sumer relative to that of the firm with the network disadvantage. That is<br />
because with a high rate of innovati<strong>on</strong> the leading firm has a high probability of getting<br />
its m<strong>on</strong>opoly profits whether it sells to the current c<strong>on</strong>sumer or not. The firm with the<br />
network disadvantage, <strong>on</strong> the other h<strong>and</strong>, has a bigger stake in winning the current<br />
c<strong>on</strong>sumer since by not letting the other firm get close to its absorbing state, it keeps its<br />
own chances of getting the innovati<strong>on</strong> <strong>and</strong> thus reaping the high m<strong>on</strong>opoly profits alive.<br />
81
HANS W. GOTTINGER<br />
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
Stochastic racing <strong>and</strong> competiti<strong>on</strong> in network markets 255<br />
However, when the leading firm has a big network advantage, that is, is close to its<br />
absorbing state, its incentive to sell is much higher since it can then guarantee itself the<br />
m<strong>on</strong>opoly profits by selling to the current customer. Thus as δ → 1 <strong>and</strong> α → 1/2, we see<br />
the firm with a network advantage of 2 selling <strong>and</strong> also the firm with a network<br />
disadvantage of 1.<br />
Until now all the discussi<strong>on</strong> has been in terms of network advantage or disadvantage.<br />
We could equally well have had the discussi<strong>on</strong> in terms of quality advantage or<br />
disadvantage since network <strong>and</strong> quality are perfect substitutes in a c<strong>on</strong>sumer’s utility.<br />
Thus without uncertainty, in the presence of network externalities, if the <strong>on</strong>ly difference<br />
between two firms was in the quality of their products then the firm with the better<br />
quality product would sell to all the c<strong>on</strong>sumers. With uncertain quality improvement<br />
however, if the probability of quality improvement were high enough then the firm with<br />
the lower quality would sell to the c<strong>on</strong>sumer. Both firms would remain in the market until<br />
<strong>on</strong>e of the firms got an innovati<strong>on</strong> that gave it an insurmountable lead over the other.<br />
6 Extensi<strong>on</strong>s <strong>and</strong> c<strong>on</strong>clusi<strong>on</strong>s<br />
We have looked at a simple model of a market with network externalities where there is<br />
technological innovati<strong>on</strong>. It would be interesting what kind of modificati<strong>on</strong>s to our basic<br />
model might change some of the results. The model is based <strong>on</strong> the assumpti<strong>on</strong> of<br />
symmetry in terms of the innovati<strong>on</strong> that the firms expect.<br />
If we introduce asymmetric innovati<strong>on</strong> this would permit us to endogenise innovati<strong>on</strong><br />
to the extent that we could look at the decisi<strong>on</strong>s of firms to improve quality. If we look at<br />
a more dynamic model of having sequential innovati<strong>on</strong> <strong>and</strong> multiple races<br />
(<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g> [2002b]) we might be able to see whether there is a clustering of innovati<strong>on</strong>s,<br />
<strong>and</strong> by examing the size of innovati<strong>on</strong>s, whether firms would rather tend toward <strong>on</strong>e big<br />
risky innovati<strong>on</strong> or rather for a series of small <strong>on</strong>es which are less risky. Under welfare<br />
theoretic c<strong>on</strong>siderati<strong>on</strong>s there may be related questi<strong>on</strong>s <strong>on</strong> whether there is too much or<br />
too little innovati<strong>on</strong> in view of a social optimum (<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, 2003).<br />
Our simple model shows how competing firms behave in a market with network<br />
externalities before the resoluti<strong>on</strong> of technological uncertainty. If our firms are in waiting<br />
(for the breakthrough innovati<strong>on</strong> to come true) but the probability of innovati<strong>on</strong> is small,<br />
then the firm that has a network advantage makes all the sales. However, if the<br />
probability of innovati<strong>on</strong> is large, then it is not always the firm with the bigger network<br />
that sells. While a firm with a network advantage of 2 would make the sale, it is the firm<br />
with a network disadvantage of 1 that makes the sale if it is lucky to catch this<br />
innovati<strong>on</strong>.<br />
Alternatively, it could also serve as an explanati<strong>on</strong> for why a firm, that is in the end,<br />
driven out of the market may just hang <strong>on</strong> there for a while, with a network disadvantage,<br />
before it finally gives up <strong>and</strong> exits (or is exited from) the market.<br />
Furthermore, as some sort of by-product of this analysis, we could pursue the<br />
interacti<strong>on</strong> between network effects <strong>and</strong> innovati<strong>on</strong> for market structures. If network<br />
effects are ‘competitive barriers’ that could <strong>on</strong>ly be overcome by radical innovati<strong>on</strong> then,<br />
by implicati<strong>on</strong>, the synergy of network effects <strong>and</strong> radical innovati<strong>on</strong> through the<br />
incumbent would c<strong>on</strong>stitute a formidable powerhouse to c<strong>on</strong>solidate m<strong>on</strong>opolistic<br />
positi<strong>on</strong>s in dynamic product markets where ‘winner-takes-all markets’ are created.<br />
82
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
HANS W. GOTTINGER<br />
256 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Acknowledgments<br />
I want to say thank to A. Aykac, Sophia Antipolis, J. McGee, Warwick, I. Png,<br />
Singapore, for their helpful comments <strong>and</strong> suggesti<strong>on</strong>s.<br />
References<br />
Belsen, K. (2003) ‘S<strong>on</strong>y again turns to designs to lift electr<strong>on</strong>ics’, New York Times<br />
(Online Editi<strong>on</strong>), Feb. 02.<br />
Choi, J.P. (1994) ‘Irreversible choice of uncertain technologies with network externalities’, R<strong>and</strong><br />
Journal of Ec<strong>on</strong>omics, Vol. 25, pp.382–401.<br />
Clark, D. (2003) ‘Sun unveils blade servers in bid to counter IBM, H-P’, Wall Street Journal<br />
(Online Editi<strong>on</strong>), Feb.22.<br />
Evans, D.S. <strong>and</strong> Schmalensee, R. (2001) ‘Some ec<strong>on</strong>omic aspects of antitrust analysis in<br />
dynamically competitive industries’, Nati<strong>on</strong>al Bureau of Ec<strong>on</strong>omic Research, Cambridge,<br />
MA. Working Paper 8268, http://www.nber.org/papers/w8268.<br />
Fudenberg, D. <strong>and</strong> Tirole, J. (1993) Game Theory, MIT Press, Cambridge, MA.<br />
Futia, C. (1980) ‘Schumpeterian competiti<strong>on</strong>’, Quarterly Journal of Ec<strong>on</strong>omics, Vol. 94,<br />
pp.675–695.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H.W (2002b) Sequential Innovati<strong>on</strong>s <strong>and</strong> R&D Racing, http://www.csef.it/ publicati<strong>on</strong>s<br />
2001/.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H.W (2003) Ec<strong>on</strong>omies of Network Industries, Routledge, L<strong>on</strong>d<strong>on</strong>.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H.W. (2002a) ‘Modeling stochastic innovati<strong>on</strong> races’, Technological Forecasting <strong>and</strong><br />
Social Change, Vol. 69, pp.607–624.<br />
Hamel, G. <strong>and</strong> Switzer, L. (2004) ‘The old guard vs. the vanguard’, The Wall Street Journal<br />
(Online Editi<strong>on</strong>), Feb. 23.<br />
Harris, C. <strong>and</strong> Vickers, J. (1987) ‘Racing with uncertainty’, Review of Ec<strong>on</strong>omic Studies, Vol. 54,<br />
pp.1–21.<br />
Katz, M. <strong>and</strong> Shapiro, C. (1992) ‘Product introducti<strong>on</strong> with network externalities’, The Journal of<br />
Industrial Ec<strong>on</strong>omics, Vol. 40, pp.55–83.<br />
Kristiansen, E.G. (1998) ‘R&D in the presence of network externalities: timing <strong>and</strong> compatibility’,<br />
R<strong>and</strong> Journal of Ec<strong>on</strong>omics, Vol. 29, pp.531–547.<br />
Lohr, S. (2003) ‘Competitors shape strategy to gain edge in web services’, New York Times (Online<br />
Editi<strong>on</strong>), Feb. 3.<br />
Markoff, J. <strong>and</strong> Lohr, S. (2003) ‘Advanced micro issues challenge to rivals’, New York Times<br />
(Online Editi<strong>on</strong>), April 19.<br />
Uttenback, J.M. (1994) ‘Mastering the dynamics of innovati<strong>on</strong>’, Harvard Business School Press,<br />
Bost<strong>on</strong>.<br />
83
HANS W. GOTTINGER<br />
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
84
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
HANS W. GOTTINGER<br />
Int. J. Business <strong>and</strong> Systems Research, Vol. X, No. Y, XXXX 1<br />
High speed technology competiti<strong>on</strong><br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g>-Werner <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
IMS, University of Maastricht,<br />
Maastricht, The Netherl<strong>and</strong>s<br />
IIR, Hitotsubashi University,<br />
Tokyo, Japan<br />
E-mail: hg528@bingo-ev.de<br />
Abstract: We c<strong>on</strong>sider a technological race in which competitive firms engage<br />
in a class of differential games <strong>and</strong> some firms have priority of moves over<br />
others. The firm that has the right to move first is called the technological<br />
leader <strong>and</strong> the other competing firm is called the follower. An example of<br />
this type of sequential move game is the Stackelberg model of duopoly.<br />
The open-loop Nash equilibrium c<strong>on</strong>diti<strong>on</strong>s in a sequential move game is<br />
derived <strong>and</strong> the results for technological race problems are derived as is a<br />
comparis<strong>on</strong> of the strategies of leader <strong>and</strong> the follower.<br />
Keywords: industrial organisati<strong>on</strong>; dynamic competiti<strong>on</strong>; differential games;<br />
technological leadership.<br />
Reference to this paper should be made as follows: <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H-W. (XXXX)<br />
‘High speed technology competiti<strong>on</strong>’, Int. J. Business <strong>and</strong> Systems Research,<br />
Vol. X, No. Y, pp.XXX–XXX.<br />
Biographical notes: <str<strong>on</strong>g>Hans</str<strong>on</strong>g>-Werner <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g> has been the Director of the<br />
Institute of Management Science, Maastricht, NL <strong>and</strong> the Professor of<br />
Ec<strong>on</strong>omics in the Institute of Innovati<strong>on</strong> Research., Tokyo, Japan. He has<br />
internati<strong>on</strong>ally taught <strong>and</strong> widely published in the areas of industrial, energy<br />
<strong>and</strong> envir<strong>on</strong>mental ec<strong>on</strong>omics <strong>and</strong> management. In these areas he has also been<br />
an advisor for internati<strong>on</strong>al organisati<strong>on</strong>s. He is the author for the recently<br />
published book titled Ec<strong>on</strong>omies of Network Industries (Routledge, 2003) <strong>and</strong><br />
Innovati<strong>on</strong>, <strong>Technology</strong> <strong>and</strong> Hypercompetiti<strong>on</strong> (Routledge, 2006).<br />
1 Introducti<strong>on</strong><br />
In highly competitive technological industries new challenges <strong>and</strong> opportunities are<br />
arising in the new product development arena. Driven by global markets, global<br />
competiti<strong>on</strong>, the global dispersi<strong>on</strong> of scientific/engineering talent <strong>and</strong> the advent of new<br />
Informati<strong>on</strong> <strong>and</strong> Communicati<strong>on</strong> Technologies (ICT) a new visi<strong>on</strong> of product<br />
development is that of a highly disaggregated, distributive process with people <strong>and</strong><br />
organisati<strong>on</strong>s spread throughout the world. At the same time, products are becoming<br />
increasingly complex requiring numerous engineering decisi<strong>on</strong>s to bring them to market.<br />
Competitive pressures mean that ‘time to market’ has become a key to new product<br />
success. However, at the same time, it is important to keep the innovati<strong>on</strong> <strong>and</strong> quality<br />
dimensi<strong>on</strong>s of the new product at their optimal level.<br />
Copyright © XXXX Inderscience Enterprises Ltd.<br />
85
HANS W. GOTTINGER<br />
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
2 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
A central questi<strong>on</strong> to address is, how should firms invest in innovati<strong>on</strong> <strong>and</strong> what are<br />
the implicati<strong>on</strong>s of such investments for competitive advantage? Underst<strong>and</strong>ing why a<br />
firm benefits from investments in innovati<strong>on</strong> <strong>and</strong> quality illuminates issues of<br />
competitive strategy <strong>and</strong> industrial organisati<strong>on</strong>. In the field of competitive strategy,<br />
much attenti<strong>on</strong> has been devoted to the c<strong>on</strong>cept of core capabilities (Teece et al., 1997).<br />
Underst<strong>and</strong>ing how firms make optimal investments in the face of competiti<strong>on</strong> reveals<br />
the nature of competiti<strong>on</strong> <strong>and</strong> provides theoretical <strong>and</strong> managerial implicati<strong>on</strong>s for<br />
developing core competence <strong>and</strong> dynamic capabilities.<br />
In major parts of competitive analysis involving R&D decisi<strong>on</strong>s the focus is <strong>on</strong><br />
breakthrough innovati<strong>on</strong>s which could create entirely new markets, for example, in<br />
studies featuring patent racing between competing firms. In more comm<strong>on</strong> competitive<br />
situati<strong>on</strong>s we observe firms, however, competing by investing in incremental<br />
improvements of products. It is an important aspect when innovati<strong>on</strong> is c<strong>on</strong>sidered to be<br />
manifested in product quality, process improvements <strong>and</strong> in the overall quality culture of<br />
an organisati<strong>on</strong> (Kanter, 2006). For example, after product launch, incremental<br />
improvement of different aspects of product quality, improvements in various business<br />
processes <strong>and</strong> an incremental adopti<strong>on</strong> of a quality culture are quite real-world<br />
phenomena. Some firms operate in a simultaneous product launch situati<strong>on</strong> while others<br />
compete sequentially by adopting the role of leader or follower. The strategic<br />
implicati<strong>on</strong>s in these diverse circumstances can be treated within a unified framework of<br />
dynamic stochastic differential games.<br />
In recent years, with the emergence of e-business <strong>and</strong> a supply chain view for<br />
product development processes, multiple firms with varying <strong>and</strong> at times c<strong>on</strong>flicting<br />
objectives enter into collaborative arrangements. In such situati<strong>on</strong>s, the competitive<br />
strategy based <strong>on</strong> quality <strong>and</strong> innovati<strong>on</strong> could potentially permeate in those<br />
collaborative setups. A recent example is the collaborative venture of S<strong>on</strong>y <strong>and</strong> Samsung<br />
to build a cutting edge plant for LCD flat screen TV though displaying <strong>on</strong>going fierce<br />
rivalry in new product launches in exactly the same product categories. When innovati<strong>on</strong><br />
<strong>and</strong> quality levels form the core of a firm’s capabilities, each member in the supply chain<br />
would have an incentive to invest <strong>and</strong> improve their dynamic capabilities. This leads to<br />
tacit competiti<strong>on</strong> am<strong>on</strong>g collaborative product development partners by means of active<br />
investment in innovati<strong>on</strong> <strong>and</strong> quality.<br />
Although the forces of innovati<strong>on</strong> are central to competiti<strong>on</strong> to emerging technically<br />
dynamic industries, they also affect mature industries where life cycles, historically, were<br />
relatively str<strong>on</strong>g, technologies were mature <strong>and</strong> dem<strong>and</strong>s stable.<br />
A strategy for technology must c<strong>on</strong>fr<strong>on</strong>t primarily what the focus of technical<br />
development will be. The questi<strong>on</strong> is what technologies are critical to the firm’s<br />
competitive advantage. In this c<strong>on</strong>text, technology must include the ‘know-how’ the firm<br />
needs to create, produce, market its products <strong>and</strong> deliver them to customers. As a major<br />
step in creating a technology strategy it has to define those capabilities where the firm<br />
seeks to achieve a distinctive advantage relative to competitors. For most firms, there are<br />
a large number of important areas of technological ‘know-how’ but <strong>on</strong>ly a h<strong>and</strong>ful where<br />
the firm will seek to create truly superior capability.<br />
Having determined the focus of technical development <strong>and</strong> the source of capability,<br />
the firm must establish the timing <strong>and</strong> frequency for innovati<strong>on</strong> efforts. Part of the timing<br />
issue involves developing technical capabilities, <strong>and</strong> the rest involves introducing<br />
technology into the market. The frequency of implementati<strong>on</strong> <strong>and</strong> associated risks will<br />
depend in part <strong>on</strong> the nature of the technology <strong>and</strong> the markets involved (e.g. disk drive<br />
86
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
HANS W. GOTTINGER<br />
High speed technology competiti<strong>on</strong> 3<br />
versus automotive technology), but in part <strong>on</strong> strategic choice. At the extreme, a firm<br />
may adopt a rapid incremental strategy, that is, frequent, small changes in technology<br />
that cumulatively lead to c<strong>on</strong>tinuous performance improvement. The polar opposite<br />
might be termed the great leap forward strategy. In this approach, a firm chooses to make<br />
infrequent but large-scale changes in technology that substantially advance the state of<br />
the art (<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, 2006, Chap. 6).<br />
As an example of the importance of innovati<strong>on</strong> strategy in product development, we<br />
notice that IBM created <strong>and</strong> c<strong>on</strong>tinues to dominate the mainframe segment, but it missed<br />
by many years, the emergence of the minicomputer architecture <strong>and</strong> market.<br />
The minicomputer was developed <strong>and</strong> its market applicati<strong>on</strong>s exploited by firms such as<br />
Digital Equipment Corporati<strong>on</strong> (DEC) <strong>and</strong> Data General (DG) back in the 1970s.<br />
Summarising, in a mainstream scenario, high speed technological competiti<strong>on</strong> is<br />
likely to be driven by<br />
1 companies competing in fast changing, diverse networks <strong>and</strong><br />
2 companies c<strong>on</strong>fr<strong>on</strong>ted with ever shortening product/technology life cycles<br />
because of multiple interacti<strong>on</strong>s between current <strong>and</strong> emerging technologies <strong>and</strong><br />
product diversity or what may be circumscribed as ‘combinatorial innovati<strong>on</strong>’<br />
(Varian et al., 2004).<br />
A clear example provided by Arthur (2000, p.3):<br />
“If you look at genomics it is certainly heavily based <strong>on</strong> biology, but it is also<br />
highly dependent <strong>on</strong> digital computing!<br />
Genomics, then, is as much a manifestati<strong>on</strong> of digital computing as it is a<br />
manifestati<strong>on</strong> of molecular biology. What happens is that some of the new<br />
technologies become base technologies, <strong>and</strong> these give rise to manifestati<strong>on</strong>s,<br />
some of which themselves become base technologies. Manfestati<strong>on</strong>s then<br />
appear at an even faster rate.”<br />
To represent the features of leader-follower type in a sequence of technological racing<br />
we c<strong>on</strong>sider a class of differential games in which some firms have priority of moves<br />
over others.<br />
The firm that has the right to move first is called the leader <strong>and</strong> the other competing<br />
firm is called the follower. A well-known example of this type of sequential move game<br />
is the Stackelberg model of duopoly. In this type of interacti<strong>on</strong> the open-loop Nash<br />
equilibrium c<strong>on</strong>diti<strong>on</strong>s in a sequential move game can be derived. It would lead to a<br />
comparis<strong>on</strong> of the strategies of leader <strong>and</strong> the follower.<br />
Secti<strong>on</strong> 2 presents the essence of leader-follower type interacti<strong>on</strong>s through the<br />
format of a differential game which is accommodated to a competitive market<br />
situati<strong>on</strong> in Secti<strong>on</strong> 3. The properties of their interacti<strong>on</strong>s are derived in Secti<strong>on</strong> 4.<br />
Secti<strong>on</strong> 5 h<strong>and</strong>les various market asymmetries in the leader follower type situati<strong>on</strong>.<br />
Finally, Secti<strong>on</strong> 6 summarises <strong>and</strong> draws up c<strong>on</strong>clusi<strong>on</strong>s <strong>and</strong> directi<strong>on</strong>s for future<br />
research.<br />
2 A differential game formulati<strong>on</strong><br />
For our modelling effort we follow the notati<strong>on</strong>s <strong>and</strong> symbols as explained below.<br />
N: number of firms, T: finite time horiz<strong>on</strong> for the strategies, i, j: superscripts to<br />
denote competing firms in a duopoly, t: an instant of time in the dynamic game setup,<br />
87
HANS W. GOTTINGER<br />
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
4 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
u(t): investments in innovati<strong>on</strong> effort (expenditure per unit time), R(t): net revenue rate<br />
for the firm at t, R o<br />
: product category net revenue rate for the existing product,<br />
R 1<br />
: product category net revenue rate for the new product, x o<br />
: quality level of the existing<br />
product.<br />
Several examples illustrate the importance of innovati<strong>on</strong> <strong>and</strong> quality as<br />
competitive weap<strong>on</strong>s. In the computer mainframe arena IBM replaced Remingt<strong>on</strong><br />
R<strong>and</strong> <strong>and</strong> Sperry Univac as market leader in the 1950s, <strong>and</strong> its subsequent growth<br />
outperformed its competitors so rapidly that in 1963 its data process revenues were<br />
four times larger than the combined revenues of its eight main rivals in the US market<br />
(Hoffman, 1976).<br />
In the model, we assume that enhancements of quality are achieved by climbing a<br />
performance ladder that may squarely embrace technology but could exp<strong>and</strong> to other<br />
criteria uniquely identified with quality. Let the quality of the product at time t be x(t).<br />
In the c<strong>on</strong>text of total quality management, as quality levels increase it becomes even<br />
more difficult to climb the performance ladder. The hypothesis behind the formulati<strong>on</strong> is<br />
that a firm needs to make higher innovati<strong>on</strong> investments targeted towards the<br />
improvement of quality. To capture this dynamics a negative feedback effect of the<br />
present state quality <strong>on</strong> the rate of change of product quality (x " (t)) is c<strong>on</strong>sidered.<br />
The state dynamics is<br />
"<br />
#<br />
x ! t" $ K# u! t "$ % %x! t"<br />
(1)<br />
where K is proporti<strong>on</strong>al to the level of capital investment in development technology <strong>and</strong><br />
L is the proporti<strong>on</strong>ality c<strong>on</strong>stant for the influence of present quality <strong>on</strong> the speed of<br />
further quality improvements, # is the innovati<strong>on</strong> resource productivity parameter.<br />
Based <strong>on</strong> (1) the quality of the product at time t is<br />
t<br />
#<br />
x! t" $ xo<br />
*<br />
&<br />
( K# u! s "$ % %x! s" '<br />
) & s<br />
(2)<br />
%<br />
The product quality provides a means for evaluating the product’s attractiveness in the<br />
market in the presence of other competing products. The firm’s market share is a<br />
functi<strong>on</strong> of both its own product quality <strong>and</strong> the product quality of rivals (the initial<br />
product quality x o<br />
may reflect the reputati<strong>on</strong>al capital of the firm, either low or high, <strong>and</strong><br />
may induce her to different innovati<strong>on</strong> investment efforts).<br />
The difference lies in the planning horiz<strong>on</strong>. In this competitive setup, the<br />
planning horiz<strong>on</strong> extends bey<strong>on</strong>d the date of launch <strong>and</strong> the competing firms c<strong>on</strong>tinue<br />
investing in product-process innovati<strong>on</strong> until the end of the growth phase of a product.<br />
Further,<br />
+ x%<br />
R%<br />
( % , t - ,<br />
i *<br />
p<br />
. x% / x%<br />
R! t"<br />
$ 0<br />
i<br />
. x !,<br />
"<br />
R<br />
'<br />
( ,<br />
i *<br />
p<br />
, t - ,<br />
.<br />
1 x !, " / x !,<br />
"<br />
The cumulative development cost of the new product at time t is given as:<br />
T*! t"<br />
$ * ,<br />
+ u! s",&<br />
s (4)<br />
%<br />
(3)<br />
88
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
HANS W. GOTTINGER<br />
High speed technology competiti<strong>on</strong> 5<br />
the firm’s cumulative profit at time t is determined as follows,<br />
, 2 ! t" $ TR! t" % T*! t"<br />
(5)<br />
where TR(t) <strong>and</strong> TC (T) are total revenues <strong>and</strong> costs at time t, respectively. The total<br />
revenue functi<strong>on</strong> is given by:<br />
TR! t"<br />
$ * ,<br />
R! s"&<br />
s (6)<br />
%<br />
where R(3) is given in (3). The firm’s decisi<strong>on</strong> set is 4 $ {u(t)}. Notice that the firm<br />
precommits <strong>on</strong> the date of product launch T p<br />
. The cumulative profit functi<strong>on</strong>, T 2(5), is<br />
defined as the total profit by end of the window of opportunity with decisi<strong>on</strong> 564.<br />
The firm’s decisi<strong>on</strong> problem can be stated as,<br />
7 9 8 7 9 8 9 7 9<br />
8<br />
./x<br />
5 64<br />
, 2 ! 5 " $ TR 5 % T* 5 $ , 2 5<br />
(7)<br />
The combinati<strong>on</strong> of Equati<strong>on</strong>s (1)–(6) generates an explicit representati<strong>on</strong> of firm i’s<br />
cumulative profit by the end of time horiz<strong>on</strong>. This substituti<strong>on</strong> yields:<br />
& x%<br />
, 2 9 75<br />
9<br />
8 $ ./x5<br />
64 : R<br />
%<br />
,<br />
i *<br />
( x% / x%<br />
,<br />
#<br />
xo<br />
/ * K<br />
%<br />
'# u! s "$ % %x! s"&<br />
s<br />
/<br />
'<br />
%<br />
,<br />
#<br />
*<br />
xo<br />
/ * K# u! s "$ % %x! s"& s / x !,<br />
"<br />
%<br />
%<br />
*<br />
,<br />
%<br />
+ u! s",&<br />
s'<br />
;)<br />
p<br />
7 p 8<br />
R , ,<br />
(8)<br />
*<br />
where x<br />
%<br />
<strong>and</strong> x j<br />
(T), respectively are the quality of existing <strong>and</strong> new products of the<br />
competitor. C<strong>on</strong>sidering a duopoly, the optimisati<strong>on</strong> problem written above can be<br />
reformulated as a differential game problem with state variable for the firm i<br />
(the competing firm is represented by superscript j) given as x i (t); the c<strong>on</strong>trol variable<br />
u i (t). In the terminology used in optimal c<strong>on</strong>trol <strong>and</strong> differential games, the salvage term<br />
for firm i, < i (T, x(T)) is defined as follows:<br />
i<br />
i<br />
x%<br />
< !, ( x !, ""141R%<br />
,<br />
i *<br />
x / x<br />
/ R<br />
'<br />
i<br />
p<br />
% %<br />
i<br />
,<br />
i i<br />
#<br />
i<br />
= x% / * K &<br />
'<br />
! "'<br />
%<br />
'<br />
! "& >7 % 8<br />
% ( u s ) % x s s , ,<br />
p<br />
i<br />
,<br />
i i<br />
#<br />
i *<br />
& '<br />
%<br />
/ *%<br />
' ( ) %<br />
'<br />
/<br />
x K u ! s " % x ! s"& s x !,<br />
"<br />
(9)<br />
where<br />
i<br />
,<br />
* *<br />
! "<br />
& *<br />
#<br />
*<br />
x , $ x &<br />
%<br />
/<br />
%<br />
2<br />
! "'<br />
%<br />
2<br />
! "<br />
'<br />
* :<br />
K ( u s ) % x s<br />
;<br />
& s (10)<br />
( )<br />
89
HANS W. GOTTINGER<br />
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
6 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
3 A sequential differential game between two competing firms<br />
We extend the c<strong>on</strong>ceptualisati<strong>on</strong> of a hyper-competitive scenario by c<strong>on</strong>sidering a<br />
sequential differential game as being representative of a leader-follower or<br />
incumbent-entrant competitive situati<strong>on</strong>. The game is characterised by informati<strong>on</strong><br />
asymmetry where the follower is aware of the innovati<strong>on</strong> <strong>and</strong> quality levels of the<br />
leader’s products. The motivati<strong>on</strong> for c<strong>on</strong>sidering this scenario is its close<br />
corresp<strong>on</strong>dence with many real life competitive cases. Knowing the investment strategy<br />
of the leader, the rival firms can formulate their own strategies. Therefore, the firm acting<br />
as a leader chooses a decisi<strong>on</strong> path that maximises the objective for all c<strong>on</strong>ceivable<br />
resp<strong>on</strong>ses that can be taken by the follower(s). In the case of sequential games a<br />
hierarchical play differential game approach is used to model the competitive situati<strong>on</strong><br />
<strong>and</strong> to obtain the open-loop Stackelberg Nash equilibrium. The issue of subgame<br />
perfectness <strong>and</strong> commitment is extremely important in these soluti<strong>on</strong>s. Adding to the<br />
previous notati<strong>on</strong> we let T p<br />
, date of product launch by leader, T p<br />
/ ?, date of product<br />
launch by follower. The leader is represented by the superscript i <strong>and</strong> the follower is<br />
represented by the superscript j. Since it is a sequential game the time of product launch<br />
for the two players is such that the leader launches the product at time T p<br />
. Later, the<br />
follower launches the product after time ? at T p<br />
/ ?. A c<strong>on</strong>tinuous improvement in the<br />
product is c<strong>on</strong>sidered <strong>and</strong> therefore the entire time interval t 6 @0, TA for the leader <strong>and</strong><br />
t 6 @T p<br />
B TA for the follower needs to be optimised. The state dynamics of the leader is<br />
" i i<br />
#<br />
i<br />
$ & '<br />
'<br />
%<br />
'<br />
x ! t" K ( u ! t ") % x ! t"<br />
(11)<br />
The rate of quality improvement x "i (t) increases with investments u i (t). The factor L 1<br />
x i (t)<br />
suggests a natural decay in quality in the absence of any investments. The differential<br />
game formulati<strong>on</strong> for leader is:<br />
i i i<br />
7 p 8 = 7 8 7 8><br />
* ,<br />
%<br />
./x , C u ! t "(, $ % u ! s"& s / < ,( x !, "<br />
(12)<br />
where < i (T, x i (T)) is defined as<br />
s. t.<br />
i i i<br />
i<br />
R% x ,<br />
% p<br />
/ R'<br />
x !,<br />
"! " '<br />
i * i * i i<br />
%<br />
/<br />
%<br />
/<br />
%<br />
/<br />
7 p 8<br />
? R x !, " , %,<br />
%?<br />
i i<br />
< 7, ( x !,<br />
" 8 D /<br />
x x x !, " x x !, " x !,<br />
"<br />
3<br />
i<br />
#<br />
i i i i<br />
x t $ K &<br />
'<br />
u t ' % %' x t x $ x<br />
%<br />
, x ,<br />
(13)<br />
! " ( ! ") ! "( !%" ( 4ixe&( ! "47ee<br />
(14)<br />
Next, the follower’s problem formulati<strong>on</strong> is discussed. Owing to the sequential nature of<br />
the game, the informati<strong>on</strong> about the leader’s quality <strong>and</strong> about the leader’s investments is<br />
known to the follower. Specifically, it is assumed that the knowledge gained by the<br />
leader’s investments ‘spills over’ to the follower’s quality improvement dynamics. It is<br />
plausible to assume that in the c<strong>on</strong>text of a ‘leader-follower’ competiti<strong>on</strong>, the level of<br />
quality improvements of the follower indeed depends not <strong>on</strong>ly <strong>on</strong> its own innovati<strong>on</strong><br />
efforts but also <strong>on</strong> the knowledge pool available because of the leader’s investments.<br />
The fact that the leader often cannot wholly c<strong>on</strong>ceal his efforts or can he credibly<br />
announce the commitment he has made, make the situati<strong>on</strong> quite complicated. To address<br />
these issues fully in this c<strong>on</strong>text it would require a subtle <strong>and</strong> rich analysis of games with<br />
incomplete informati<strong>on</strong>s (Robs<strong>on</strong>, 1990).<br />
90
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
HANS W. GOTTINGER<br />
High speed technology competiti<strong>on</strong> 7<br />
A formal treatment of ‘spillover effects’ would enrich this model. The spillover<br />
effect is modelled by c<strong>on</strong>sidering a linear additive term M 2<br />
u i (t). The follower’s state<br />
dynamics is<br />
3 * *<br />
#<br />
* i<br />
$ & '<br />
2<br />
%<br />
2<br />
/<br />
2<br />
x ! t" K ( u ! t ") % x ! t" - u ! t"<br />
(15)<br />
It is assumed that the spillover is a functi<strong>on</strong> of the investments made by the leader u i (t).<br />
In this term M 2<br />
is a very small number which quantifies the amount of spillover from the<br />
leader to the follower. The term M 2<br />
u i (t) would restrict the analysis to the case where the<br />
follower emulates the innovati<strong>on</strong> <strong>and</strong> product quality of the leader, as opposed to setting<br />
his own technology <strong>and</strong> quality st<strong>and</strong>ards. Now with all c<strong>on</strong>siderati<strong>on</strong>s wrapped up, a<br />
leader-follower dynamics can be stated.<br />
The follower’s differential game formulati<strong>on</strong> is given as:<br />
,<br />
7 p 8 * = > 7 8<br />
, p<br />
, %,<br />
p * * *<br />
$ %*<br />
= u 7s / ,<br />
p 8> & s / < 7, ( x !,<br />
" 8<br />
* * * * *<br />
./x , 2 u ! t "(, $ % u ! s" & s / < , ( x !,<br />
"<br />
where < j (T, x (T)) is defined as,<br />
s.t.<br />
*<br />
x<br />
< !, ( x !, "" 4R %<br />
,<br />
x<br />
*<br />
%<br />
i *<br />
%<br />
/ x%<br />
p<br />
%<br />
*<br />
= x !, " 7, %,<br />
p<br />
%?<br />
8><br />
x<br />
/ R<br />
/<br />
x !, " x x !, " x !,<br />
"<br />
*<br />
% i<br />
%<br />
? R<br />
* '<br />
/<br />
%<br />
i<br />
/<br />
*<br />
* *<br />
#<br />
* i<br />
$ & '<br />
2<br />
%<br />
2<br />
/<br />
2<br />
x ! t" K ( u ! t ") % x ! t" - u ! t"<br />
(18)<br />
where M 2<br />
u i<br />
(t) represent the spillover of knowledge, assumed to be a functi<strong>on</strong> of<br />
investments by the leader, M 2<br />
is a small c<strong>on</strong>stant such that 0 - M 2<br />
-- 1.<br />
where<br />
Furthermore, x<br />
* !%" $ x<br />
*<br />
%<br />
( T fixed, x j (T) free<br />
i<br />
t<br />
i i<br />
! "<br />
& i<br />
#<br />
i<br />
x , $ x &<br />
%<br />
/<br />
%<br />
'<br />
! "'<br />
%<br />
'<br />
! "<br />
'<br />
* :<br />
K ( u s ) % x s<br />
;<br />
& s<br />
( )<br />
*<br />
t<br />
* *<br />
! "<br />
& *<br />
#<br />
* i<br />
x , $ x &<br />
%<br />
/<br />
2<br />
! "'<br />
%<br />
2<br />
! " /<br />
2<br />
! "<br />
'<br />
* :<br />
K ( u s ) % x s - u s<br />
;<br />
& s<br />
, p ( )<br />
(16)<br />
(17)<br />
4 Analysis of the model <strong>and</strong> discussi<strong>on</strong><br />
A c<strong>on</strong>venti<strong>on</strong>al tool for solving the problem is the applicati<strong>on</strong> of P<strong>on</strong>tryagin’s maximum<br />
principle for open-loop Stackelberg equilibrium c<strong>on</strong>diti<strong>on</strong>s. Interested readers can refer<br />
Dockner et al. (2000) for details regarding P<strong>on</strong>tryagin’s approach to solving sequential<br />
91
HANS W. GOTTINGER<br />
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
8 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
differential game problems. The equilibrium results are expressed in the form of<br />
properties as given below.<br />
4.1 Equilibrium results<br />
Property 1: The maximised costate variables for the follower are functi<strong>on</strong> of time <strong>and</strong> are<br />
given by,<br />
E<br />
9 $ E e<br />
(19)<br />
'<br />
E<br />
% 2t<br />
%'<br />
E e<br />
9 '<br />
$ (20)<br />
2 %2<br />
% t<br />
where<br />
E 9 '<br />
<strong>and</strong> E 9<br />
2<br />
are the costate variables reflecting the marginal price for a unit<br />
increase in follower firm’s own state <strong>and</strong> the state of the leader i, E 1<br />
(0) $ E 01<br />
, E 2<br />
(0) $ E 02<br />
are known c<strong>on</strong>stants.<br />
Property 2: The Stackelberg equilibrium investment by the follower in product<br />
development is given as:<br />
*<br />
* 9 * %<br />
'8!' "<br />
2t<br />
%#<br />
! " &<br />
2<br />
# E '<br />
%'<br />
u t $ ( K e ) (21)<br />
Next the leader’s problem is investigated. The leader knows the follower’s best resp<strong>on</strong>se<br />
to each c<strong>on</strong>trol path u i (3).<br />
Property 3: The maximised costate variables for the leader are given by,<br />
F<br />
F<br />
F e<br />
9 %'<br />
t<br />
'<br />
$<br />
%'<br />
(22)<br />
9 % 2t<br />
2<br />
$ F<br />
%2e<br />
(23)<br />
F F e<br />
%<br />
9 $ % 2t<br />
(24)<br />
9 %9<br />
where F 9 ( '<br />
F 9 2<br />
<strong>and</strong> F 9<br />
9<br />
are the costate variables reflecting the marginal price for a unit<br />
increase in the leader’s own state, the state of the follower <strong>and</strong> the costate of the follower.<br />
F 01<br />
, F 02<br />
<strong>and</strong> F 03<br />
are c<strong>on</strong>stants.<br />
Property 4: The Stackelberg equilibrium investment by the leader in product<br />
development is given as:<br />
i<br />
u ! t"<br />
t<br />
i %<br />
'8!' "<br />
' t<br />
%#<br />
9<br />
& K'<br />
# F<br />
%'<br />
e '<br />
$ : % 2t<br />
;<br />
'%<br />
-<br />
2<br />
F<br />
%2e<br />
( )<br />
Property 5: The equilibrium state trajectory of performance improvement of the follower<br />
is given as:<br />
(25)<br />
&<br />
x t % x e -<br />
i<br />
'8!'%#<br />
"<br />
%% 2t<br />
i %'<br />
t<br />
* e<br />
&<br />
2<br />
'# F '<br />
%'<br />
! " :<br />
*<br />
% t K e<br />
$<br />
2 %<br />
/ 7% '/ 8 2 : % ;<br />
2t<br />
%<br />
2<br />
: ('%<br />
-<br />
2<br />
F<br />
%2e<br />
)<br />
(<br />
% 2t * % 2t<br />
7 ' e 8 K2 7K2 # E%'<br />
e 8<br />
/ % /<br />
* * 2<br />
8!' % "<br />
# #<br />
'<br />
;)<br />
(26)<br />
92
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
HANS W. GOTTINGER<br />
High speed technology competiti<strong>on</strong> 9<br />
Property 6: The equilibrium state trajectory of performance improvement of the leader is<br />
given as:<br />
i i<br />
'<br />
i %<br />
8!' "<br />
' t<br />
# %#<br />
%% t<br />
i e<br />
&<br />
i<br />
%'<br />
i<br />
& K'<br />
# F<br />
%'<br />
e '<br />
'<br />
x ! t " $ :%' x% / 7% '/ e 8 K'<br />
: % 2t<br />
;<br />
;<br />
%' : ('%<br />
-<br />
2<br />
F<br />
%2e<br />
) ;<br />
( )<br />
For the analysis, the leader <strong>and</strong> the follower are assumed to be symmetric <strong>and</strong> therefore<br />
the corresp<strong>on</strong>ding values for the follower j are also assumed to be the same.<br />
The first observati<strong>on</strong> is that the costate variable of both the leader <strong>and</strong> the follower<br />
increases more rapidly with time<br />
The marginal utility of a unit increase in quality, exhibits a c<strong>on</strong>vex increasing trajectory.<br />
With comparable values of parameters the plot of the costate trajectory of the follower<br />
has an upward exp<strong>on</strong>ential drift with time. The plot begins at time t $ T r<br />
, since the<br />
follower initiates product development activities <strong>on</strong>ly after the leader has already<br />
launched the product. Regarding the costate trajectory of the leader we see that F 1<br />
represents marginal utility from a unit increase in quality of the leader’s product.<br />
The costate variable for both the leader <strong>and</strong> the follower firm increase for the<br />
entire planning horiz<strong>on</strong>. This follows since the leader <strong>and</strong> follower are competing <strong>on</strong> the<br />
basis of their quality levels. Firms increase their market shares based <strong>on</strong> their<br />
relative quality to that of the competitor’s. After launch the product faces its introducti<strong>on</strong><br />
<strong>and</strong> growth stage.<br />
In these phases, maintaining higher quality becomes even more important since the<br />
product sales are directly influenced by product quality. This leads to a c<strong>on</strong>vex<br />
increasing trajectory for the costate variable. Thus, as a sec<strong>on</strong>d observati<strong>on</strong>.<br />
The investment in innovati<strong>on</strong> made by the follower increases more rapidly with time<br />
The investment strategies of the follower are purely a functi<strong>on</strong> of its own costate<br />
variable. The follower’s costate variable increases in a c<strong>on</strong>vex fashi<strong>on</strong>. This in turn<br />
results in a c<strong>on</strong>vex increase in follower’s c<strong>on</strong>trol trajectory. With identical parameter<br />
values we can derive that the follower compensates for a delayed entry into the market<br />
by increasing its investment intensity. Such an increase in investment results in increased<br />
quality <strong>and</strong> therefore high total revenue for the follower. Thus we observe.<br />
The investment in innovati<strong>on</strong> made by the leader initially increases rapidly with time but<br />
later increases at a decreasing rate<br />
The leader’s investment trajectory is sigmoidal or S-shaped. Because of the nature of the<br />
game, the leader enters the market before the follower. While formulating his<br />
equilibrium investment strategy, the leader takes into account the evoluti<strong>on</strong> of his own<br />
costate <strong>and</strong> also the costate of the follower.<br />
Being early in the market, the leader takes into account all possible courses of acti<strong>on</strong><br />
that a follower may choose.<br />
We note that the leader’s rate of investment increases in the initial phase <strong>and</strong> then<br />
starts to decrease. That is, the leader capitalises <strong>on</strong> the advantage of early market entry by<br />
increasing the intensity of investments <strong>and</strong> gaining a high market share. Subsequently,<br />
after the follower’s entry, the leader reduces the intensity of investments <strong>and</strong> thereby<br />
reduces the amount of spillover that is potentially possible.<br />
(27)<br />
93
HANS W. GOTTINGER<br />
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
10 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
When the parameter values of the leader <strong>and</strong> follower are identical, the follower invests<br />
higher than the leader<br />
In the course of time, initially the follower maps its investment strategy to that of the<br />
leader. However, subsequently the leader starts reducing the rate of investments, while<br />
the follower c<strong>on</strong>tinues with the high investment rate based strategy.<br />
As a further observati<strong>on</strong>, the rate of increase of the follower’s product quality increases<br />
with time<br />
The quality of follower’s product increases for the entire time-horiz<strong>on</strong>. Moreover,<br />
this increase is a c<strong>on</strong>vex functi<strong>on</strong>. As can be inferred from the problem formulati<strong>on</strong>, the<br />
quality level of the follower is a functi<strong>on</strong> of investments made by the leader <strong>and</strong> the<br />
follower. Specifically, the quality trajectory is influenced by both the c<strong>on</strong>vex profile of<br />
investments made by the follower <strong>and</strong> also .<br />
by the spillover effect of the investments<br />
made by the leader. This results in a c<strong>on</strong>vex increase in quality improvements. The state<br />
trajectory is therefore upward sloping in c<strong>on</strong>vex form.<br />
The follower starts accruing revenue <strong>on</strong>ly after T p<br />
/ ?, where T p<br />
is the date of launch<br />
of the leader. In the game c<strong>on</strong>text of quality-based competiti<strong>on</strong>, the follower<br />
compensates for the delayed entry by increasing the quality levels at a fast rate.<br />
The rate of increase of the leader’s product quality decreases with time<br />
The leader’s investment trajectory follows an S-shaped trajectory. In the problem<br />
formulati<strong>on</strong>, the leader's quality improvement is <strong>on</strong>ly a functi<strong>on</strong> of its own investments.<br />
Moreover, it is assumed that the leader doesn’t obtain the advantages of spillover of<br />
knowledge gained by the follower’s investments. Furthermore, the leader needs to<br />
deliberately avoid maintaining very high investments to ensure that the follower<br />
doesn’t achieve huge gains from spillovers. Under such a setup it must be noted<br />
that the leader chooses an S-shaped trajectory for investments. In such an investment<br />
profile the investment increases rapidly in the initial phase but then eventually<br />
tapers off. Corresp<strong>on</strong>ding to this investment profile, the rate of quality improvement<br />
is high in the initial phase but then eventually the rate of improvement starts<br />
decreasing. Thus, the quality trajectory of the leader is c<strong>on</strong>cave. This leads to another<br />
observati<strong>on</strong>.<br />
When the parameter values of the leader <strong>and</strong> follower are identical, the quality of the<br />
follower’s product is higher than that of the leader’s product<br />
With identical parameter values the follower <strong>and</strong> leader are perfectly symmetric in their<br />
capabilities. Moreover, the leader has the advantage of earlier market entry, whereas the<br />
follower obtains the gains of informati<strong>on</strong> asymmetry <strong>and</strong> the associated knowledge<br />
spillovers from the leader. It is reas<strong>on</strong>able to assume that the follower employs his<br />
inherent competence <strong>and</strong> the benefit of spillover of knowledge from the leader’s<br />
investments to increase quality at a much faster rate. Under such circumstances the<br />
overall gain <strong>and</strong> loss in the competitive game is dictated by the relative revenues<br />
achieved <strong>and</strong> the costs incurred by the two players. With higher investments than the<br />
leader the follower incurs higher costs. At the same time these investments also lead to<br />
higher quality levels <strong>and</strong> therefore revenues for the follower. In the event of the value<br />
of R 1<br />
being very high the follower wins the game while the results favour the leader if<br />
R 1<br />
is low.<br />
94
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
HANS W. GOTTINGER<br />
High speed technology competiti<strong>on</strong> 11<br />
5 Firm asymmetries<br />
We now turn to firm asymmetries <strong>and</strong> their implicati<strong>on</strong>s. It is likely that firms take up the<br />
role of leader <strong>and</strong> follower based <strong>on</strong> inherent strengths <strong>and</strong> weaknesses. Therefore, an<br />
explicit c<strong>on</strong>siderati<strong>on</strong> of firm asymmetries is very important <strong>and</strong> could potentially lend<br />
more insights in a sequential game setup. Asymmetries may be addressed in at least four<br />
ways: Firstly, it is possible that organisati<strong>on</strong>al approaches <strong>and</strong> techniques (such as total<br />
quality management) can be used to make the product development process more<br />
cost-efficient <strong>and</strong> effective. This would influence the value of the resource productivity<br />
parameter #. Sec<strong>on</strong>dly, it is possible to have an advantage in product development by<br />
making capital investment in technology development. This is equivalent to c<strong>on</strong>sidering<br />
asymmetries in the value of K. Thirdly, it is possible that the obsolescence parameter<br />
of firms denoted by L is different. Finally, different values of M suggest different<br />
levels of spillover of knowledge from the leader’s investments to the follower. The<br />
firm with higher value of L has a higher obsolescence or decay in quality. We<br />
restrict assessment of asymmetries by c<strong>on</strong>sidering different values for the parameters:<br />
(# i <strong>and</strong> # j ), (K 1<br />
<strong>and</strong> K 2<br />
), (L 1<br />
<strong>and</strong> L 2<br />
) <strong>and</strong> (M l<br />
<strong>and</strong> M 2<br />
).<br />
5.1 Innovati<strong>on</strong> resource productivity<br />
Parameters # i <strong>and</strong> # j denote the innovati<strong>on</strong> resource productivity parameter of the two<br />
firms. An asymmetry could result if the competing firms have differing capabilities in<br />
making a productive use of investments. The skill set of employees, training <strong>and</strong><br />
development activities, the quality culture are some of the reas<strong>on</strong>s for such an<br />
asymmetry. For example, (# i $ 0.2) G (# j $ 0.1) suggest that firm i has a higher<br />
innovati<strong>on</strong> productivity over j. An investigati<strong>on</strong> of different values of # suggest that the<br />
firm with higher innovati<strong>on</strong> productivity also invests a higher amount in product<br />
development. From the results we observe that the follower has a relatively higher level<br />
of investments in innovati<strong>on</strong> as compared to the leader.<br />
Therefore, an increase in the value of # of the follower will <strong>on</strong>ly result in making<br />
these investments still higher. Instead, an increase in leader’s # would provide some<br />
interesting insights in that it will show that the leader sails ahead with the follower<br />
having no chance in catching up. It clearly shows that the leader has a higher investment<br />
than that of the follower for almost the entire planning horiz<strong>on</strong>. As can also be noted, the<br />
quality of the leader’s product is always higher than that of the follower. Hence, an<br />
asymmetry in terms of resource productivity parameters provides a totally different result<br />
than what was observed when the competing firms were symmetric. It suggests that the<br />
leader with an advantage in terms of early market entry as well as a higher level of<br />
resource productivity indeed c<strong>on</strong>tinues having higher revenues. Under this situati<strong>on</strong> the<br />
leader doesn’t worry much about the spillover to the follower since its own resource<br />
productivity enables attaining higher revenues by increasing quality levels relative to that<br />
of the follower. With a high value of R 1<br />
it can be c<strong>on</strong>jectured that the result favours<br />
the leader.<br />
5.2 Capital investment parameters<br />
Firm asymmetries can be analysed by evaluating different values for K <strong>and</strong> the<br />
implicati<strong>on</strong>s <strong>on</strong> c<strong>on</strong>trol <strong>and</strong> state trajectories. K l<br />
<strong>and</strong> K 2<br />
are proporti<strong>on</strong>al to the level of<br />
95
HANS W. GOTTINGER<br />
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
12 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
capital investment in development technology by the two competing firms. If<br />
(K 1<br />
$ 20) G (K 2<br />
$ 10), it suggests that the leader i has higher levels of capital<br />
investment in development technology as compared to the follower j. The implicati<strong>on</strong>s of<br />
a higher value of K for the leader can easily be derived. The parameter K exerts a<br />
positive effect <strong>on</strong> innovati<strong>on</strong> investments. Similar to the c<strong>on</strong>jecture regarding different<br />
values of #, the intuiti<strong>on</strong> behind this effect is that with a higher level of capital<br />
investments in development technology, a firm targets the investments towards<br />
increasing the product quality. Thus with a higher value of K the investment by the<br />
leader is relatively higher than that of the follower for most part of the planning horiz<strong>on</strong>.<br />
However, in the latter part of the planning horiz<strong>on</strong>, the follower’s investment in fact<br />
shoots up while that of the leader tapers off. Such an investment profile would have an<br />
impact <strong>on</strong> the state trajectory. Furthermore, the product quality of the leader is higher<br />
than that of the follower for most part of the planning horiz<strong>on</strong>. Eventually, at the end<br />
of the planning horiz<strong>on</strong>, the high growth in the lower’s investment results in higher<br />
quality than that of the leader. In the model, the relative difference in resources due to<br />
strategic investments is captured in asymmetries in the value of K. Given adequately high<br />
values of R 1<br />
, the leader i would have relatively higher profits than the follower j.<br />
5.3 The obsolescence parameter<br />
L 1<br />
<strong>and</strong> L 2<br />
characterise the decay or in other words the obsolescence effect of quality.<br />
Different values of L 1<br />
<strong>and</strong> L 2<br />
help evaluate the difference am<strong>on</strong>g the two firms regarding<br />
this effect. (L 1<br />
$ 1) G (L 2<br />
$ 0.7) suggests that the obsolescence effect for leader i is higher<br />
than that for the follower j. An investigati<strong>on</strong> into the dynamics of innovati<strong>on</strong> investments<br />
reveals that with higher value of the parameter L, a firm would make higher level of<br />
investments. Interestingly, with a difference in L 1<br />
<strong>and</strong> L 2<br />
the shape of investment<br />
trajectory of the leader is now c<strong>on</strong>vex. That is when the leader faces a high obsolescence<br />
effect the investments are increased at a faster rate to ensure an increasing state<br />
trajectory. However, as opposed to parameter a <strong>and</strong> K such an increased c<strong>on</strong>trol<br />
trajectory doesn’t translate into higher quality. In fact, in this situati<strong>on</strong>, the leader has<br />
lower product quality inspite of higher investments. A high decay effect puts the leader<br />
into a disadvantageous positi<strong>on</strong>. In such a situati<strong>on</strong>, under equilibrium c<strong>on</strong>trol the leader<br />
always loses the competitive game.<br />
5.4 Spillover factor M 2<br />
M 2<br />
characterises the amount of spillover from the leader to the follower. In this secti<strong>on</strong><br />
the impact of reducti<strong>on</strong> of the spillover factor M 2<br />
is c<strong>on</strong>sidered. Evidently in this case the<br />
leader would be less worried about the amount of advantage the follower obtains from<br />
the leader’s own investments. If the spillover effect is low, the leader pursues<br />
investments more aggressively, c<strong>on</strong>sistent with the incentive enhancement of spillover<br />
effects in the c<strong>on</strong>text of network externalities (Saaskilahti, 2006). Under such<br />
circumstances, the c<strong>on</strong>trol trajectory of the leader is also c<strong>on</strong>vexly increasing <strong>and</strong> is <strong>on</strong>ly<br />
marginally lower than that of the follower.<br />
Regarding the impact of such investments <strong>on</strong> the state trajectory, the quality of the<br />
follower is higher than that of the leader <strong>and</strong> both obtain a c<strong>on</strong>vex increasing state<br />
trajectory. But unlike #, K <strong>and</strong> L, a different value of M 2<br />
doesn’t corresp<strong>on</strong>d to any<br />
advantage related to resources, core competence or dynamic capabilities. This is purely<br />
96
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
HANS W. GOTTINGER<br />
High speed technology competiti<strong>on</strong> 13<br />
an exogenously defined variable based <strong>on</strong> technology <strong>and</strong> industry types. Under changed<br />
circumstances with respect to M 2<br />
the leader doesn’t really gain much in the competitive<br />
game. The results in fact point to the follower winning the game due to much higher<br />
levels of quality <strong>and</strong> <strong>on</strong>ly marginally higher levels of investments.<br />
6 C<strong>on</strong>clusi<strong>on</strong>s <strong>and</strong> extensi<strong>on</strong>s<br />
There is apparently a paradox in trying to assess, both empirically <strong>and</strong> theoretically, the<br />
impact of competitive pressure <strong>on</strong> innovati<strong>on</strong> <strong>and</strong> growth. On <strong>on</strong>e h<strong>and</strong>, according to the<br />
traditi<strong>on</strong> originating in Schumpeter (1942) the prospective reward provided by m<strong>on</strong>opoly<br />
rent to a successful innovator is required to stimulate sufficient R&D investment <strong>and</strong><br />
technological progress.<br />
On the other h<strong>and</strong>, the incentives to innovate are weaker for an incumbent<br />
m<strong>on</strong>opolist than for a firm in a competitive industry (Arrow, 1962). When competiti<strong>on</strong> is<br />
intense in the product market, innovati<strong>on</strong> may even be seen as the <strong>on</strong>ly way for a firm to<br />
survive. In neo-Schumpeterian models of endogenous growth (Aghi<strong>on</strong> <strong>and</strong> Howitt, 1992;<br />
Grossman <strong>and</strong> Helpman, 1991), innovati<strong>on</strong> allows a firm in an industry to take the lead<br />
<strong>and</strong> gain profit. But the m<strong>on</strong>opoly rent enjoyed by the winner is <strong>on</strong>ly temporary, <strong>and</strong> a<br />
new innovator, capitalising <strong>on</strong> accumulated knowledge, is always able to ‘leapfrog’ the<br />
leader unless the leader is endowed with advantages of firm asymmetries. In the recent<br />
research literature, Aghi<strong>on</strong> et al. (1997) <strong>and</strong> Aghi<strong>on</strong> et al. (2001), supposing a duopoly in<br />
each sector, both at the research <strong>and</strong> producti<strong>on</strong> levels, have introduced what they call<br />
‘step-by-step innovati<strong>on</strong>’, according to which technological progress allows a firm to<br />
take the lead, but with the lagging firm remaining active <strong>and</strong> eventually capable of<br />
catching up. This model has been extended by Encaoua <strong>and</strong> Ulph (2000), allowing for<br />
the possibility that the lagging firm leapfrogs the leader, without driving it out of the<br />
market. Here we adopt a simplified approach to c<strong>on</strong>sider spillover effects. Griliches<br />
(1979) lays out the c<strong>on</strong>ceptual framework <strong>and</strong> provides an early discussi<strong>on</strong> of the<br />
importance of spillover effects of R&D. Later in Griliches (1992) reviewed the recent<br />
empirical evidence <strong>on</strong> spillovers, <strong>and</strong> tentatively c<strong>on</strong>cluded that spillover effects may be<br />
substantial. The model can be extended by evaluating the improvement in product quality<br />
with learning effect. This effect is very well documented in the literature <strong>and</strong> thus the<br />
extensi<strong>on</strong> to incorporate learning effects is straightforward. Specifically, learning can be<br />
used to characterise the dynamics of evoluti<strong>on</strong> of product quality <strong>and</strong> the characterisati<strong>on</strong><br />
of costs associated with innovati<strong>on</strong> investments.<br />
For analytical simplicity the revenue functi<strong>on</strong> is treated as a salvage value. As an<br />
extensi<strong>on</strong> the incorporati<strong>on</strong> of the dynamics of revenue in the analysis of the model can<br />
be explored. In the sequential play game, the state dynamics of the follower can be<br />
c<strong>on</strong>sidered to be dependent <strong>on</strong> the state of the leader at previous time instant.<br />
This modificati<strong>on</strong> would allow for analysis of a model in which the leader uses<br />
open-loop Nash equilibrium strategy for innovati<strong>on</strong> investments whereas the follower<br />
would adopt a Markovian Nash equilibrium strategy.<br />
Additi<strong>on</strong>al insights can be gained by this modificati<strong>on</strong> for leader-follower<br />
competitive dynamics in new product development. Firm asymmetries can be explored<br />
by c<strong>on</strong>sidering multiple parameters at the same time. This would enable a richer<br />
underst<strong>and</strong>ing of the strategies that a leader <strong>and</strong> a follower should adopt based <strong>on</strong> their<br />
strengths <strong>and</strong> weaknesses.<br />
97
HANS W. GOTTINGER<br />
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
14 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
The c<strong>on</strong>venti<strong>on</strong>al approach used in innovati<strong>on</strong> <strong>and</strong> R&D research viewed the process<br />
as <strong>on</strong>e with c<strong>on</strong>stant returns, competitive output <strong>and</strong> factor markets <strong>and</strong> no externalities.<br />
However, such a framework doesn’t offer a full explanati<strong>on</strong> of productivity growth.<br />
For a better underst<strong>and</strong>ing it therefore becomes very important to c<strong>on</strong>sider increasing<br />
returns to scale, R&D spillovers <strong>and</strong> other externalities <strong>and</strong> disequilibria.<br />
References<br />
Aghi<strong>on</strong>, P. <strong>and</strong> Howitt, P. (1992) ‘A model of growth through creative destructi<strong>on</strong>’, Ec<strong>on</strong>ometrica,<br />
Vol. 60, pp.323–352.<br />
Aghi<strong>on</strong>, P., Harris, C. <strong>and</strong> Vickers, J. (1997) ‘Competiti<strong>on</strong> <strong>and</strong> growth with step-by-step<br />
innovati<strong>on</strong>: an example’, Review of Ec<strong>on</strong>omic Studies, Vol. 41, pp.771–782.<br />
Aghi<strong>on</strong>, P., Harris, C., Howitt, P. <strong>and</strong> Vickers, J. (2001) ‘Competiti<strong>on</strong>, imitati<strong>on</strong> <strong>and</strong> growth with<br />
step-by-step innovati<strong>on</strong>’, Review of Ec<strong>on</strong>omic Studies, Vol. 68, pp.467–492.<br />
Arrow, K. (1962) ‘The ec<strong>on</strong>omic implicati<strong>on</strong>s of learning by doing’, Review of Ec<strong>on</strong>omic Studies,<br />
Vol. 29, pp.155–173.<br />
Arthur, W.B. (2000) ‘Myths <strong>and</strong> realities of the high-tech ec<strong>on</strong>omy’, Talk given at Credit Suisse<br />
First Bost<strong>on</strong> (CSFB) Thought Leader Forum, 10 September.<br />
Dockner, E., Jorgensen, S., VanL<strong>on</strong>g, N. <strong>and</strong> Sorger, G. (2000) Differential Games in Ec<strong>on</strong>omics<br />
<strong>and</strong> Management Science, Cambridge: Cambridge University Press.<br />
Encaoua, D. <strong>and</strong> Ulph, D. (2000) ‘Catching-up or leapfrogging’, Cahiers de la MSE, Vol. 5,<br />
pp.10–22.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H.W. (2006) Innovati<strong>on</strong>, <strong>Technology</strong> <strong>and</strong> Hypercompetiti<strong>on</strong>, L<strong>on</strong>d<strong>on</strong>: Routledge.<br />
Griliches, Z. (1979) ‘Issues in assessing the c<strong>on</strong>tributi<strong>on</strong> of Research <strong>and</strong> <strong>Development</strong> to<br />
productivity growth’, The Bell Journal of Ec<strong>on</strong>omics, Vol. 10, No. 1, pp.92–116.<br />
Griliches, Z. (1992) ‘The search of R&D spillovers’, The Sc<strong>and</strong>inavian Journal of Ec<strong>on</strong>omics,<br />
Vol. 94, pp.S29–S47.<br />
Grossman, G. <strong>and</strong> Helpman, E. (1991) Innovati<strong>on</strong> <strong>and</strong> Growth in the Global Ec<strong>on</strong>omy, Cambridge,<br />
MA: MIT Press.<br />
Hoffman, W.D. (1976) ‘Market structure <strong>and</strong> strategies of R&D behavior in the data processing<br />
market – theoretical thoughts <strong>and</strong> empirical findings’, Research Policy, Vol. 5, pp.334–353.<br />
Kanter, R.M. (2006) ‘Innovati<strong>on</strong>: the classical traps’, Harvard Business Review, November,<br />
pp.72–83.<br />
Robs<strong>on</strong>, A.J. (1990) ‘Duopoly with endogeneous strategic timing: stackelberg regained’,<br />
Internati<strong>on</strong>al Ec<strong>on</strong>omic Review, Vol. 31, pp.25–35.<br />
Saaskilahti, P. (2006) ‘Strategic R&D <strong>and</strong> network compatibility’, Ec<strong>on</strong>omics of Innovati<strong>on</strong> <strong>and</strong><br />
New <strong>Technology</strong>, Vol. 15, No. 8, pp.711–733.<br />
Schumpeter, J.A. (1942) Capitalism, Socialism <strong>and</strong> Democracy, New York: Harper.<br />
Teece, D.J., Pisano, G. <strong>and</strong> Shuen, A. (1997) ‘Dynamic capabilities <strong>and</strong> strategic management’,<br />
Strategic Management Journal, Vol. 18, No. 7, pp.509–533.<br />
Varian, H.R., Farrell, J. <strong>and</strong> Shapiro, C. (2004) The Ec<strong>on</strong>omics of Informati<strong>on</strong> <strong>Technology</strong>,<br />
Cambridge: Cambridge University Press.<br />
98
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
HANS W. GOTTINGER<br />
Appendix<br />
High speed technology competiti<strong>on</strong> 15<br />
Mathematical Preliminaries, Definiti<strong>on</strong>s <strong>and</strong> Theorems<br />
This secti<strong>on</strong> covers some of the salient analytical aspects specific to a sequential play<br />
game.<br />
For a finite horiz<strong>on</strong> T let L <strong>and</strong> F denote the leader <strong>and</strong> follower respectively. Let<br />
x denote the vector of state variables, u L the vector of c<strong>on</strong>trol variables of the leader <strong>and</strong><br />
u F the vector of c<strong>on</strong>trol variables of the follower. Assume x 6 R n , u L 6 R mL <strong>and</strong> u F 6 R mF .<br />
Definiti<strong>on</strong> 1: The initial value of the follower’s costate variable E i<br />
is said to be<br />
n<strong>on</strong>-c<strong>on</strong>trollable if E i<br />
(0) is independent of the leader’s c<strong>on</strong>trol path u L (t). Otherwise, it is<br />
said to be c<strong>on</strong>trollable.<br />
The definiti<strong>on</strong> suggests that if the costate variable is c<strong>on</strong>trollable, the follower’s c<strong>on</strong>trol<br />
variable u F (t) at time t depends also <strong>on</strong> future values of u L (3); that is, <strong>on</strong> values u L (s) with<br />
s G t.<br />
Theorems <strong>and</strong> Proofs<br />
Theorem 1: The maximised costate variables for the follower are a functi<strong>on</strong> of time <strong>and</strong><br />
is given by<br />
where<br />
E<br />
9 $ E e<br />
(A1)<br />
'<br />
E<br />
% 2t<br />
%'<br />
E e<br />
9 '<br />
$ (A2)<br />
2 %2<br />
% t<br />
E 9 '<br />
<strong>and</strong> E 9<br />
2<br />
are the costate variables reflecting the marginal price for a unit<br />
increase in the follower firm's own state <strong>and</strong> the state of the leader i; E 1<br />
(0) $ E 01<br />
, E 2<br />
(0) $<br />
E 02<br />
are c<strong>on</strong>stants.<br />
Proof: Stackelberg equilibrium c<strong>on</strong>diti<strong>on</strong>s are derived by c<strong>on</strong>structing the Hamilt<strong>on</strong>ians.<br />
The analytical soluti<strong>on</strong> is derived for the follower firm j by writing the Hamilt<strong>on</strong>ian as:<br />
*<br />
* * #<br />
! " E &<br />
* * i<br />
/ $ % & '<br />
' 2<br />
! "<br />
2<br />
! "<br />
2<br />
! "'<br />
( u t ) / K u t % % x t / - u t<br />
( )<br />
i<br />
#<br />
/ E &<br />
i<br />
i<br />
2<br />
K' u ! t " % %'<br />
x ! t"<br />
'<br />
( )<br />
(A3)<br />
The necessary c<strong>on</strong>diti<strong>on</strong>s for optimality are:<br />
*<br />
/ $ %<br />
(A4)<br />
*<br />
u<br />
From (A5)<br />
E $ % / : E $ % /<br />
(A5)<br />
9 * 9 *<br />
' *<br />
x 2<br />
i<br />
x<br />
x !%" $ x<br />
(A6)<br />
* *<br />
%<br />
E E E E<br />
9 9 % 2t<br />
'<br />
$ %<br />
2 ':<br />
'<br />
$<br />
%'<br />
e<br />
(A7)<br />
E E E E<br />
9 9 %'<br />
t<br />
2<br />
$ %' 2:<br />
2<br />
$<br />
%2e<br />
(A8)<br />
99
HANS W. GOTTINGER<br />
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
16 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
where E 1<br />
(0) $ E 01<br />
is the known positive c<strong>on</strong>stant denoting the initial value of the<br />
costate E 1<br />
<strong>and</strong> E 2<br />
(0) $ E 02<br />
is a known negative c<strong>on</strong>stant denoting the initial value of the<br />
costate E 2<br />
.<br />
Theorem 2: The Stackelberg equilibrium investment by the follower in product<br />
development is given as<br />
*<br />
* 9 * %<br />
'8!' "<br />
2t<br />
%#<br />
! " &<br />
2<br />
# E '<br />
%'<br />
u t $ (<br />
K e )<br />
(A9)<br />
Proof: Differentiating (A3) with respect to u j (t), we obtain:<br />
*<br />
* *<br />
# %'<br />
u*<br />
$ %'/ E<br />
'#<br />
&<br />
2 ( ! ")<br />
'<br />
/ K u t<br />
(A10)<br />
By Equati<strong>on</strong> (A10) to zero <strong>and</strong> some algebraic manipulati<strong>on</strong>s, the following expressi<strong>on</strong> is<br />
obtained for optimal effort in product development:<br />
* 9 * 9<br />
2 '<br />
*<br />
' 8!' %#<br />
"<br />
u ! t"<br />
$ & ( K # E ' )<br />
(A11)<br />
Substituting the expressi<strong>on</strong> for E i<br />
in (A11):<br />
*<br />
* 9 * %<br />
'8!' "<br />
2t<br />
%#<br />
! " &<br />
2<br />
# E '<br />
%'<br />
u t $ ( K e )<br />
(A12)<br />
Theorem 3: The maximised costate variables for the leader are give n by,<br />
F<br />
9 $ F e<br />
(A13)<br />
'<br />
%'<br />
t<br />
%'<br />
F<br />
F e<br />
9 % 2t<br />
2<br />
$<br />
%2<br />
(A14)<br />
F F %<br />
9 $ % % 2t<br />
9 %9e<br />
(A15)<br />
where F 9 ( '<br />
F 9 2<br />
<strong>and</strong> F 9<br />
9<br />
are the costate variables reflecting the marginal price for a unit<br />
increase in the leader’s own state, the state of the follower <strong>and</strong> the costate of the<br />
follower; F 01<br />
, F 02<br />
<strong>and</strong> F 03<br />
are c<strong>on</strong>stants.<br />
Proof: The Stackelberg equilibrium c<strong>on</strong>diti<strong>on</strong>s are derived by c<strong>on</strong>structing the<br />
Hamilt<strong>on</strong>ian. An analytical soluti<strong>on</strong> is derived for the leader firm I by writing the<br />
Hamilt<strong>on</strong>ian as:<br />
/ & u t ' & K u t % x t '<br />
i<br />
i i i # i<br />
$ % ( ! ") / F<br />
'<br />
! " %<br />
'<br />
! "<br />
( )<br />
/ F & % / ' / F E<br />
( )<br />
*<br />
* #<br />
* i<br />
2<br />
K2u ! t " %<br />
2x ! t" -<br />
2u ! t "<br />
9 '%<br />
2<br />
@<br />
A<br />
(A16)<br />
The necessary c<strong>on</strong>diti<strong>on</strong>s for optimality are<br />
i<br />
/ $ %<br />
(A17)<br />
i<br />
u<br />
F $ %/ : F $ %/ : F $ %/ E<br />
(A18)<br />
9 i 9 i 9 i<br />
' i 2 *<br />
x<br />
x 9<br />
'<br />
i<br />
x !%" $ x<br />
(A19)<br />
i<br />
%<br />
100
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
HANS W. GOTTINGER<br />
From (A18)<br />
High speed technology competiti<strong>on</strong> 17<br />
F F F F<br />
9 9 %'<br />
t<br />
'<br />
$ %' ':<br />
'<br />
$<br />
%'<br />
e<br />
(A20)<br />
F F F F<br />
9 9<br />
% 2t<br />
2<br />
$ %<br />
2 2:<br />
2<br />
$<br />
%2e<br />
(A21)<br />
9 % 2t<br />
9<br />
$ %<br />
9% 2:<br />
9<br />
9<br />
$ %<br />
%9e<br />
(A22)<br />
F F F F %<br />
where F 9 ( '<br />
F 9 2<br />
<strong>and</strong> F 9<br />
9<br />
are the costate variables reflecting the marginal price for a unit<br />
increase in the leader’s own state, the state of the follower <strong>and</strong> the costate of the<br />
follower; F 01<br />
, F 02<br />
<strong>and</strong> F 03<br />
are c<strong>on</strong>stants.<br />
Theorem 4: The Stackelberg equilibrium investment by the leader in product<br />
development is given as:<br />
i<br />
u ! t"<br />
i<br />
i %<br />
'8!' "<br />
' t<br />
%#<br />
9<br />
& K'<br />
# F<br />
%'<br />
e '<br />
$ : % 2t<br />
;<br />
'%<br />
-<br />
2<br />
F<br />
%2e<br />
( )<br />
Proof: Differentiating (A3) with respect to u i (t), we obtain:<br />
(A23)<br />
i<br />
i i<br />
# %'<br />
& '<br />
' '<br />
F<br />
2 2<br />
/ i $ %'/ F # K u ! t"<br />
/ -<br />
u<br />
( )<br />
(A24)<br />
By Equati<strong>on</strong> (A24) to zero <strong>and</strong> some algebraic manipulati<strong>on</strong>s, the following expressi<strong>on</strong> is<br />
obtained for optimal effort in product development:<br />
i<br />
u ! t"<br />
i<br />
i 9<br />
'8!'%#<br />
"<br />
9<br />
& K'<br />
# F '<br />
'<br />
$ : 9 ;<br />
'%<br />
-<br />
2F<br />
2<br />
( )<br />
(A25)<br />
Substituting the expressi<strong>on</strong> for<br />
i<br />
u ! t"<br />
i<br />
i %<br />
'8!' "<br />
' t<br />
%#<br />
9<br />
& K'<br />
# F<br />
%'<br />
e '<br />
$ : % 2t<br />
;<br />
'%<br />
-<br />
2<br />
F<br />
%2e<br />
( )<br />
F 9 '<br />
<strong>and</strong> F 9 2<br />
in (A25) yields<br />
(A26)<br />
Theorem 5: The equilibrium state trajectory of performance improvement of the follower<br />
is given as<br />
&<br />
* *<br />
x ! t " $ % x / %'/ e - : ;<br />
i<br />
'8!'%<br />
# "<br />
%% 2t<br />
i %'<br />
t<br />
e<br />
&<br />
2<br />
'# F '<br />
:<br />
% t K<br />
%'<br />
e<br />
2 % 7 8 2<br />
% 2t<br />
% :<br />
2 (: '%-2 F%2e<br />
;<br />
(<br />
)<br />
* * 2<br />
# 8!' % # "<br />
% 2t * % 2t<br />
'<br />
/ %'/ e K K # E e ; )<br />
7 8 2 7 2 %' 8<br />
(A27)<br />
Proof: The expressi<strong>on</strong> for the optimal state trajectory can be obtained by c<strong>on</strong>sidering the<br />
state dynamics given in Equati<strong>on</strong> (11). Substituting u i (t) $ u i (t) 9 <strong>and</strong> u j (t) $ u i (t) 9 in<br />
(11) the following first-order differential equati<strong>on</strong> can be obtained<br />
* i i 9<br />
* * 9<br />
# 8!' %# "<br />
*<br />
& K'<br />
# F '<br />
'<br />
x ! t" $ K &<br />
2 ( K<br />
2<br />
# E '<br />
' ) % %<br />
2x ! t " 7, p<br />
/ H % t8<br />
/ -<br />
2 : 9 ;<br />
('%<br />
-<br />
2F<br />
2 )<br />
i<br />
'8!'%#<br />
"<br />
(A28)<br />
101
HANS W. GOTTINGER<br />
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
18 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
The first order differential equati<strong>on</strong> can be solved with the initial c<strong>on</strong>diti<strong>on</strong> x<br />
* !%" $ x<br />
*<br />
%<br />
;<br />
The resulting expressi<strong>on</strong> for x j (t) 9 is:<br />
&<br />
x t % x e -<br />
i<br />
'8!'%#<br />
"<br />
%% 2t<br />
i %'<br />
t<br />
* e<br />
&<br />
2<br />
'# F '<br />
%'<br />
! " :<br />
*<br />
% t K e<br />
$<br />
2 %<br />
/ 7% '/ 8 2 : % ;<br />
2t<br />
%<br />
2<br />
: ('%<br />
-<br />
2<br />
F<br />
%2e<br />
)<br />
(<br />
% 2t * % 2t<br />
7 ' e 8 K2 7K2 # E<br />
%'<br />
e 8<br />
/ % /<br />
* * 2<br />
# 8!' %#<br />
"<br />
'<br />
;)<br />
(A29)<br />
Theorem 6: The equilibrium state trajectory of performance improvement of the leader is<br />
given as:<br />
i i<br />
'<br />
i %<br />
8!' "<br />
' t<br />
# %#<br />
%% t<br />
i e<br />
&<br />
i<br />
%'<br />
t<br />
& K'<br />
# F<br />
%'<br />
e '<br />
'<br />
x ! t " $ :%' x% / 7% '/ e 8 K'<br />
: % 2t<br />
;<br />
;<br />
%' : ('%<br />
-<br />
2<br />
F<br />
%2e<br />
) ;<br />
( )<br />
(A30)<br />
Proof: The expressi<strong>on</strong> for the optimal state trajectory can be obtained by c<strong>on</strong>sidering the<br />
state dynamics given in Equati<strong>on</strong> (14). Substituting u i (t) $ u i (t) 9 in (14) the following<br />
first-order differential equati<strong>on</strong> can be obtained:<br />
i i<br />
i 9<br />
# 8!' %#<br />
"<br />
*<br />
& K'<br />
# F '<br />
'<br />
i<br />
! " $<br />
' : '<br />
! "<br />
9 ; %<br />
'%<br />
-<br />
2F<br />
2<br />
x t K % x t<br />
( )<br />
(A31)<br />
i<br />
i<br />
The first order differential equati<strong>on</strong> can be solved with the initial c<strong>on</strong>diti<strong>on</strong> x !%" $ x%<br />
;<br />
The resulting expressi<strong>on</strong> for x i (t) 9 is:<br />
i i<br />
'<br />
i %<br />
8!' "<br />
' t<br />
# %#<br />
% t<br />
i<br />
%'<br />
t<br />
& K'<br />
# F<br />
%'<br />
e '<br />
:<br />
' % 7 8<br />
;<br />
'<br />
% 2t<br />
'<br />
: '%<br />
2<br />
F<br />
%2<br />
;<br />
%<br />
i e<br />
& '<br />
x ! t " $ % x / %'/ e K : ;<br />
% ( - e<br />
(<br />
)<br />
)<br />
(A32)<br />
102
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACING<br />
HANS W. GOTTINGER<br />
103
HANS W. GOTTINGER<br />
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACINGN<br />
104
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACING<br />
HANS W. GOTTINGER<br />
105
HANS W. GOTTINGER<br />
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACINGN<br />
106
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACING<br />
HANS W. GOTTINGER<br />
107
HANS W. GOTTINGER<br />
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACINGN<br />
108
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACING<br />
HANS W. GOTTINGER<br />
109
HANS W. GOTTINGER<br />
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACINGN<br />
110
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACING<br />
HANS W. GOTTINGER<br />
111
HANS W. GOTTINGER<br />
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACINGN<br />
112
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACING<br />
HANS W. GOTTINGER<br />
113
HANS W. GOTTINGER<br />
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACINGN<br />
114
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACING<br />
HANS W. GOTTINGER<br />
115
HANS W. GOTTINGER<br />
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACINGN<br />
116
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACING<br />
HANS W. GOTTINGER<br />
117
HANS W. GOTTINGER<br />
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACINGN<br />
118
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACING<br />
HANS W. GOTTINGER<br />
119
HANS W. GOTTINGER<br />
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACINGN<br />
120
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACING<br />
HANS W. GOTTINGER<br />
121
HANS W. GOTTINGER<br />
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACINGN<br />
122
1.8 INTERNET OF THINGS ECONOMIC AND INDUSTRIAL DIMENSIONS<br />
HANS W. GOTTINGER<br />
!"#$%"&#'("&)*+(,%"&)*(-*.,/'"$//*&"0*12("(3'2/*4$/$&%25*<br />
"#$%&!'()*+!$$),$"-!<br />
.//0+112223456785709:;64.68?1@16@:7=!<br />
A>6+!$#3$$'BC1@36@:7=3"#$%#'#)3$)!<br />
DEEF+!"-"C,%)B-!(G=68/*&!DEEF+!"-"C,%)'H!(I8;687*! !<br />
!<br />
65$*!"#$%"$#*(-*65'"7/8*12("(3'2*&"0*!"0,/#%'&)*<br />
9'3$"/'("/*<br />
!"#$%&'(#'(%)*++,#-'(%<br />
E/=J/78>?654L!EMNOMKP!P>849;/683!)L!"#$%L!003!$$),$"-3!A>6+!$#3$$'BC1@36@:7=3"#$%#'#)3$)!<br />
6'4',7'1+!O9,2JS! 5>??9865J/6>8! :75J?7!<br />
5>??>8L!2.65.!78J:;7A!947=!0J=/6560J/6>8L!5>;;J:>=J/6>8!J8A!<br />
68/7=J5/6>83! W7:! "3#! /75.8>;>56J;! 87/2>=]68;>?7!74478/6J;!/>!?>A7=8!<br />
4>56J;!68/7=J5/6>8!J4!27;;!J4!X>=!:J;!:94687443!W.6;7!W7:!<br />
"3#!59==78/;S!A>?68J/74!/.7!D8/7=87/L!87/2>=]!78=]682J=A4!J8>/.7=!J;L!5>??>8;S!=7X7==7A!/>!J4!<br />
/.7!E7?J8/65!W7:!J8A!4>?7/6?74!=7X7==7A!/>!J4!W7:!-3#3!M.7!<br />
J;! >X! /.7! E7?J8/65! W7:! (7?:=J5689=]4*!64!/>!?J=]!90!27:!5>8/78/!<br />
68!J!2JS!/.J/!?J]74!6/!98A7=4/J8AJ:;7!:S!?J5.6874L!J;;>268! :7.J[7! ?>=7! 68/7;;6=?J/4!2>9;A!J;;>2!<br />
?J5.6874!/>!0=>5744!J8A!4.J=7!AJ/J!>8!/.76=!>28L!26/.>9/!/.7!<br />
877A! X>=! .9?J8! ?7A6J/6>83! O;>80?78/4! 68!<br />
D8/7=87/! /75.8>;>;>=! F7/2>=]4! J8A!<br />
F7J=! d67;A! P>??9865J/6>8! (FdP*! 9468! d=7c9785S!<br />
DA78/6/S! (NdDY*! /J! :778! 7[>;[688[7=X! /.747! /2>! /75.8>;>=!<br />
F7/2>=]4L! 64! ;7JA68! 872! 0>446:6;6/674! J8A! [646>843! M.7!<br />
0>446:6;6/S!>X!J!X=J?72>=]!/.J/!2>9;A!J;;>2!A6=75/!?J5.687,<br />
/>,! ?J5.687! 5>??9865J/6>8! >[7=! /.7! D8/7=87/! .J4! ;7A!<br />
=747J=5.7=4! />! 78[646>8! /.7! :787X6/4! >X! :=688;687!J8A!J;;>268!0J=/6560J/7!68!/.7!27:!<br />
J4! J! [J4/! 87/2>=]! >X! J9/>8>?>94L! 47;X,>=A957A!J!0J=JA6
HANS W. GOTTINGER<br />
1.8 INTERNET OF THINGS ECONOMIC AND INDUSTRIAL DIMENSIONS<br />
! D8/7=8J/6>8J;!e>9=8J;!>X!a9468744!J8A!K5>8>?654!N747J=5.!"#$%&!'()*+!$$),$"-! $$'!<br />
!<br />
D8/7=87/,;6]7!87/2>=]3!Y7[6574!J/!.>?7!8>!?J//7=!/.76=!46`7L!<br />
J=7!=!0J446[7!NdDY!/J8A6/6>87=! J=7! J//J5.7A! />! J! /=J84576[7=! />!<br />
5>??9865J/7! 26/.! J! ;>5J;! 47=[7=3! M.64! ;>5J;! 47=[7=! 90AJ/74!<br />
/.7! 4784>=! >9/09/! AJ/J! 68/>! /.7! D8/7=87/! /.=>95J;!<br />
47=[7=3! M.7! AJ/J! X=>?! D8/7=87/! 64! J557447A! X=>?! J! A74]/>0L!<br />
?>:6;7!0.>87!>=!J8S!A7[657!/.J/!64!5>8875/7A!/>!/.7!D8/7=87/3!<br />
I87! >X! /.7! ?J68! JA[J8/JX! /.64! ]68A! >X! 87/2>=]68;[68?7! J00;6J8574! J8A! A7[6574! 64! /.J/!<br />
27! 5J8! .J[7! 5>8/=>;! >[7=! >9=! A7[6574! /.=>99/! J! :J;!<br />
=7J5.!2.7=7!5>??9865J/6>8!64!0>446:;73!M.7!0>/78/6J;!J=6474!<br />
>8! 87J=;S! 7[7=S! 45J;7! J4! 27! 5>9;A! 6?J8>?>94! 47;X,A=6[68874*L!<br />
4?J=/! .>?74! J8A! 4?J=/! 56/6743! D>M! 5J8! :7! 6?0;7?78/7A! 68!<br />
:9468744!>=!68A94/=6J;!78[6=>8?78/4!:J47A!>8!/.7!=7c96=7?78/!<br />
78J:;7A! /75.865J;;S! :S! ?J5.687,/>,?J5.687! (Q"Q*!<br />
5>??9865J/6>8!(O;0JSA68!["]*3!<br />
M.7=7! J=7! ?J8S! A7[6574! 5>8875/7A! />! /.7! D8/7=87/3! d>=!<br />
7\J?0;7L! />! :! 27! .JA! />! 84!?J89J;;SL!:9/!8>2!27!.J[7!0=68/7=4!/.J/!5J8!:7!<br />
5>8875/7A! 26=7;744;S! /.=>9?7! 87/2>=]3! M.7!<br />
0=68/68:!5J8!:7!A>87!X=>?!J8S2.7=7!J8A!>8!J8S!A7[657!<br />
495.!J4!J!5>?09/7=!>=!J!4?J=/0.>873!D8!/.7!4J?7!2JSL!J;;!/.7!<br />
J00;6J8574!J8A!A7[6574!J/!.>?7!5J8!:7!5>8/=>;;7A!=7?>/7;S3!D/!<br />
64! J5.67[7A! :S! J==J8=]!J8A!5>8875/68!J!A75646>8,<br />
?J]68=]! :75J947! /.7!<br />
A75646>8,?J]688!>8!/.7!4/J/94!>X!/.7!<br />
A7[6574!68!/.7!87/2>=]3!<br />
D8!J8!D>M!78[6=>8?78/L!/.7!5>8/=>;;68??9865J/6>8!<br />
>X! A7[6574! 68[>;[74! ?95.! 4784>=! 87/2>=]4! 26/.! 6809/! J8A!<br />
>9/09/!0>=/43!<br />
W.J/! A64/68M! X=>?! /=JA6/6>8J;! 87/2>=]!<br />
/75.8>;>,X>;A+! (6*! /.7! ^/.68X! A6[7=47!<br />
8J/9=7! 495.! J4! ?J5.687! 0J=/4L! 5>8/=>;! 986/4L! 5>??9865J/6>8!<br />
A7[6574LfL! (66*! /.7S! 78=7! 5>?0;7\!<br />
5>??9865J/6>8! 46?6;J=! J8A! 90! />! /.7! ;7[7;! >X! .9?J8!<br />
5>??9865J/6>83! a3! O=/.9=! [-]L! /.7! E/J8X>=A! 75>8>?64/! J8A!<br />
XJ/.7=! >X! ^685=7J4688A!K5>8>?S!J8A!7\7?0;6X674!6/!;6]7!/.64+!<br />
gf!6X!D_?!A=6[684!O82!>X!/=JXX65!2.7=7!?S!5J=!64!68!<br />
J! 5>8[7=4J/6>8! 26/.! /.7! 5J=4! J=>98A! 6/! /.J/! J=7! 68! J!<br />
5>8[7=4J/6>8! 26/.! 8A! 75>8>?S! 64! 5=7J/68=! 94! ,,,! 4;>2;SL! c967/;SL! J8A!<br />
4/7JA6;S!,,,!J!A6XX7=78/!2>=;Ah3!<br />
CA%D"4E-(*;#1%6'$'"(45%*#%B#+'(#'+%*F%<br />
35,#-$%<br />
D>M! ?7J84! 5>8875/68! /.7! D8/7=87/L! 6/! 5J8!:7! J8!<br />
>:@75/! >=! J! A7[6573! M.7! 5>??9865J/6>8! >X! A7[6574! >8! /.7!<br />
D8/7=87/,;6]7!87/2>=]!5J8!:7!>87!46A7A!>=!/2>!46A7A!68!/7=?4!<br />
>X! 87/2>=]68??9865J/6>8! 5J8! :7! J!<br />
46?0;7\! ]68A! >X! 5>??9865J/6>8! >=! A90;7\! 5>??9865J/6>83!<br />
M.7! D8/7=87/! .J4! :778! 9468X! /75.8>;>?! (5>??7=56J;*&! /.7! 5>??9865J/6>8!<br />
:7/2778! /.7! A7[6574! .J4! 7[>;[7A! J4! D8/7=87/! /75.8>;>07A3!M.7!872! 27:!A7[7;>0?78/!>8!/.7!27:!J=7!<br />
W7:! "! J8A! W7:! -! /75.8>;>;>;>=!A7[6574!<br />
/>! 5>??9865J/7! 26/.! /.7! 27:! A6=75/;SL! ?J]688,?J]6807=!<br />
J;=6/.?!/>!?J]7!J!A75646>8!J8A!J5/!J55>=A68=!684/J857L!5>846A7=!J8!J5/6[7!NdDY!5.60!7c96007A!26/.!<br />
J! =7;7[J8/! 4784>=&! /.64! NdDY! ;J:7;! 949J;;S! ?! /.7! 0>27=7A! 4784>=! 68! /.7! 5.60! J8A! 478A4! /.J/!<br />
AJ/J! />! /.7! 47=[7=! :9/! A>74! 8>/! /J]7! J8S! J5/6>83! V>27[7=L!<br />
2.78!27!??J8A4!68!5>?09/7=!;J8>47! J8!<br />
J00=>0=6J/7!J5/6>8!X=>?!/.7!X!J5/6>843!D8!/.64!5J47L!<br />
27! J=7! /J;]689/! ?J5.687,/>,?J5.687! (Q"Q*!<br />
5>??9865J/6>8! 2.7=7! /.7! ?J5.687! 5>??9865J/74! 26/.68!<br />
/.7?47;[74!J8A!07=X>=?4!J8!J00=>0=6J/7!J5/6>83!%<br />
D8A94/=6J;!A=6[68=574!68!/.7!D>M!5>8/7\/!!<br />
O55>=A68! D>M! A7[7;>0?78/L! D>M! :9468744! 7[>;[74!<br />
/.=>9! /7;7,<br />
>07=J/6>8! J8A! /7;7,0=747857! J4! J! =749;/! >X! /.7! 68/7=J5/6>8!<br />
X>=574! >X! :>/.! ?J=]7/! 09;;! J8A! /75.8>;>//680?78/!=>JA?J0!68A65J/74!/.J/!/.7!0=>;>8/689>94;S! 5>8/=6:9/7! />! /.7!<br />
A7[7;>0?78/! >X! D>ML! 2.6;7! /.7! 5>??7=56J;6`J/6>8! 68! /.7!<br />
?J=]7/!0;J57!64!J8>/.7=!]7S!64497!=7;J/68!/.7!0=>X!<br />
D>M3! M.7=7X>=7L! D>M! :9468744! 64! 0=>?>/7A! :S! 68A94/=6J;!<br />
A=6[68=574!>X!:>/.!/75.8>;>;>8!:76890?78/! >X! =7;J/7A! /75.8>;>0?78/!>X!D>M!/75.8>;>94!0>/78/6J;!/>!<br />
i=?4! 68! D>M! :94687443! d>=! 7\J?0;7L! ?J8S! 6?0>=/J8/!<br />
/75.8>;>9A!5>?09/68;>=! 87/2>=]! /75.8>;>?>/7A! /.7! A7[7;>0?78/!<br />
>X!D>M!:9468744!/>!J!872!;7[7;!(W.6/?>=7!7/!J;3![)]*3!<br />
E75>8AL!?J=]7/!09;;!64!A7i87A!J4!J8!688>[J/6[7!X>=57!:76807A! :S! i=?4! 68! =740>847! />! J8! 6A78/6i7A! J8A1! >=!<br />
0>/78/6J;! ?J=]7/! 877A! 5>8A6/6>8J;! >8! J! =?>94! ?J=]7/! A7?J8A! 68! D>M! :9468744!<br />
0=>[6A74! 980=757A78/7A! ?J=]7/! >00>=/986/674! />! i=?43! d>=!<br />
7\J?0;7L! D>M! J00;65J/6>84! 68! .>?7! J00;6J8574L! .7J;/.5J=7L!<br />
J8A! J9/>?>:6;74! J=7! 09;;7A! :S! 872! ?J=]7/! A7?J8A4! ;6]7!<br />
.>947.>;A!A7?J8A!X>=!.>?7!J9/>?J/6>8L!0J/678/!A7?J8A!X>=!<br />
594/>?6`7A! 47=[657L! J8A! 594/>?7=! A7?J8A! X>=! 68/7;;68J;! M7;75>??9865J/6>8! j86>8! (DMj*! X>=!<br />
684/J857! A7i874! /.7! D8/7=87/! >X! M.68=!/.7!D8X>=?J/6>8!E>567/SL!78J:;688875/688L! 7\64/68;[6807=J:;7! 68X>=?J/6>8!<br />
J8A! 5>??9865J/6>8! /75.8>;>
1.8 INTERNET OF THINGS ECONOMIC AND INDUSTRIAL DIMENSIONS<br />
HANS W. GOTTINGER<br />
$$%! VJ84!W7=87=!R>//68X!M.688>?65!J8A!D8A94/=6J;!Y6?7846>84! !<br />
!<br />
/.7! /.68?7! 5>8875/7A! 68! /.7! D>M3! I/.7=!<br />
A7i86/6>84!X>594!>8!D8/7=87/,=7;J/7A!J4075/4!>X!/.7!D>ML!495.!<br />
J4! D8/7=87/! 0=>/>5>;4! J8A! 87/2>=]! /75.8>;>8!47?J8/65!5.J;;78M!=7;J/68!/.7!<br />
4/>=J=8! >X! ;J=;9?74! >X!<br />
68X>=?J/6>8!(O/`>=6!7/!J;3![%]*3!!<br />
W.6;7!/.7=7!64!8>!986[7=4J;!A7X686/6>8!X>=!/.7!D>ML!/.7!5>=7!<br />
5>8570/! 64! /.J/! 7[7=SAJS! >:@75/4! 5J8! :7! 7c96007A! 26/.!<br />
6A78/6XS682! /.7?! />! 5>??9865J/7! 26/.! >87! J8>/.7=! J8A!<br />
26/.!>/.7=!A7[6574!J8A!47=[6574!>[7=!/.7!D8/7=87/!/>!J5.67[7!<br />
4>?7! 947X9;! >:@75/6[73! M.7! 5>=7! 4688570/4! 98A7=;S68M!J=7!8>/!8723!d>=!S7J=4L!/75.8>;>=! 87/2>=]4! .J[7! :778! 947A! 68! 68A94/=6J;!<br />
J8A! ?J89XJ5/9=688/7\/4! X>=! /=J5]685]3!M.7!6A7J!>X!A6=75/!?J5.687,/>,!<br />
?J5.687!5>??9865J/6>8!64!J;4>!8>/!/.J/!872L!J4!6/!64!:J465!/>!<br />
/.7!6A7J!>X!/.7!D8/7=87/!68!2.65.!5;678/4L!47=[7=4!J8A!=>9/7=4!<br />
5>??9865J/7!26/.!7J5.!>/.7=3!W.J/!/.7!D>M!=70=7478/4!64!J8!<br />
7[>;9/6>8!>X!/.7!947!>X!/.747!7\64/68;>X!<br />
/.7! 89?:7=! J8A! ]68A4! >X! A7[6574! J4! 27;;! J4! /.7!<br />
68/7=5>8875/6>8! >X! 87/2>=]4! >X! /.747! A7[6574! J5=>44! /.7!<br />
D8/7=87/3!d>=!7\J?0;7L!?>4/!A7[6574!59==78/;S!>8!/.7!D8/7=87/!<br />
27=7!>=6X!/.7!D8/7=87/!J8A!.J[7!<br />
68/7574468=J=]! 5J0J:6;6/6743!<br />
M.747!A7[6574!685;9A7!47=[7=4L!A74]/>04! ;J0/>04L!/J:;7/4!J8A!<br />
4?J=/!0.>8743!W.J/!/.7!D>M!0=>0>474!64!/>!J//J5.!/75.8>;>! 7[7=SAJS! A7[6574L! 495.! J4! J9A6>1[6A7>! =7576[7=4L! 4?>]7!<br />
A7/75/>=4L! .>?7! J00;6J8574L! 7/53! J8A! ?J]688;687L!<br />
7[78!6X!/.7S!27=7!8>/!686/6J;;S!A746/.7=!?J@>=!7[>;9/6>8J=S!5.J8?647A!:S!/.7!<br />
D>ML!64!/.7!68/78!>X!87/2>=]4!/.J/!5>8/J68!/.747!A7[6574L!<br />
?J]689=!7\J?0;7L!NdDY!.J4!:778!947A!X>=!S7J=4!/>!/=J5]!0=>A95/4!<br />
/.=>9X!/.7!4900;S!5.J683!V>27[7=L!>857!/.7!<br />
0=>A95/! ;7X/! /.7! 4.7;X! >X! J! =7/J6;! >9/;7/L! /.7! ?J89XJ5/9=7=! _4!<br />
J:6;6/S!/>!/=J5]!/.7!>:@75/!2J4!;>4/3!!<br />
b6]72647L! 5>849?7=4! 27=7! 98J:;7! />! ! /.7!<br />
;6X75S5;7!68X>=?J/6>8!>X!0=>A95/4! /.7S!09=5.J47A3!aS! 9M! 0=>?6474! />! 78J:;7! 0=>A95/!<br />
/=J57J:6;6/S!/.=>99/!/.7!78/6=7!0=>A95/!;6X75S5;73!<br />
M.7! 7?7=;>X!A7[6574!J8A!<br />
/.7=7X>=7! /.7! >=68570/! >X! D>ML! .7857! 64! 7\/78A68!<br />
J?:678/! 68/7;;68>?>94! 5>8/=>;3! M>! AJ/7L! J!<br />
89?:7=!>X!/75.8>;>;[7A!68!D>ML!495.!J4!26=7;744!<br />
4784>=! 87/2>=]4! (WEF4*L! :J=5>A74L! 68/7;;684L! 5;>9A!<br />
5>?09/68=]!A7078A688!/.7!4/J8AJ=A6`J/6>83!<br />
Y7078A688! [J=6>94! /75.8>;>=! /.7!<br />
6?0;7?78/J/6>8L! /.7! A7X686/6>8! >X! /.7! D>M! [J=6743! V>27[7=L!<br />
/.7!X98AJ?78/J;!>X!D>M!6?0;674!/.J/!>:@75/4!68!J8!D>M!5J8!:7!<br />
6A78/6X67A! 986c97;S! 68! /.7! [6=/9J;! =70=7478/J/6>843! W6/.68! J8!<br />
D>ML! J;;! /.68! 7\5.J85744!AJ/J!J55>=A68!0=7A7X687A!45.7?743!<br />
GA%H9'4,F,4%3'45#*0*-,'$%"#1%I$"-'%<br />
O/!/.7!5>=7!>X!/.7!6A7J!>X!/.7!D8/7=87/!>X!M.68/6>8!/.J/!7[7=SAJS! ^/.68A4! 26;;! :7!<br />
7c96007A! 26/.! /=J5]688! 64! X9;;S! J5/9J;6`7AL! ^/.68! 5>8/J68! ?>=7!<br />
4>0.64/65J/7A!0=>574468=]68:@75/4!/>!98A7=4/J8A!/.76=!78[6=>8?78/4L!<br />
68/7=J5/!26/.!07>0;7!J8A!:75>?7!4?J=/3!b6]7!J8S!68X>=?J/6>8!<br />
4S4/7?L! /.7! D>M! 26;;! =7;S! >8! J! 5>?:68J/6>8! >X! .J=A2J=7L!<br />
4>X/2J=7! J8A! J=5.6/75/9=74L! J8A! /.=>9X/2J=7! ?>=7! J8A!<br />
?>=7!O=/6X656J;!D8/7;;68>?>94! J8A! 47=[7! J4! 68/7;;68,?J]68
HANS W. GOTTINGER<br />
1.8 INTERNET OF THINGS ECONOMIC AND INDUSTRIAL DIMENSIONS<br />
! D8/7=8J/6>8J;!e>9=8J;!>X!a9468744!J8A!K5>8>?654!N747J=5.!"#$%&!'()*+!$$),$"-! $$C!<br />
!<br />
E784>=!87/2>=]4!J=7!A7[6574!/.J/!?>86/>=!5.J=J5/7=64/654!>X!<br />
/.7! 78[6=>8?78/! >=! >/.7=! >:@75/4! 495.! J4! /7?07=J/9=7L!<br />
.9?6A6/SL! ?>[7?78/L! J8A! c9J8/6/S3! W.78! ?9;/60;7! 4784>=4!<br />
J=7! 947A! />! J4! J!<br />
26=7;744! 4784>=! 87/2>=]! (WEF*3! W6=7;744! 4784>=! 87/2>=]4!<br />
5>8/J68! /.7! 4784>=4! /.7?47;[74! J8A! ?JS! J;4>! 5>8/J68!<br />
;;75/!AJ/J!X=>?!/.7!4784>=4!J8A!0J44!6/!>8!/>!J!<br />
47=[7=3!!<br />
W.6;7! 4784>=4! ^47847_! /.7! 4/J/7! >X! J8! 78[6=>8?78/! >=!<br />
>:@75/L!J5/9J/>=4!07=X>=?!J5/6>84!/>!JXX75/!/.7!78[6=>8?78/!>=!<br />
>:@75/!68!4>?7!2JS3!O5/9J/>=4!5J8!JXX75/!/.7!78[6=>8?78/!:S!<br />
7?6//6898AL! ;6! 2J[74! >=! 7[78! 4?7;;43! M.747!<br />
5J0J:6;6/674! J=7! >87! 2JS! /.J/! D>M! >:@75/4! 5J8! 5>??9865J/7!<br />
26/.! 07>0;73! O5/9J/>=4! J=7! X=7c978/;S! 947A! 68! 5>?:68J/6>8!<br />
26/.! 4784>=4! />! 0=>A957! 4784>=,J5/9J/>=! 87/2>=]43! I87!<br />
7\J?0;7!>X!/.7!947!>X!J5/9J/>=4!68!495.!J!87/2>=]!2>9;A!:7!<br />
/.7!947!>X!J!4784>=!/>!A7/75/!/.7!0=747857!>X!5J=:>8!?>8>\6A7!<br />
68!J!=>>?!J8A!/.7!947!>X!J8!J5/9J/>=!/>!0=>A957!J!;>9A!8>647!<br />
J;7=/680;7!/>!/.7!A7/75/6>8!>X!/.7!.J=?X9;!?:68J/6>8! >X! 4784>=4! J8A! J5/9J/>=4! 5J8! 78J:;7! >:@75/4! />!<br />
46?9;/J87>94;S! :7! J2J=7! >X! /.76=! 78[6=>8?78/! J8A! 68/7=J5/!<br />
26/.!07>0;7L!:>/.!J;4!>X!/.7!D>M3!<br />
D8!0=J5/65J;!68A94/=6J;!94JM!0=7478/;S!7?:=J574!<br />
NdDY,78J:;7A! 6A78/6X65J/6>8! J8A! /=J5]68;>[6A7! 49XX65678/! =7J;,/6?7!<br />
68X>=?J/6>8! J:>9/! /.68ML! 2.65.! J=7! [7=S! 947X9;! />!<br />
?J89XJ5/9=7=4L!A64/=6:9/>=4L!J8A!=7/J6;7=43!d>=!7\J?0;7L!NdDY!<br />
J00;65J/6>8! 68! 4900;S! 5.J68! ?J8J[7!<br />
68[78/>=S! ?J8J?7! 6A78/6X67A! JA[J8/J=! 5>4/L! 46?0;6X67A! :9468744! 0=>574474L! J8A!<br />
6?0=>[7A!7XX656785S3!<br />
d6[7!S7J=4!JL!6/!2J4!=70>=/7A!/.J/!-!07=578/!>X!Kj!:J47A!<br />
5>?0J8674! J=7! 9468![$#]*!<br />
:9/! 26/.! J! 4/=>82/.! 0>/78/6J;3! D8! NdDY,:J47A!<br />
J00;65J/6>84L! )'! 07=578/! >X! X6=?4! ! X>=! J55744! 5>8/=>;L! "k!<br />
07=578/! X>=! 4900;S! 5.J68L! ")! 07=578/! X>=! X=772JS! />;;4L! "B!<br />
07=578/! X>=! 4759=6/S! 5>8/=>;L! "$! 07=578/! X>=! 0=>A95/! 5>8/=>;L!<br />
J8A!$)!07=578/!X>=!J447/!?J8J;[7! [7=S! A6[7=46X67A! 40756X65J/6>84!<br />
68!/7=?4!>X!5>??9865J/6>8L!5>?09/J/6>8L! ?7?>=SL!J8A!AJ/J!<br />
4/>=J=! /=J84?6446>8! 5J0J56/6743! O8! D>M!<br />
J00;65J/6>8! 5>8464/4! >X! ?J8S! /S074! >X! A7[65743! O;;! /S074! >X!<br />
.J=A2J=7! A7[6574! 4.>9;A! :7! 27;;! >=9=]! J8A! :7! J557446:;7! [6J! J[J6;J:;7! 5>??9865J/6>83!<br />
MS065J;;SL! A7[6574! 5J8! :7! >=??9865J/6>8!09=0>47!>[7=!/.7!D8/7=87/3!<br />
D>M! 5J8! :7! J8! J94! 87/2>=]4L!<br />
495.! J4! WEF4L! 26=7;744! ?74.! 87/2>=]4L! ?>:6;7! 87/2>=]4L!<br />
J8A! W6=7;744! b>5J;;S! O=7J! F7/2>=]! (WbOF*3! M.747!<br />
87/2>=]4!.7;0!/.7!/.688,?J]68?09/J/6>8L! J8A! AJ/J! 7\5.J88L! /.7! =7;6J:;7! 5>??9865J/6>8! :7/2778! 8!26/.!=74075/!<br />
/>! D>M3! M.7! X! =98868?0;65J/7!<br />
>0/6?6`J/6>8! J;=6/.?! ;>5J;;S! :S! 7\0;>6/68=]!<br />
]8>2;7A?09/J/6>8J;! 5>?0;7\6/S! 64! 4.6X/7A! X=>?!<br />
/.68!/.7!0/6?J;!=>9/7!J8A!0J=J?7/7=!<br />
[J;974!X>=!/.7!:/J687A3!M.64!64!X7J46:;7!46857!<br />
/.7!46`7!>X!/.7!?J68!64!68!/.7!>=A7=!>X!J!X72!>X!/784!<br />
68!5>?0J=64>8!26/.!/.7!46`74!>X!/.68??9865J/6>8!<br />
=7c96=7?78/4! [J=S! X=>?! >87! A7[657! /S07! />! J8>/.7=3! M.7!<br />
/.68M! 5J8! .J[7! [7=S! A6XX7=78/! 5J0J:6;6/674! X>=!<br />
5>?09/J/6>8L! ?7?>=SL! 0>27=L! >=! 5>??9865J/6>83! d>=!<br />
684/J857L! J! 57;;9;J=! 0.>87! >=! J! /J:;7/! .J4! ?95.! :7//7=!<br />
5>??9865J/6>8!J8A!5>?09/J/6>8!5J0J:6;6/674!/.J8!J!40756J;,<br />
09=0>47! 7;75/=>865! 0=>A95/! 495.! J4! J! .7J=/! =J/7! ?>86/>=!<br />
2J/5.3!E6?6;J=;SL!/.68X! m9J;6/S! >X! E7=[657! (m>E*L! 68! 0J=/659;J=L! 68! /.7! J4075/4! >X!<br />
A7;JSL! 787=849?0/6>8L! J8A! =7;6J:6;6/S3! d>=! 7\J?0;7L!<br />
?686?6`68=! 5>??9865J/6>815>?09/J/6>8!<br />
09=0>474! 64! J! ?J@>=! 5>84/=J68/! X>=! /.7! :J//7=S,0>27=7A!<br />
A7[6574! 26/.>9/! 7XX65678/! 787=/! 5=6/65J;! X>=! /.7! A7[6574! 26/.! 0>27=!<br />
4900;S!5>8875/6>83!<br />
JA%D('"1+5%*F%8990,4"+,*#%8('"$%<br />
D>M! 64! J8! 68/7=5>8875/7A! 87/2>=]! >X! 4?J=/! >:@75/4! J8A!<br />
A7[65743! I:@75/4! 68! /.7! 2>=;A! >X! /.7! D8/7=87/! >X! M.684/;S! =7X7==7A! />! J4! 4?J=/! >:@75/4L! J=7! J9
1.8 INTERNET OF THINGS ECONOMIC AND INDUSTRIAL DIMENSIONS<br />
HANS W. GOTTINGER<br />
$$k! VJ84!W7=87=!R>//68X!M.688>?65!J8A!D8A94/=6J;!Y6?7846>84! !<br />
!<br />
/.>47!0=>?0/7A!:S!/.7!87/2>=]68X!7[7=SAJS!>:@75/4h!(DMj!<br />
[']*3!<br />
M.7! D>M! 64! J! 5>8[7=X! [J=6>94! [646>84L! 7[>;[688! /.7! 4/J]7.>;A7=4_! X>594! >X! A7[7;>0?78/3!<br />
O/`>=6!7/!J;3![%]!6A78/6i74!/.=77!5;J4474!>X![646>84!26/.68!/.7!<br />
=747J=5.! 5>??986/S! >X! /.7! D>M+! M.68=678/7AL! D8/7=87/,<br />
>=678/7A! J8A! E7?J8/65,>=678/7A3! M.7! M.68=678/7A! [646>8!<br />
X>59474!>8!/.7!M.68X!M.68??9865J/6>84!(FdP*L!W6=7;744!E784>=!<br />
J8A! O5/9J/>=! F7/2>=]4! (WEOF*L! ?65=>0=>5744>=4L!<br />
7?:7AA7A! 4S4/7?4L! -Y! J8A! :J=5>A74! />=678/7A![646>8!64!5>857=87A!26/.!/.7!0=>/>5>;4!<br />
X>=!5>??9865J/6>8!87/2>=]43!E>?7!0=>/>5>;4!J=7!/.7!D8/7=87/!<br />
G=>/>5>;L! DG['! J8A! /.7! W7:! >X! M.68
HANS W. GOTTINGER<br />
1.8 INTERNET OF THINGS ECONOMIC AND INDUSTRIAL DIMENSIONS<br />
! D8/7=8J/6>8J;!e>9=8J;!>X!a9468744!J8A!K5>8>?654!N747J=5.!"#$%&!'()*+!$$),$"-! $"#!<br />
!<br />
78>9!4900>=/!=>:94/!785=S0/6>83!M>!78J:;7!785=S0/6>8!>8!<br />
/.7!D>ML!J;=6/.?4!877A!/>!:7!?JA7!?>=7!7XX65678/!J8A!;744!<br />
787=849?688! 45.7?74!<br />
J=7!877A7A!(pJ8!J8A!W78![""]*3!!<br />
D8! JAA6/6>8! />! 785=S0/6>8L! 6A78/6/S! ?J8J=/J8/! 5>?0>878/! >X! J8S! 4759=6/S! ?>A7;! J8A! 986c97!<br />
6A78/6X67=4!J=7!74478/6J;!/>!D>M!A7[65743!M.747!6A78/6X67=4!?JS!<br />
:7! 947A! />! 74/J:;64.! 07=4>8J;! 6A78/6/674! J/! X68J856J;!<br />
684/6/9/6>84L!6A78/6XS!6;;7/.7=!X985/6>843!!<br />
M.94L!7849=68:@75/4!J=7!2.>!/.7S!4JS!/.7S!J=7!<br />
64!74478/6J;!/>!D>M!4955744!(QJ.J;;7!7/!J;3!["-]&!N>?J8!7/!J;3!<br />
["B]*3!<br />
G=6[J5S!!<br />
O4!?>=7!J8A!?>=7!>:@75/4!:75>?7!/=J57J:;7!/.=>9ML!<br />
/.=7J/4!/>!07=4>8J;!0=6[J5S!:75>?7!?>=7!47=6>943!D8!JAA6/6>8!<br />
/>! 4759=68! ?J]7! 49=7! /.J/! 6/! A>748_/! XJ;;! 68/>! /.7!<br />
2=>8X!AJ/J!>287=4.60!877A!/>!:7!JAA=7447A!<br />
68!>=A7=!/>!7849=7!/.J/!947=4!X77;!5>?X>=/J:;7!0J=/6560J/68M3!!<br />
M.94L!/.7!>287=4.60!>X!AJ/J!5>;;75/!7A!X=>?!4?J=/!>:@75/4!<br />
?94/!:7!5;7J=;S!74/J:;64.7A3!M.7!AJ/J!>287=!?94/!:7!J449=7A!<br />
/.J/! /.7! AJ/J! 26;;! 8>/! :7! 947A! 26/.>9/! .641.7=! 5>8478/L!<br />
0J=/659;J=;S! 2.78! /.7! AJ/J! 26;;! :7! 4.J=7A3! G=6[J5S! 0>;65674!<br />
5J8!:7!>87!J00=>J5.!/>!7849=68X!68X>=?J/6>83!<br />
E?J=/! >:@75/4! J8A! =7JA68M! 5J8! 7J5.! :7!<br />
7c96007A!26/.!0=6[J5S!0>;656743!W.78!/.7!>:@75/!J8A!=7JA7=!<br />
5>?7!68/>!5>8/J5/L!/.7S!5J8!7J5.!5.75]!/.7!>/.7=!_4!0=6[J5S!<br />
0>;65S!X>=!5>?0J/6:6;6/S!:7X>=7!5>??9865J/68?J8!7/!J;3!<br />
["B]*3!!<br />
LA%.4*#*/,4%D'#'F,+$%<br />
D8!J!49=[7S!>8!D>M!J8A!6/4!^75>8>?65!787=
1.8 INTERNET OF THINGS ECONOMIC AND INDUSTRIAL DIMENSIONS<br />
HANS W. GOTTINGER<br />
$"$! VJ84!W7=87=!R>//68X!M.688>?65!J8A!D8A94/=6J;!Y6?7846>84! !<br />
!<br />
5>9;A!27;;!.J[7!0>46/6[7!75>8>?65!6?0J5/4!:9/!5>9;A!27;;!:7!<br />
87X!4759=6/S!:=7J5.74L!5S:7=!06=J5S!J8A!<br />
;>44! >X! 0=6[J5S! 0>/78/6J;;S! 68A95688>?65!<br />
AJ?J9;A! :7! 7[78! >9/! >X! 0=>0>=/6>8! 26/.! J!<br />
986[7=4J;! 45J;J:;7! J8A! 45J;68X! D>M3! O4! 68! /.7!<br />
75>8>?654!/=7J/?78/!>X!J556A78/!;J2!/.64!=7c96=74!J44744?78/!<br />
>X!J!A7;65J/7!:J;J857!>X!/=JA7>XX4!J8A!5>?0784J/6>83!!<br />
E9??J=6`68M! 0=>?6474! ?9;/60;7!<br />
75>8>?65!:787X6/4!495.!J4!!<br />
($*! ;>27=688! 5>4/4! X>=! [J=6>94! 68A94/=674! J8A!<br />
/.7! 78/6=7! 75>8>?SL! 6?0=>[68X! 0=>A95/4L! 47=[6574!<br />
X>=!5>849?7=4L!J8A!c9J;6/S!>X!;6X7!68!78[6=>8?78/J;!0=>/75/6>8!<br />
J8A!787=[646>8L!<br />
("*!78J:;68=?J857!68!J!:=>JA!<br />
47/! >X! [J;97! 8>?65! 2/.! J8A! 27;;! :76899/!5=7J/68:4!68!4>?7!68A94/=6743!<br />
MA%N;+;('%O,('4+,*#$%<br />
E6857! /.7! D>M! .J4! 8>/! S7/! :778! X9;;S! =7J;6`7A! s!<br />
5>8570/9J;;S!J8A!68!0=J5/65J;!6?0;7?78/J/6>84L!6/!?656>94! />! X>=75J4/! /.7! X9/9=7! A6=75/6>84L! /.7! 7\/78/! >X!<br />
68A94/=6J;! :=>JA7868X! D>M3! p7/L! /.7!<br />
?J89XJ5/9=68=! />8! /.7! W>=;A! K5>8>?65! d>=9?_4! (WKd*! "#$'! O889J;!<br />
Q77/684L! J;4>! 27! .J[7! .JA! J! 5>?0=7.7846[7!<br />
4.>25J47! J/! /.7! PKaDM! J8A! /.7! VJ88>[7=! D8A94/=6J;! dJ6=!<br />
("#$'L! "#$%*! 26/.! 5>?0J8674! 495.! J4! E67?784L! OaaL! D8/7;L!<br />
DaQL! V9J276L! tMKL! a>45.L! l9]JL! dJ895L! lJ2J4J]6! 68!<br />
0J=/87=4.60! 26/.! 4>X/2J=7! X6=?4! 4.>268@75/4L! J8A!<br />
7\0>468[7=! $##! 5J474! >X! =7J;,2>=;A! 6?0;7?78/J/6>84! >X!<br />
D8A94/=S!B3#3!d9/9=7![646>84!>X!/.7!D>M!26;;!JXX75/!6/4!59==78/!<br />
A7[7;>0?78/!J8A!?94/!/.7=7X>=7!:7!5>846A7=7A3!!<br />
I87! X9/9=7! [646>8! X>=! /.7! D>M! 64! /.7! W7:! >X! M.68X! M.680>474! /.7! 947! >X! 27:! 4/J8AJ=A4! />! X9;;S!<br />
68/7:@75/4! 68/>! /.7! W>=;A! W6A7! W7:! (WWW*3!<br />
j468;>=!A7[7;>07=4!/>!<br />
:96;A! J00;65J/6>84! 9468:@75/4! J8A! 7\64/68/>5>;4! 5J8! ?>=7! 7J46;S! 78J:;7! /.7! 68/7=>07=J:6;6/S! J8A!<br />
5>??9865J/6>8!>X!A6XX7=78/!A7[65743!O!?J4.90!64!J!W7:!"3#!<br />
5>8570/! 2.7=7! J8! J00;65J/6>8! 9474! AJ/J! J8A! X985/6>8J;6/S!<br />
X=>?!J![J=67/S!>X!27:!=74>9=5743!E>?7!=747J=5.7=4!0=>0>468X! M.68A7;! 49X!J00;65J/6>84!(R968J=A!7/!J;3![-"]*3!<br />
W6/.!/.7!X!/.7!W7:L!/.7!W>M!J8A!D>M!J=7!>8!<br />
/.7! 7AX! 7\07=678568AJSL! 27! J=7! >87!<br />
4/70! 5;>47=! />! /.64! [646>8! A97! />! ;J/74/! JA[J8574! 68! 27:!<br />
47=[6574L! 6A78/6i5J/6>8! /75.8>;>8[7==]4L! 2.65.! ?J]7! 5>??9865J/6>8! 5J0J:6;6/674!<br />
J8A! 0=>57446827=! J[J6;J:;7! 68! 685=7J468;[68! /.7! 4>,<br />
5J;;7A! ^W7:! >X! M.68M*L! J8! 78[6=>8?78/! 2.7=7!<br />
7[7=SAJS!A7[6574!495.!J4!/=JXi5!;6A6/674!J=7!=75>8/=>;;J:;7! [6J! /.7! D8/7=87/3! P7=/J68;SL! 4/J=/68?! J8! D8/7=87/! >X! 87J=;S! >87! /.>94J8A! ?6;;6>8! 5>?09/7=4L!<br />
/.7!W7:!8>2!/9=84!/>!:7!J8!D8/7=87/!>X!87J=;S!$##!:6;;6>8!>X!<br />
/.68=!A7[6574!0=74J?!J8!D>M!/>!J!W>M3!<br />
M.94L!68!J!W>M!0;J/X>=?!/.7!2>=];>JA!64!09/!J/!/.7!7\/=7?7!<br />
;7[7;!J8A!45J;J:6;6/S!64!J!5>?09;4>=S!=7c96=7?78/3!!<br />
M.64! A>74! 8>/! =7X7=! />! J8S! /75.8>;>=! J8S! 87/2>=]!<br />
4/=95/9=7L!:9/! >8;S! />! /.7! 6A7J! >X! 68/7=5>8875/688875/68?09/7=4! 26/.! /.7! D8/7=87/3! j47! >X!<br />
/.7! W7:! J4! /.7! 0;J/X>=?! .>4/684688875/7A!<br />
A7[6574! 5J8! :7! 7\0;J687A! :S! ?9;/60;7! /75.8>;>8>?65!:787i/4L!J!X72!>X!2.65.!685;9A7!A70;>S?78/L!.6X! 4/J8AJ=A6`7A!<br />
5>??9865J/6>8! 0=>/>5>;4! J8A! /.7! 75>4S4/7?! 5=7J/7A! /.J8]4!<br />
/>! W7:! "3#! 0J=JA6
HANS W. GOTTINGER<br />
1.8 INTERNET OF THINGS ECONOMIC AND INDUSTRIAL DIMENSIONS<br />
! D8/7=8J/6>8J;!e>9=8J;!>X!a9468744!J8A!K5>8>?654!N747J=5.!"#$%&!'()*+!$$),$"-! $""!<br />
!<br />
899'#1,R2%H/"(+%!*/'%HE'0'+*#%O'$,-#%<br />
S%8#%B00;$+("+,7'%.R"/90'%%<br />
O! .>?7! :75>?74! 4?J=/! 6X! /.7! J55744! J8A! 5>8/=>;! >X! /.7!<br />
A7[6574!J8A!9/6;6/674!68!/.7!.>?7!4/J=/!/J]688!(27!<br />
J=7! 8>/! /J;]689/! O=/6X656J;! D8/7;;607=! 47/! >X!<br />
684/=95/6>84!64!!/.7?L!J8A!>8;S!6X!QJ5.687!/>!?J5.687!<br />
5>??9865J/6>8!64!.J007868!?J5.687!;7J=868446:;7! 6X! 4784>=! 87/2>=]4! 5>8/=>;! /.747! A7[65743!<br />
M75.8>;>/>5>;4L! 87/2>=]4! ;6]7! NdDYL! FdPL! t6>/.! J8A! [J=6>94! 4784>=4!<br />
(/7?07=J/9=7L! GDNL! 4?>]7L! J8A! .9?6A6/S*! 5J8! ?J]7! J!<br />
A75646>8,?J]682! 6;;94/=J/6[7;S!<br />
7\0;J68!/.7!5>??9865J/6>8!0=>57443!<br />
E?J=/!.>?7!87/2>=]3!<br />
W7!5J8!477!4?J;;!56=596/4!(4784>=4!J8A!4?J=/!56=596/4*!X6\7A!<br />
>8!J;;!/.7!A7[6574!J8A!J00;6J8574L!/.747!4784>=4!J8A!A75646>8,<br />
?J]68??9865J/7! 26/.! J55744! ! J! ;>5J;! =>9/7=! /.J/! 0=>[6A74!<br />
D8/7=87/! 5>8875/6>8! /.=>9!<br />
/.7!D8/7=87/3!<br />
E?J=/!.>?74!.J[7!/.7!X>;;>2688!?J]6827=!KXX65678/!O00;6J8574!<br />
V6!9/09/!<br />
J4! /.7! =7849?7! ?686?J;! 0>27=3!<br />
O00;6J8574! ;6]7! .627=! 4J[68! ?J]7! 4?J=/! 9/6;6`J/6>8! >X! 787=9/!5>?0=>?64688!5>?X>=/!;7[7;43!!<br />
-3!N7?>/7!O55744!<br />
M.7! ?J68! 09=0>47! >X! 4?J=/! .>?74! 64! />! J5.67[7! 5>8/=>;!<br />
>[7=!J00;6J8574!=7?>/7;S3!<br />
Q>4/!>X!/.7!/6?74!27!477L!07>0;7!X>=!426/5.!>XX!/.7!<br />
J00;6J8574!2.6;7!=94.68!2>=]!A97!/>!2.65.!/.7S!.J[7!/>!<br />
0JS! .628! >X!<br />
J00;6J857_4! A97! />! 7\57446[7! 94J! J! 986[7=4J;! 5>8/=>;J=7! J557446:;7! X=>?! J8S2.7=7!<br />
=7?>/7;S!26/.!>9=!4?J=/0.>874!J8A!5>?09/7=4L!/.7!947=!5J8!<br />
426/5.!/.7?!J4!.714.7!2J8/4!J8A!?>86/>=!/.76=!.>?73!!<br />
B3!E759=6/S!!<br />
!<br />
W7!:96;A!>9=!.>?7!/.7!2JS!27!2J8/L!J8A!27!J;4>!2J8/!6/!<br />
/>!4/JS!/.7!4J?7!2JSL!4>!4JX7/S!64!J;4>!/.7!0=6?J=S!5>857=8!68!<br />
J! 4?J=/! .>?73! d>=! 4JX7/S! X=>?! X6=7! 2J/7=! J8A! .60=6J/7! 4784>=4! 4.>9;A! :7! 684/J;;7A! />! J;7=/!<br />
94!68!5J474!>X!X6=7!:=7J]>9/!>=!2J/7=!;7J]J?7!<br />
5J8!J;4>!:7![9;87=J:;7!/>!=>::7=S3!M>!0=7[78/!>9=!.>947!X=>?!<br />
=A!/.7!X>>/J9A3! O;4>! />! =?7A! >8! J8S! 98J9/.>=6`7A!<br />
78/=674! J8A! 4940656>94! :7.J[6>=L! 68X=J=7A! J;J=?! 4S4/7?! (6/!<br />
5>9;A! :7! J! GDN! 4784>=! >=! :9=X!<br />
/.7! ;J/74/! /75.8>;>0;7*L! 4>! /.7! >07=J/6>8! >X!<br />
/.747!J00;6J8574! 4.>9;A!:7!7J46;S!98A7=4/J8AJ:;7!X>=!07>0;7!<br />
>X!J;;!J9:;74.>>/!>8!6/4!47;X3! M.7!68/7=XJ57!>X!5>8/=>;!4.>9;A!<br />
J;4>!:7!J4!46?0;7!J8A!X=678A;S!J4!0>446:;73!<br />
d>=! 6;;94/=J/6[7! 09=0>474L! 27! .J[7! A746A7;! />!<br />
A7?>84/=J/7! D8/7=87/! >X! M.68=! =7?>/7! ?>86/>=688/=>;;68X! J00;6J8574! =7?>/7;S! /.=>9X! /.64! ?>A7;! 64! />!<br />
0=>[6A7!8>80.S465J;!5>8/=>;!>[7=!J00;6J8574!J8A!/>!8! /.7! /7?07=J/9=7! J8A! .9?6A6/S! ;7[7;4! J/! .>?7!<br />
/.=>9?!J8S2.7=7!68!/.7!2>=;A3!M.64!<br />
]68A! >X! J! ?>A7;! .7;04! /.7! 0.S465J;;S! 5.J;;78! 5>8/=>;! J00;6J8574! J4! 6/! ?6=! /.7?! />! ?>[7! />! /.7! 426/5.! 7[7=S! /6?7! /.7S! 2J8/! />!<br />
5>8/=>;! 6/3! Q>=7>[7=L! /.7! 0=>0>47A! ?>A7;! 64! 5>4/! 7XX75/6[7!<br />
J8A!0>27=!7XX65678/3!!<br />
M.7! 5>8570/! >X! D8/7=87/! >X! M.68[7A! :S! 5=7J/6807=! 68/7=XJ57! J8A! A746??9865J/7! 26/.! /.7!<br />
D8/7=87/3!O[J6;J:;7!>A4!J8A!c9J8/6/S!5J8!:7!]8>28!68!J!<br />
468:@75/4! J8A! >A4! J=7! 98A!<br />
>:@75/4!5J8!:7!/=J5]7A!94688570/!:S!/=J5]680=6J/7! 4784>=4! X6\7A! />! /.7?3! O! ?>:6;7!<br />
J00;65J/6>8!5J8!:7!5=7J/7A!/>!5>8/=>;!/.747!J00;6J8574!X=>?!<br />
>9=! 4?J=/! 0.>8743! aS! J==J886/>=7A!X=>?!<br />
J8S2.7=7! :S! ;>;>94!J00;65J/6>84!X>=!<br />
7J4S!J557446:6;6/S3!<br />
6'F'('#4'$%<br />
u$v!<br />
u"v!<br />
u-v!<br />
uBv!<br />
K[J84L! G3! P3! J8A! O8898`6J/JL! Q3! ("#$"*3! D8A94/=6J;! D8/7=87/L!<br />
R787=J;!K;75/=65!(RK*!N70>=/L!F72!p>=]3!<br />
O;0JSA68L! K3! ("#$'*3! QJ5.687! b7J=868
1.8 INTERNET OF THINGS ECONOMIC AND INDUSTRIAL DIMENSIONS<br />
HANS W. GOTTINGER<br />
$"-! VJ84!W7=87=!R>//68X!M.688>?65!J8A!D8A94/=6J;!Y6?7846>84! !<br />
!<br />
u)v!<br />
W.6/?>=7L! O3L! O=6L!b3L!D7=JL!O3!J8A!Q>=J:6/>L!R3!("#$#*3!hM.7!D8/7=87/!>X!<br />
M.68?09/7=!F7/2>=]4L!)B($)*L!"%C%s"C#)3!<br />
d68?JL!D3L!E6?07=;L!K3!J8A!V785.L!R3!("##k*3!gO!@>68/!=>JA?J0!<br />
X>=! E7?J8/65! /75.8>;>X! M.68;>=J8A6L! Y3L! E65J=6L! E3L! Y7! G7;;7X! /.688L! J00;65J/6>84! J8A! =747J=5.!<br />
5.J;;785!F7/2>=]4L!$#!(%*L!$Bk%s$)$'3!<br />
u$Bv! b6L!b3L!J8A!b69L!e3!("#$"*3!gO8!7XX65678/!J8A!X;7\6:;7!27:!<br />
47=[6574,:J47A! ?9;/6A64560;68J=S! A7460/6?64J/6>8!<br />
X=J?72>=]! X>=! 5>?0;7\! 788!ES4/7?4L!'(-*L!-B)s-%$3!<br />
u$)v! E5.JXX7=4L!V3L!l>?868>4L!F3L!GJ;;>/L!Q3L!M=>9447L!a3L!F6;44>8L!<br />
Q3L! J8A! I;6[76=JL! O3! ("#$$*3h! E?J=/! 56/674! J8A! /.7! X9/9=7!<br />
68/7=87/+! M>2J=A4! 5>>07=J/6>8! X=J?72>=]4! X>=! >078!<br />
688>[J/6>8hL! 68! Y>?68X! M.68=! J?:678/!<br />
J4464/7A! ;6[68577A68X! /.7! E7[78/.! D8/7=8J/6>8J;!<br />
P>8X7=7857! >8! D8X>=?J/6>8! M75.8>;>84!<br />
(DMFR*3!<br />
u$Cv! d;n84/=95/68??986,! 5J/6>84!<br />
68!5>?09/7=!J8A!68X>=?J/6>8!4567857!(Z>;3!-"*L!$'s"'3!a7=;68+!<br />
E0=68874L! e3! Y3! ("#$"*3! g7,<br />
a9468744! 4S4/7?4! 68/78+! J! 4S4/7?4! 07=4075/6[7hL!<br />
D8X>=?J/6>8!M75.8>;>84! =747J=5.! (IN*! 68! 47=[657!<br />
68A94/=674+! J! 5>?0=7.7846[7! =7[672hL! ES4/7?4! N747J=5.! J8A!<br />
a7.J[6>=J;!E567857L!-#!(-*L!-##s-)-3!!<br />
u""v! pJ8L!M3!J8A!W78L!m3!("#$$*3!ga96;A68X!M.68/>5>;!:J47A!>8!68X>=?J/6>8!<br />
/75.8>;,! >?09/3! E563! D8/7;;3! ES4/3! K8[6=>83! $#BL!<br />
$B-s$Bk3!!<br />
u"-v! QJ.J;;7L! G3L! aJ:J=L! E3L! G=J4JAL! F3! N3! J8A! G=J4JAL! N3!<br />
("#$#*3hDA78/6/S! ?J8J=]! />2J=A4! D8/7=87/! >X!<br />
M.68M*+!N>JA?J0!J8A!]7S!5.J;;78?J8L! N3L! FJ@7=JL! G3! J8A! b>07`L! e3! ("#$$*3hE759=68X!M.68?09/7=4L!BB(k*L!)$s)C3!<br />
u")v! d;7645.L! K3! ("#$-*3! gW.J/! 64! /.7! D8/7=87/! >X! /.68?!.//0+!1!1!2223!6?3!K/.`3!5.!1!<br />
7A95J/6>8!1!VE$#!1!OjMIDYbOaE,WG,aDtOGG,)-30AX3!!<br />
u"'v! R>//68X! R>[7=8J857L!<br />
F72!p>=]+!I\X>=A!j86[3!G=7443!<br />
u"kv! R>//68X! P6/6743!<br />
M.7! 68/7=87/! >X! K[7=S/.682! 27! ;6[7hL! 68!<br />
E0756J;!D4497!>X!d>=76X!M.68!/.7!W7:!>X!M.689=57,<br />
>=678/7A! J=5.6/75/9=7! J8A! :74/! G=J5/6574hL! 68! j5]7;?J88L! Y3L!<br />
VJ==64>8L! Q3! J8A! Q65.J.7;;74L! d3L! KA43! ("#$#*3! O=5.6/75/68X!M.68
HANS W. GOTTINGER<br />
1.8 INTERNET OF THINGS ECONOMIC AND INDUSTRIAL DIMENSIONS<br />
132
2.<br />
MARKETS<br />
AND<br />
COMPETITION
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
HANS W. GOTTINGER<br />
!"#$%&'&()*+),-(%.&(()/&(&0"#$)1)2*3456)7*48)<br />
9-:3%#0;%*.)
HANS W. GOTTINGER<br />
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
!"##$%&'()*+,*-,*./0123,*4%%"56#$"%)*78%69$:;*"8!4='!B4,C8#!:#,A@#A,8)!I@"AB$8#8,!E4:!+B$,8::8'!5;!#"8!JA4D+#4#+98!'+4#+*=!*8!<br />
+=!,8:$8@#+98!+='A:#,+8:!SI@"8,8,?!.X30?.XX6T)!<br />
!<br />
G*,! "+-"U#8@"! +='A:#,+8:! E+#"! 4! -D*54D!
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
HANS W. GOTTINGER<br />
7+%8&9':!$*!;0:&,'::!A!!<br />
!<br />
M80:! *$+! *&+5:! &,! 8&48G('%8! &,.0:(+&':! %$,(&,0$0:! &,,$9-(&$,! &:! (8'! ,-5'! $*! (8'! 4-5'! -:!<br />
%$5K'(&(&$,!&,!&,,$9-(&$,!.+&9':!(8'!$0(%$5':!&,!4+$S(8!1!S'@*-+'!%8-,4':!-,.!:0+9&9-@A!<br />
!<br />
V&(8$0(! &(! :(+0%(0+-@! 4+$S(8! &,! 5-+W'(:! S$0@.! ,$(! F'! K$::&F@'! S8'(8'+! &,! K+$.0%(!<br />
('%8,$@$4&':1!:'+9&%':!$+!$K'+-(&$,-@!K'+*$+5-,%'A!C,,$9-(&$,!F-:'.!*&+5!4+$S(8!S$0@.!F'!-!<br />
W')!&,.&%-($+!*$+!-!*&+5T:!:0:(-&,-F@'!*0(0+'A!<br />
!<br />
7*('+!+':$@9&,4!(8'!&::0'!8$S!-,.!S8)!@-+4'!*&+5:!-+'!-%(&9'@)!',4-4'.!&,!&,,$9-(&9'!-%(&9&()!-,!<br />
&5K@&%-('.! ,'L(! :('K! %$0@.! F'! 8$S! (8')! %$,(&,0'! ($! F'! &,,$9-(&9'! $+4-,&:-(&$,:A! 2,.'+! (8'!<br />
K+':05K(&$,!(8-(!@-+4'!*&+5:!F'%-5'!T@-+4'T!F'%-0:'!$*!T4+$S&,4!(8+$048!&,,$9-(&$,T!(8'+'!&:!<br />
(8'!X0':(&$,!8$S!(8')!W''K!$,!F'&,4!&,,$9-(&9'!*$+!(8'&+!@&*'(&5'A!<br />
!<br />
M8+$048!:(0.&':!$*!&,,$9-(&$,!&,!&,.0:(+&-@!$+4-,&Y-(&$,:!S'!@'-+,!(8-(!+$048@)!(S$G(8&+.!$*!<br />
&,,$9-(&$,:!-+'!$*!(8'!K+$.0%(!&5K+$9'5',(!()K'!S8&@'!$,'G(8&+.!&:!$*!(8'!K+$%'::!&,,$9-(&$,!<br />
()K'!=%$:(!+'.0%&,4!$+!K+$.0%(&9&()!',8-,%&,4>A!V'!5-W'!0:'!$*!(8&:!$F:'+9-(&$,!F)!-::&4,&,4!<br />
(8'!*&+:(!W&,.!($!K+&9-('!$+4-,&Y-(&$,:!=&,.0:(+&-@!*&+5:>!S8&@'!(8'!:'%$,.!($!-!K0F@&%!:'+9&%'!<br />
%',('+'.!$+4-,&Y-(&$,1!.':&4,'.!-,.!+'40@-('.!-:!:0%8A!<br />
!<br />
!*+,-././*012+345-.26.47!.74-23082/00*93./*02<br />
V'! *$%0:! 8'+'! $,! :$5'! W')! +':0@(:! *+$5! F'8-9&$+-@! '%$,$5&%! 5$.'@:! &,! &,.0:(+&-@!<br />
$+4-,&Y-(&$,=! CZ>! (8-(! @'-9'! *$$(K+&,(:! *$+! 'LK@$+&,4! (8'! &,,$9-(&$,! .+&9'! -,.! %-K-F&@&()! $*!<br />
*&+5:!&,!K-+(&%0@-+!&,.0:(+)!:'((&,4:A!!<br />
!<br />
C,!%8-+-%('+&Y&,4!-,.!K$+(+-)&,4!(8'!&,('+,-@!'%$,$5)!$*!@-+4'!*&+5:!
HANS W. GOTTINGER<br />
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
!"##$%&'()*+,*-,*./0123,*4%%"56#$"%)*78%69$:;*"!#"6!:*7'+#+*79!-*A6,7+7-!896>!47'!9*;6!:"47:6)!<br />
!<br />
B"6! :?499+:4?! $*+7#! *!<br />
*76! C*8?'! *596,A6! ;*,6! +77*A4#+*7! +7! #"*96! +7'89#,+69! #"4#! 4,6! :"4,4:#6,+O6'! 5@! 4! "+-"6,!<br />
'6-,66!*!4?#"*8-"!7*!:4894?+#@!9"*8?'!56!+;$8#6'!#*!#"+9!,6?4#+*79"+$)!<br />
!<br />
N*;6!;*'6?9!:*79+'6,!#"6!,6?4#+*79"+$!56#C667!$,*'8:#!;4,=6#!9#,8:#8,6!D+)6)!:*7:67#,4#+*7E!<br />
47'! +77*A4#+*7)! ! S7! #"6! :*7#6(#! *! #"6! $,*545+?+#@! *!C"67!+;+#4#+*7!+9!<br />
7*#!$*99+5?6>!47!+7:,6496!+7!#"6!78;56,!*
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
HANS W. GOTTINGER<br />
7+%8&9':!$*!;0:&,'::!! .$5&,-(':! (8'! *&+5:R! .'%&:&$,A! Y8',! :0%%'::! &:! :($%8-:(&%1! 8$N'9'+1! (8'! (8+'-(!<br />
*+$5!(8'!+&9-@!&,,$9-(&,4!&:!@'::!-%0('A!C,!(8'!%-:'!$*!.+-:(&%!&,,$9-(&$,:1!(8'!%$5O'(&(&9'!(8+'-(!<br />
&:!(8'!:-5'!*$+!F$(8!*&+5:!-,.!&(!&:!(8'!O+$*&(!&,%',(&9'!=N8&%8!&:!@-+4'+!*$+!%8-@@',4'+:>1!,'Z(!($!<br />
.&9'+:'!,$,G5$,'(-+)!&,%',(&9':1!(8-(!.'('+5&,':!(8'!@'9'@!$*!
HANS W. GOTTINGER<br />
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
!"##$%&'()*+,*-,*./0123,*4%%"56#$"%)*78%69$:;*"?!4!,:?@9#A!4!D+':!;4,+:#8!*99!#":?:!#":*,+:?!?"4,:!#":!=*BB*7!<br />
4+B! *
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
HANS W. GOTTINGER<br />
7+%8&9':!$*!;0:&,'::!
HANS W. GOTTINGER<br />
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
!"##$%&'()*+,*-,*./0123,*4%%"56#$"%)*78%69$:;*"! 94,! -,74#7,!<br />
7=*! #"7! 7:$"48+8! *!4!.P31>!N"4$)!2E)!<br />
!<br />
H,4'+#+*!*
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
HANS W. GOTTINGER<br />
7+%8&9':!$*!;0:&,'::!A!<br />
!<br />
S8':'! %$,:&.'+-(&$,:! :044':(! (8-(! &(! &:! 0:'*0@! ($! .&:(&,40&:8! F'(L'',! .&**'+',(! '%$,$5&%!<br />
',9&+$,5',(:1! -%%$+.&,4! ($! (8'! +'@-(&9'! &5N$+(-,%'! $*! %$5N'(&(&$,! &,! &,,$9-(&$,! -,.!<br />
%$5N'(&(&$,!-(!(8'!N+$.0%(!5-+M'(!@'9'@!(8-(!%-,!F'!'TN'%('.!&,!(8'!5-+M'(A!S8'!@&('+-(0+'!$,!<br />
,'(L$+M!&,.0:(+&':!&:!N-+(&%0@-+@)!0:'*0@!&,!'TN@-&,&,4!L8)!-!5-+M'(!5-)!'T8&F&(!L&,,'+G(-M':G<br />
-@@!N+$N'+(&':A!<br />
!<br />
4,)5"%/2677,8)02<br />
U'(L$+M! '**'%(:! 5-)! F'! -,! &5N$+(-,(! *-%($+! &,! .'('+5&,&,4! (8'! %$5N'(&(&9'! ',9&+$,5',(A! 7!<br />
=N$:&(&9'>! .&+'%(! ,'(L$+M! 'T('+,-@&()! 'T&:(:! L8',! (8'! .'5-,.! *$+! -! :'+9&%'! &,%+'-:':! F)! -,!<br />
'T(+-! 0,&(A! C,! $+.'+! *$+! (8&:! '**'%(! ($! F'! O.&+'%(P! (8'! +'-:$,! *$+! (8'! &,%+'-:'! &,! .'5-,.! 50:(!<br />
%$5'!.&+'%(@)!*+$5!(8'!-..&(&$,-@!%$,:05N(&$,!$*!(8'!:'+9&%'!&,!V0':(&$,1!L&(8$0(!,''.!*$+!-!<br />
:(+-('4&%! +':N$,:'! F)!:0NN@&'+:!&,!(8'!:-5'!$+! +'@-('.! 5-+M'(A! E$+! 'T-5N@'1! (8'! .'5-,.! *$+!<br />
N0F@&%! ('@'N8$,)! :'+9&%':! =&A'A! (8'! .'5-,.! *$+! :0F:%+&N(&$,:! ($! ,'(L$+M:>! 5-)! &,%+'-:'! -:!<br />
5$+'!0:'+:!%$,:05'!(8&:!N+$.0%(!='A4A!',('+!&,($!:0F:%+&N(&$,!-4+''5',(:!L&(8!N0F@&%!,'(L$+M!<br />
$N'+-($+:>!:&5N@)!F'%-0:'!5$+'!0:'+:!-@@$L!-!4+'-('+!,05F'+!$*!N$(',(&-@!%$,,'%(&$,:!($!F'!<br />
5-.'!$*!(8'!,'(L$+MA!<br />
!<br />
;)! %$,(+-:(1! -! =N$:&(&9'>! &,.&+'%(! ,'(L$+M! 'T('+,-@&()! 'T&:(:! L8'+'! (8'! .'5-,.! *$+! -! :'+9&%'!<br />
:0NN@&'.!&,!$,'!+'@'9-,(!5-+M'(!8-:!-!:&4,&*&%-,(!'**'%(!$,!:0NN@)!$*!-,$(8'+!N+$.0%(!=()N&%-@@)!<br />
&,! -! .&**'+',(! +'@'9-,(! 5-+M'(>! :0%8! (8-(! (8&:! &,! (0+,! &,%+'-:':! .'5-,.! &,! (8'! *&+:(! +'@'9-,(!<br />
5-+M'(A!<br />
!<br />
7,!'T-5N@'!$*!&,.&+'%(!,'(L$+M!'**'%(:!&:!(8-(!$*!%$5N0('+!:$*(L-+'!*$+!-!N-+(&%0@-+!$N'+-(&,4!<br />
:):('5!N@-(*$+5A!7,!&,%+'-:'!&,!%$,:05N(&$,!&,!(8'!5-+M'(!*$+!:0NN@)!$*!$N'+-(&,4!:):('5:!&:!<br />
@&M'@)! ($! F+&,4! F','*&(:! ($! (8'! 5-+M'(:! *$+! :0NN@)! $*! %$5N-(&F@'! :$*(L-+'1! F)! 'TN-,.&,4! (8'!<br />
5-+M'(!:0%8!(8-(!-!4+'-('+!9-+&'()!$*!:$*(L-+'!&:!$**'+'.!$+!(8'!0,&(!N+&%'!.'%+'-:':!=:&,%'!*&T'.!<br />
%$:(:!$*!:$*(L-+'!.'9'@$N5',(!%-,!F'!:N+'-.!$9'+!-!4+'-('+!,05F'+!$*!N$(',(&-@!F0)'+:>A!C,!<br />
(0+,1!(8'!F','*&(:!($!(8'!:$*(L-+'!5-+M'(:!+',.'+!(8'!:-5'!$N'+-(&,4!:):('5!5$+'!9-@0-F@'1!<br />
-,.! (80:! &,%+'-:'! .'5-,.! *$+! &(A! ;0(! (8'! '**'%(! &:! &,.&+'%(! F'%-0:'! &(! +'@&':! $,! -! :(+-('4&%!<br />
+':N$,:'!$*!:$*(L-+'!:0NN@&'+:!($!(8'!&,%+'-:'!&,!(8'!%$,:05N(&$,!$*!$N'+-(&,4!:):('5:A!<br />
!<br />
9"%&:"+).12.+;2
HANS W. GOTTINGER<br />
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
!"##$%&'()*+,*-,*./0123,*4%%"56#$"%)*78%69$:;*""6,65!47'!<br />
56##6,! $,*'?:#9! '+9$84:6! *8'6,! *59*86#6! $,*'?:#9! =,*A! #"6! A4,F6#)! O7! A4,F6#9! 9?5R6:#! #*!<br />
;6,#+:48!'+==6,67#+4#+*7@!4!=+,A!B*,!A*,6!4::?,4#68!#"6!$4:6!*=!<br />
#6:"7*8*-+:48!4';47:6!A4I!6(+9#9@!*?#9+'6!>"+:"!<br />
=+,A9!:477*#!9?,;+;6!+7!#"6!A4,F6#C)!N"+9!6:*7*A+:!$,*:699!:47!-?+'6!*7!#"6!74#?,6!*=!A4,F6#!<br />
'+96!<br />
+77*;4#+*7!A4*,96V!!N"+9!:477*#!56!<br />
479>6,6'!4#!#"6!-676,+:!86;68@!-+;67!#"6!:?,,67#!86;68!*=!?7'6,9#47'+7-!*=!#"696!6==6:#9)!W47"+:"! A4
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
HANS W. GOTTINGER<br />
7+%8&9':!$*!;0:&,'::!! ($! -%%'::!<br />
'R('+,-@! :$0+%':! $*! M,$N@'.4'! ='A4A! :%&',(&*&%! M,$N@'.4'>S! 8-9&,4! -%%'::! ($! :$5'! (-%&(!<br />
M,$N@'.4'!N8&%8!&:!,$(!(+-,:*'+-E@'!=L'+8-L:!E'%-0:'!'5E$.&'.!&,!$+4-,&O-(&$,-@!+$0(&,':>S!<br />
-,.!8-9&,4!-%%'::!($!:$5'!:L'%&*&%!%$5L@'5',(-+)!+':$0+%':1!'A4A!:L'%&*&%!&,L0(!*-%($+:A!<br />
!<br />
*+,&)-./"0123'2-<br />
J8'!%-L-E&@&()!$*!-!*&+5!($!&,,$9-('!&,!L-+(&%0@-+!-+'-:!N&@@!E'!-**'%('.!E)!(8'!'R(',(!$*!(-%&(!<br />
M,$N@'.4'! :0LL$+(&,4! &(:! &,,$9-(&9'! -%(&9&()1! -,.! N8'(8'+! (8'! *&+5! 8-:! -%%'::! ($! (8&:!<br />
M,$N@'.4'!='A4A!-%T0&+'.!(8+$048!-!L+$%'::!$*!@'-+,&,4FE)F.$&,4>A!<br />
!<br />
J-%&(!M,$N@'.4'!%-,,$(!E'!%$.&*&'.1!(+-,:*'++'.!$+!&5&(-('.!E)!'R('+,-@!L-+(&':!E0(!&:!+-(8'+!<br />
'5E$.&'.!&,!+$0(&,':!-,.!:M&@@:!$*!(8'!*&+5A!J8'!('+5!U(-%&(!M,$N@'.4'K!&:!0:'.!($!',%-L:0@-('!<br />
(8'! &.'-! $*! M,$N@'.4'! (8-(! &:! *$0,.! ($! E'! 0:'*0@! ='A4A! &,! -,! $+4-,&O-(&$,-@! +$0(&,'! $+! -! :M&@@>!<br />
N&(8$0(! E'&,4! -+(&%0@-('1! -,.! (8'+'*$+'! N&(8$0(! E'&,4! .&+'%(@)! (+-,:*'+-E@'! ='A4A! :$@.! -:!<br />
&,*$+5-(&$,>A!<br />
!<br />
455678*465-859-:;8A!<br />
!<br />
Y'!:8$0@.!-%%'L(!(8-(!(8&:!.&:(&,%(&$,!&:!,$(!-E:$@0('A!/-%8!&,,$9-(&$,1!&,!'9'+)!5-+M'(1!N$0@.!<br />
4','+-@@)!&,9$@9'!E$(8!'@'5',(:!$*!&,9',(&$,!-,.!-.$L(&$,A!D'N!&,,$9-(&9'!&5L+$9'5',(:!N&@@!<br />
E'!4',0&,'@)!0,&T0'!($!(8'!&,,$9-(&,4!*&+5!:&,%'!(8')!N&@@!-@N-):!&,9$@9'!-!.'4+''!$*!&5&(-(&$,!<br />
-,.! -.$L(&$,! *+$5! $E:'+9-(&$,! $*! (8'! N$+@.! $0(:&.'! $*! (8'! *&+5A! 7,)! &,9',(&$,! N&@@! .+-N! ($!<br />
:$5'!'R(',(!$,!L-:(!&,,$9-(&$,:!-,.!M,$N@'.4'A!P$,9'+:'@)1!'9',!N8'+'!-!*&+5!:''5:!($!E'!<br />
.&+'%(@)! -.$L(&,4! &,,$9-(&$,:! -::$%&-('.! N&(8! &,9',(&9'! '**$+(! E)! $(8'+! $+4-,&O-(&$,:1! &(! N&@@!<br />
,''.!($!-.-L(!(8':'!($!&(:!$N,!%8-+-%('+&:(&%:!-,.!%&+%05:(-,%':1!N8&%8!+'T0&+':!-,!'@'5',(!$*!<br />
&,9',(&$,A!/9',!N8'+'!&,,$9-(&$,!&:!@-+4'@)!.+&9',!E)!&,9',(&9'!&5L+$9'5',(:!&,!(8'!:'+9&%':!<br />
L+$9&.'.!E)!:0LL@&'+:1!*&+5:!0:&,4!:0%8!:'+9&%':!5-)!,''.!:$5'!%+'-(&9'!'**$+(!($!&,%$+L$+-('!<br />
(8':'!&,($!(8'&+!L+$.0%(&$,!L+$%'::':!-,.!L+$.0%(:A!J8&:!9&'N!&:!0,.'+@&,'.!E)!%$55',(:!*+$5!<br />
Z'@:$,! =6[[I1! LAGBB>! *$@@$N&,4! -,! 'R(',:&9'! %+$::F%$0,(+)! '5L&+&%-@! :(0.)! $,! &,,$9-(&$,Q!<br />
\](8'!E0@M!$*!(8'!'**$+(!&,,$9-(&$,!,''.:!($!E'!.$,'!E)!(8'!*&+5:!(8'5:'@9':A!Y8&@'!(8')!5-)!<br />
.+-N! $,! $0(:&.'! .'9'@$L5',(:1! :&4,&*&%-,(! &,('+,-@! '**$+(! -,.! :M&@@! &:! ,''.'.! ($! %$5L@'5',(!<br />
-,.!&5L@'5',(!(8':'A^!<br />
Z$,'(8'@'::1!N'!%-,!%$,%'&9'!$*!.&:(&,40&:8&,4!(8'!'R(',(!($!N8&%8!(8'!&,,$9-(&$,!-::$%&-('.!<br />
N&(8!:0LL@)!$*!-!*0(0+'!5-+M'(!&:!-.$L(&9'!+-(8'+!(8-,!&,,$9-(&9'1!'9',!&*!E$(8!'@'5',(:!50:(!<br />
E'!L+':',(!($!:$5'!.'4+''A!<br />
!<br />
!<br />
!"#$%&'()!"!#$%&'()!*$+!#%&',%'!-,.!/.0%-(&$,1!2,&('.!3&,4.$5! 66!<br />
145
HANS W. GOTTINGER<br />
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
!"##$%&'()*+,*-,*./0123,*4%%"56#$"%)*78%69$:;*"!4!?@,#";,!'+!
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
HANS W. GOTTINGER<br />
8+%9&:';!$*!(9'!M+&5-+)!5-+K'(?B!<br />
!<br />
8! %$55$,A)! %&('.! 'O-5MA'! &;! (9-(! $*! M+&,('+;! -,.! %$5M-(&FA'! &,KG+'*&AA! %-+(+&.4';1! L9'+'!<br />
.'5-,.! *$+! (9'! A-(('+! &;! .'+&:'.! ',(&+'A)! *+$5! %$,;05M(&$,! $*! (9'! *$+5'+B! D,! (9&;! %-;'1! (9'!<br />
;%$M'! *$+! %$5M'(&(&$,! &,! &,,$:-(&$,! .&+'%('.! ($! +'*&AA! %-+(+&.4';! L&AA! F'! -**'%('.! F)! L9'(9'+!<br />
%-+(+&.4'!5-,0*-%(0+'+;!%-,!&,('+-%(!;0%%';;*0AA)!L&(9!(9'!M+&,('+;!5-+K'(B!<br />
!<br />
8! A';;! $F:&$0;! 'O-5MA'! &;! 5-+K'(;! %$,%'+,'.! L&(9! -.:'+(&;&,4! >(9'! ;0MMA)! $*! -.:'+(&;&,4!<br />
;M-%'?B!D,!5$;(!%-;';1!(9'!'O&;(',%'!$*!(9'!-.:'+(&;&,4!5-+K'(!&;!.'M',.',(!$,!%$,;05M(&$,!&,!<br />
-!+'A-('.!%$,(',(!5-+K'(!>F'!(9-(!(9'!;0MMA)!$*!('A':&;&$,!;'+:&%';1!(9'!;0MMA)!$*!,'L;M-M'+;1!<br />
5-4-P&,';1!$+!(9'!;0MMA)!$*!%$,(',(!$,!(9'!D,('+,'(?B!8,-A);&;!$*!%$5M'(&(&$,!&,!(9'!&,,$:-(&:'!<br />
-%(&:&()!+'A-('.!($!',(+)!&,!(9'!-.:'+(&;&,4!5-+K'(;!50;(!M-)!-((',(&$,!($!(9'!,''.!*$+!(9'!*&+5!<br />
($! ;&50A(-,'$0;A)! $M'+-('! ;0%%';;*0AA)! &,! (9'! %$,(',(! 5-+K'(;B! #$! (9'! +'A-(&$,;9&M! F'(L'',!<br />
-.:'+(&;&,4!+'A':-,(!5-+K'(;!-,.!%$,(',(!+'A':-,(!5-+K'(;!*-AA;!0,.'+!(9'!%-('4$+)!$*!-.N0,%(!<br />
5-+K'(;B!<br />
!<br />
Q*!M-+(&%0A-+!&5M$+(-,%'!&;!L9'+'!-!%$5M'(&(&:'!M+&5-+)!5-+K'(!&;!-;;$%&-('.!L&(9!-,!-.N0,%(!<br />
5-+K'(!&,!L9&%9!(9'+'!&;!5$,$M$A)B!R90;!*&+5;!%$5M'('!&,!-!F+$-.!M+&5-+)!5-+K'(!-,.!'-%9!<br />
-%9&':';!5$,$M$A)!;0MMA)!&,!-!,-++$L!-.N0,%(!5-+K'(!>M+&,('+;!-,.!+'*&AA!%-+(+&.4';?B!!D,!;0%9!<br />
%-;';!L'!5-)!'OM'%(!*&+5;!($!0;'!-,)!5-+4&,-A!&,%+'-;'!&,!M+$*&(;!+'-A&P'.!&,!(9'!;'%$,.-+)!<br />
5-+K'(! ($! *0,.! -! .'%+'-;'! &,! M+&%'! &,! (9'! %$5M'(&(&:'! M+&5-+)! 5-+K'(! -,.! (9'+'F)! &5M+$:'!<br />
(9'&+! +'A-(&:'! M$;&(&$,! -4-&,;(! %$5M'(&($+;B! S$+'$:'+1! L'! L$0A.! 'OM'%(! (9'+'! ($! F'! %9'%K;!<br />
-4-&,;(!%'+(-&,!()M';!$*!-,(&G%$5M'(&(&:'!F'9-:&$+!&,!(9'!5$,$M$A)!A&,K'.!5-+K'(;!.0'!($!(9'!<br />
(+-,;*'+!$*!%$5M'(&(&:'!'**'%(;!*+$5!(9'!M+&5-+)!5-+K'(B!<br />
!<br />
R90;! &,! -.N0,%(! 5-+K'(! %-;';1! (9'! '**'%(! 5-)! F'! ;0%9! (9-(! &,,$:-(&:'! -%(&:&()! +'A-('.! ($! (9'!<br />
;'%$,.-+)! 5-+K'(;! L&AA! .'M',.! $,! (9'! -F&A&()! $*! (9'! *&+5! ($! ',('+! (9'! M+&5-+)! 5-+K'(!<br />
;0%%';;*0AA)T! $+! ! (9'! -F&A&()! $*! (9'! *&+5! ($! U4-&,! -%%';;V! ($! (9'! ,'%';;-+)! A&,K;! F'(L'',! (9'!<br />
M+&5-+)!-,.!!A&,K'.!5-+K'(;B!<br />
!<br />
.6)7"%804996:);0<br />
R9'+'! ;''5! ($! F'! %A'-+! ;&5&A-+&(&';! F'(L'',! -.N0,%(! 5-+K'(;! -,.! ,'(L$+K! '**'%(;1! F0(! (9'!<br />
*$+5'+!&;!.'*&,'.!-F$:'!-;!L9'+'!%$,;05M(&$,!$*!-!M+$.0%(!&,!$,'!5-+K'(!%+'-(';!.'5-,.!&,!<br />
-,$(9'+1! L9&A'! ,'(L$+K! '**'%(;! -+'! 5$+'! +'A-('.! ($! 9$L! %$,;05M(&$,! $*! -! M+$.0%(! &,! $,'!<br />
5-+K'(!&,%+'-;';!.'5-,.!*$+!'&(9'+!>&?!(9-(!M+$.0%(!$+!>&&?!-!M+$.0%(!&,!-!.&**'+',(!5-+K'(!(9-(!<br />
&,! (0+,! 5-)! A'-.! ($! &,%+'-;';! &,! .'5-,.! *$+! (9'! $+&4&,-A! M+$.0%(! >;''! W$((&,4'+! >HII7?! *$+!<br />
.&;%0;;&$,!$*!,'(L$+K!'**'%(;?B!<br />
84-&,!,'(L$+K!'**'%(;!9-:'!&5M$+(-,(!&5MA&%-(&$,;!*$+!(9'!%-M-F&A&()!$*!*&+5;!($!;0MMA)!*0(0+'!<br />
5-+K'(;! -;;$%&-('.! L&(9! &,,$:-(&$,B! #0MMA)! ($! (9'! *0(0+'! 5-+K'(! L&AA! +'X0&+'! -! ;(+-('4)! *$+!<br />
&,('+-%(&,4!L&(9!(9'!,'(L$+K!'**'%(;!&,!;0%9!-!L-)!-;!($!5-K'!(9'!',(+)!;0%%';;*0A!-,.!;0;(-&,!-!<br />
M$;&(&$,! &,! (9'! 5-+K'(B! D,! M-+(&%0A-+1! ,'(L$+K! '**'%(;! %-,! 4','+-('! &5M$+(-,(! F','*&(;! *$+!<br />
&,%05F',(! ('%9,$A$4&';! -,.! (90;! %+'-('! &,('+G('5M$+-A! A&,K;! F'(L'',! %0++',(! -,.! *0(0+'!<br />
5-+K'(;1!-;!.&;%0;;'.!&,!(9'!*$AA$L&,4!;'%(&$,B!<br />
!<br />
!<br />
!"#$%&'()!"!#$%&'()!*$+!#%&',%'!-,.!/.0%-(&$,1!2,&('.!3&,4.$5! 67!<br />
147
HANS W. GOTTINGER<br />
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
!"##$%&'()*+,*-,*./0123,*4%%"56#$"%)*78%69$:;*"
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
HANS W. GOTTINGER<br />
8+%9&:';!$*!$+!&(;!.';%',.-,(?!&,!(9'!*0(0+'!5-+O'(B!<br />
!<br />
D,!(9'!%-;'!$*!&,.&+'%(!,'(L$+O!'**'%(;1!(9'!+'A':-,(!A&,O!F'(L'',!5-+O'(;!&;!&,.&+'%(!F'(L'',!<br />
(9'! %0++',(! +'A':-,(! 5-+O'(! -,.! (9'! *0(0+'! 5-+O'(1! F'%-0;'! &(! 'P&;(;! (9+$049! -! 9$+&S$,(-AA)!<br />
+'A-('.!5-+O'(B!<br />
!<br />
#0%9!+'A-(&$,;9&M;!-+'!(9'+'*$+'!&5M$+(-,(!&,!-!A-('+!;(-4'1!-,.!.&;%0;;'.!*0+(9'+!F'A$LB!K9'!<br />
*$AA$L&,4! (-FA'! ;044';(;! M$;;&F&A&(&';! *$+! &,('+G('5M$+-A! A&,O;! F'(L'',! 5-+O'(;! (9-(! 5-)! M0(!<br />
*&+5;!L9&%9!%0++',(A)!;0MMA)!-!+'A-('.!%0++',(!5-+O'(!-(!-!;&4,&*&%-,(!%$5M'(&(&:'!-.:-,(-4'!&,!<br />
(9'!%-M-F&A&()!($!;0MMA)!(9'!*0(0+'!5-+O'(!(9-,!*&+5;!(9-(!.$!,$(B!<br />
!<br />
*+,+%+-.+/0<br />
849&$,1!TB!-,.!=BU&**&(9!>HII7?!R$5M'(&(&$,!-,.!U+$L(91!R-5F+&.4'1!Q-B!QDK!T+';;!<br />
8++$L1!3BVB!>6WJH?1X/%$,$5&%!N'A*-+'!-,.!(9'!8AA$%-(&$,!$*!=';$0+%';!*$+!D,:',(&$,X1!&,!(9'!=-('!-,.!Y&+'%(&$,!$*!<br />
D,:',(&:'!8%(&:&()Z!/%$,$5&%!-,.!#$%&-A!E-%($+;1!T+&,%'($,Z!T+&,%'($,!2,&:B!T+';;1!JIWGJH7!<br />
>6W[\?1!=&:-A+)!&,!6W[W?1_/5M&+&%-A!#(0.&';!$*!D,,$:-(&$,!-,.!Q-+O'(!#(+0%(0+'_1!&,!`-,.F$$O!$*!<br />
D,.0;(+&-A!a+4-,&S-(&$,1!@$AB!DD1!=B!#%95-A',;''!-,.!=BYBN&AA&41!'.;1!85;('+.-5Z]$+(9!`$AA-,.!<br />
Y$;&1!RB!>6W[[?1!_#$0+%';1!T+$%'.0+';!-,.!Q-%+$'%$,$5&%!/**'%(;!$*!D,,$:-(&$,X1!<br />
V$0+,-A!$*!/%$,$5&%!^&('+-(0+'!HJ>b?1!66HIG66\6!<br />
/-;A')1!YB!-,.!VB3A'&,F'+4!>HI6I?1!]'(L$+O;1!R+$L.;!-,.!Q-+O'(;1!R-5F+&.4'Z!R-5F+&.4'!2,&:B!T+';;!<br />
/:-,;1!YB1!`-4&01!8B!-,.!=B!#%95-A',;''!>HIIJ?1!D,:&;&FA'!/,4&,';1!`$L!#$*(L-+'!TA-(*$+5;!Y+&:'!D,,$:-(&$,!-,.!<br />
K+-,;*$+5!D,.0;(+&';1!R-5F+&.4'1!Q-B!QDK!T+';;!<br />
!<br />
!"#$%&'()!"!#$%&'()!*$+!#%&',%'!-,.!/.0%-(&$,1!2,&('.!3&,4.$5! 67!<br />
149
HANS W. GOTTINGER<br />
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
!"##$%&'()*+,*-,*./0123,*4%%"56#$"%)*78%69$:;*"**'!B.C32D:!7*E'9':!F$+?'E9':!4?'!;G#+E4#9':!A#"!+?!#"9!=E*54E!A!M?'G8#,+4E!^,-4?+_4#+*?:!;4,W9#!7*,
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
HANS W. GOTTINGER<br />
!"#$%&'()*%)+,&*$-)+&*'%&'().'('$/&0)1,2/%$-)3)4,-567)8,59)<br />
:2;-*&$%9K*!L:$E3>) M#:>! 78A$9:! G:9K>8&8D%:4*! N>>8O3;%8>*! P:;Q8$R! @98>8E%94*! S%9:>4%>D*! NG!<br />
N>?A4;$?!T:O:&8#E:>;!UHVTW!%4!3!9$A9%3&!#K:>8E:>8>!X8;K!=$8E!;K:!#8%>;!8=!O%:Q!8=!;K:!<br />
%>?%O%?A3&!=%$E!3>?!;K:!:98>8E:=%;4!;8!3!=%$E!=$8E!<br />
%>>8O3;%>D!3$:!&%R:&;!=$8E!;K:!489%3&!X:>:=%;4'!N>!;K:!3X4:>9:!8=!3>%4E!<br />
#$:O:>;%>D!%;*!;K:!X:>:=%;4!;8!3>!%>>8O3;%>D!=%$E!3$:!&%R:&;$D!=%$E4'!N>!4A9K!3!49:>3$%8*!=%$E4!3$:!A>&%R:&>8O3;:'!GKA4!3998$?%>D!;8!<br />
98>O:>;%8>3&! ;K%>R%>D*! =%$E4! >::?! ;8! K3O:! 48E:! 48$;! 8=! $:Q3$?! =8$! %>>8O3;%>D'! N>;:&&:9;A3&!<br />
#$8#:$;;4!3>?!98#43;%8>'!"!X%D!#8$;%8>!8=!;K:!<br />
HVT!&%;:$3;A$:!K34!=89A4:?!8>!;K:!8#;%E3&!#3;:>;4[!?A$3;%8>!3>?!X$:3?;K!3>?!;K:!%>9:>;%O:!8=!<br />
=%$E4!;8!%>>8O3;:'!<br />
!<br />
Y8Q:O:$*!3!?%==:$:>;!;$:>?!K34!:E:$D:?!;K:4:!?3D!#$8&%=:$3;%8>!<br />
8=! K%\;:9K! U>:;Q8$RW! %>?A4;$%:4'! N>4;:3?! 8=! ;$D! ;8! D:;! :]9&A4%O:! 8Q>:$4K%#! $%DK;4*! 3>!<br />
%>9$:34%>D!>AEX:$!8=!=%$E4!3$:!E3R%>D!;K:%$!;:9K>8&8D8&8D8!&8>D:$!#$8#$%:;3$!48A$9:_!UI3>:*!@98>8E%4;*!(++/W'!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!<br />
!<br />
N>! ;K:! 48\93&&:?! X$8Q4:$! Q3$*! %>! ;K:! ,``+4*! Q:! K3O:! Q%;>:44:?! %>;:>4:! 98E#:;%;%8>! %>! ;K:!<br />
E3$R:;! =8$! %>;:$>:;! X$8Q4:$4! X:;Q::>! J%9$848=;! 3>?! P:;493#:'! P:;493#:! K3?! 3! E358$! K:3?!<br />
4;3$;!8>!J%9$848=;*!98>;$8&&%>D!`+a!8=!;K:!X$8Q4:$!E3$R:;!XD!%>!;K:!E3$R:;'!b%;K!;K:!:>;$D3D:?!%>!3!$39:!;8!<br />
K3O:!;K:!X:4;!3O3%&3X&:!#$8?A9;'!!<br />
!<br />
L%O:>!;K:!%>;:>4:!98E#:;%;%8>!X:;Q::>!;K:!;Q8!=%$E4*!X?!8=!,``-!J%9$848=;!Q34!#$%9%>D!<br />
;K:!N>;:$>:;!@]#&8$:$!=$::'!N>!98>;$34;*!P:;493#:!Q34!9K3$D%>D!98$#8$3;%8>4!&%9:>4%>D!=::4!=8$!<br />
!<br />
!<br />
!<br />
@,AB/*C09:!3>?!@?A93;%8>*!B>%;:?!C%>D?8E!<br />
!<br />
,F/!<br />
151
HANS W. GOTTINGER<br />
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
!"#$%&'()*%)+,&*$-)+&*'%&'().'('$/&0)1,2/%$-)3!++.14)<br />
5,-678)9((2'):)!;/*-?:)<br />
)<br />
2(*%G) A0'*/) K/,L('/6) MB) A0') '%") ,C) ?NN:8) O*&/,(,CA) 0$") (A,-'%) $) -$/G') &02%P! ,C) Q'A(&$;'R()<br />
H$/P'A)(0$/'6)Q'A(&$;')'#'%A2$--B)C,--,L'")(2*A)$%")(A$/A'")G*#*%G)$L$B)*A()K/,L('/)C/''6)<br />
)<br />
S0') 'TA'%"'") K$AA-') K'AL''%) O*&/,(,CA) $%") Q'A(&$;') 0$") *A() A,--) ,%) A0') ;/,C*A() ,C) K,A0)<br />
&,H;$%*'(6)9%)?NNU)Q'A(&$;')&$H')2;)L*A0)$)%'L)(A/$A'GB)$%")"'&*"'")A,)/'-'$(')*A()(,2/&')<br />
&,"'8)A0')$&A2$-)-*%'),C);/,G/$HH*%G)-$%G2$G'8)C,/)A0')Q'A(&$;')V,HH2%*&$A,/6)S0*()$--,L'")<br />
2('/()$%")"'#'-,;'/()A,)-,,P)*%(*"')A0')L,/P*%G(),C)A0')K/,L('/8)A,)H,"*CB)A0')(,CAL$/')$%")<br />
'#'%)A,)/'"*(A/*K2A')A0')%'L)#'/(*,%)2%"'/)A0'*/),L%)K/$%")%$H'8);/,#*"'")A0$A)A0')H,"*C*'")<br />
(,2/&')&,"')L$()$-(,)C/''-B)$#$*-$K-'6)S0')L0,-')*"'$)*()A,)A2/%)A0')'%A*/')*%A'/%'A)&,HH2%*AB)<br />
*%A,)$)#$(A)/'('$/&0)"*#*(*,%)C,/)Q'A(&$;')K/,L('/6))<br />
)<br />
S0') A'/H) W,;'%) (,2/&') (,CAL$/'X) 0$() K''%) L*"'-B) 2('") *%) A0') ;,;2-$/) $%") ;/,C'((*,%$-)<br />
-*A'/$A2/') 35$/*$%) $%") +0$;*/,8) =>>YZ) ['/%'/) $%") S*/,-'8) =>>746) 9%(A'$") ,C) P'';*%G) A0'*/)<br />
A'&0%,-,GB) ;/,;/*'A$/B8) A0') C*/H() L*--) "*(A/*K2A') *A) C/''-B6) 9A) *() A0*() ;0'%,H'%,%) A0$A) A0*()<br />
&0$;A'/) $AA'H;A() A,) 'T;-,/'6) \') L*(0) A,) (A2"B) A0') "'&*(*,%() ,C) C*/H() L0'A0'/) A,) P'';) A0'*/)<br />
A'&0%,-,G*'();/,;/*'A$/B),/)%,A6)<br />
)<br />
D#'%)A0,2G0)A0')L0,-')2%,/A0,",T),;'%)$;;/,$&0)H$B)(''H)&,2%A'/*%A2*A*#'8)Q'A(&$;')L$()<br />
%,A)A0'),%-B),%')L0,)'H;-,B'")*A6)!;$&0'8)$);/,G/$H)C,/)('/#*%G)L,/-")L*"')L'K)(*A'(8)$%")<br />
+'%"H$*-8) $) ;/,G/$H) A0$A) /,2A'() $%") "'-*#'/() *%A'/%'A) '-'&A/,%*&) H$*-8) $/') 'T$H;-'() ,C) C/'')<br />
,;'%)(,2/&');/,G/$H()A0$A)",H*%$A')A0')H$/P'A6)];'%)(,2/&')$;;/,$&0'()0$#')K''%)'T;$%"'")<br />
A,) A0') K*,A'&0%,-,GB) $%") 0'$-A0) &$/') *%"2(A/*'(8) D&,%,H*(A) 3=>>7$46) [*%2T8) $%) *%&/'$(*%G-B)<br />
;,;2-$/),;'/$A*%G)(B(A'H)&/'$A'")*%)?NN?)*()$%,A0'/)&-$((*&)'T$H;-'),C)(2&&'((C2-),;'%)(,2/&')<br />
(,CAL$/'6) O$%B) ,C) A0') ;/,G/$HH'/() $%") (,CAL$/') "'(*G%'/() $"#,&$A*%G) ]++) H$B) (0$/') $)<br />
2A,;*$%)#*(*,%),C)(,CAL$/')"'#'-,;H'%A8),/)A0'B)H$B)(*H;-B)L$%A)A,);/,#')A0'H('-#'()A,)K')<br />
K'AA'/) A0$%) (,CAL$/') G*$%A) O*&/,(,CA6) ^,L'#'/8) A0') L0,-') *"'$) ,C) ]++) H$B) %,A) K') (,) $%A*<<br />
&$;*A$-*(A*&) $() *A) (''H(6) 9A) *() 0$/") A,) K'-*'#') A0$A) ;/,C*A
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
HANS W. GOTTINGER<br />
!"##$%&'()*+,*-./012,*3'#4"(5*67"%"8$'9*:"(*;
HANS W. GOTTINGER<br />
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
!"#$%&'()*%)+,&*$-)+&*'%&'().'('$/&0)1,2/%$-)3!++.14)<br />
5,-678)9((2'):)!;/*-?:)<br />
)<br />
&,%(2H'/) J$('6) K0$A) *() J'&$2(') A0') ",L%(A/'$H) (,CAL$/') &,H;$%*'() A02() &$%) A$;) *%A,) A0*()<br />
-$/G'/)%'AL,/M),C)&2(A,H'/(6)K0*()*H;/,#'()A0')N2$-*AB),C)A0')O+6)K0*()*()A0')*%"*/'&A)'CC'&A),C)<br />
%'AL,/M),%)&,%(2H'/(P)2A*-*A*'(6)<br />
)<br />
O;'%)(,2/&')(B(A'H)$;;/,$&0'()'%A$*-)(,H')A/$%(*A*,%$-);/$&A*&$J-')"*($"#$%A$G'()$G$*%(A)A0')<br />
;/'#$*-*%G)(A$%"$/"6)<br />
)<br />
Q'&$2('),C)J2*-"*%G)2;)$)%'AL,/M)A0'/')L*--)J')C'L'/)$;;-*&$A*,%()L0*&0)$/*(')$()/,$"J-,&M()<br />
A,L$/"() C2/A0'/) "*(('H*%$A*,%8) C,/) 'R$H;-'8) $() *%) A0') &$(') ,C) S*%2R8) ,%-B) $) (*%G-') "*G*A)<br />
;'/&'%A$G') L,/M() C,/) &,H;2A'/) G$H'() 3D&,%,H*(A8) =>>7J46) 9%&,H;$A*J*-*AB) *((2'() L*A0) A0')<br />
",H*%$%A),;'/$A*%G)(B(A'H)$/')$-(,),%'),C)A0')J*GG'(A),J(A$&-'()C,/)C2/A0'/)"*(('H*%$A*,%8)*%)<br />
;$/A*&2-$/8) C,/) -$/G') &,/;,/$A*,%() L0,) L,2-") C$&') (*G%*C*&$%A) (L*A&0*%G) &,(A() L*A0) ;,((*J-')<br />
,;'/$A*,%$-)"*(/2;A*,%(6)O%)A0')J'%'C*A)(*"'8)0,L'#'/8),;'%)(,2/&')(B(A'H()H$B)*%"2&').TU)*%)<br />
&,H;-'H'%A$/B);/,"2&A()$%")('&A,/()J'&$2('),C)-,L'/)-*&'%(')C''(8)A02()-,L'/)&,(A(),C).TU8)<br />
A0'')$/')H,/')*%&'%A*#'()A,)'%-$/G')$%")'R;$%");/,"2&A)"'#'-,;H'%A)$%")$;;-*&$A*,%()J'B,%")<br />
&,/') *%%,#$A*,%(6) Q'&$2(') ,C) *A() &,%A*%2,2(-B) ,%G,*%G) &0$/$&A'/) *A) *() H,/') -*M'-B) A,) J')<br />
*%&/'H'%A$-) A0$%) ('H*%$-) L0*&0) ;,('() N2'(A*,%() ,C) A0'*/) ,L%) $() A,) L0'A0'/) ,;'%) (,2/&')<br />
(B(A'H()&$%)G'%2*%'-B)C,(A'/)*%%,#$A*,%6)V,/)"'"*&$A'")$;;-*&$A*,%()$()*%*A*$--B)J2*-A)2;)C/,H)<br />
(&/$A&0)(2&0)$()('$/&0)A0/,2G0)W,,G-')*A)H$B)B*'-")(,H')J*G)$"#$%A$G'()8)%,A)A0')-'$(A)J'&$2(')<br />
,C)&,H;$/$A*#'-B)-,L)*H;-'H'%A$A*,%)$%"),;'/$A*,%$-)&,(A(6)<br />
)<br />
K0*();$;'/)*(),/G$%*X'")$()C,--,L(Y)K0')%'RA)('&A*,%)-,,M()$A)A0')-*A'/$A2/')/'-$A'")A,)A0*()L,/M6)<br />
9%)+'&A*,%)Z6)L');/'('%A)A0')H,"'-6)9%)+'&A*,%)78)L')-,,M)$A)A0')'N2*-*J/*2H)/'(2-A()$%")G*#')<br />
*%A'/;/'A$A*,%(6)+'&A*,%)[)*%"*&$A'()'RA'%(*,%()A,),2/)H,"'-)$%")"/$L()&,%&-2(*,%(),%)C2/A0'/)<br />
2('6)<br />
*<br />
+,)-.)&.'/*0&1-.2&.'*,.3*4#-.*5"6%1-*7-1(."8"'&-2*<br />
K0') *((2'() &,#'/'") *%) A0*() ;$;'/) /'-$A') A,) L,/M) ,%) %'AL,/M) 'RA'/%$-*A*'(8) .TU8) '%A/B)<br />
"'A'//'%&')$%")-*&'%(*%G6)<br />
)<br />
.TU) *() $%) 'RA'%(*#'-B) /'('$/&0'") $/'$) *%) *%"2(A/*$-) ,/G$%*X$A*,%) (A$/A*%G) L*A0) !//,LP()<br />
;*,%''/*%G)$/A*&-')L0'/')0')$(M'"Y)\]0$A)*()A0')G$*%)C/,H)*%%,#$A*,%)A,)$)C*/H)A0$A)*()A0'),%-B)<br />
,%')A,)2%"'/A$M').TU8)G*#'%)A0$A)*A()*%%,#$A*,%)*();/,A'&A'")JB)$);$A'%A),C)2%-*H*A'")"2/$A*,%^_)<br />
3!//,L8)?`I=46))+*%&')A0'%)A0'/')0$()J''%)$)(;$A'),C)/'('$/&0),%).TU)&,#'/*%G)*((2'()(2&0)$()A0')<br />
*%&'%A*#'() A,) *%%,#$A'8) ;$A'%A) /$&'(8) L'-C$/') *H;-*&$A*,%() ,C) .TU8) &0,*&') $%") $",;A*,%) ,C)<br />
A'&0%,-,G*'(6)a,L'#'/8)H,(A),C)A0*()L,/M)0$()C,&2('"),%)&,%#'%A*,%$-)H$/M'A()/$A0'/)A0$%),%)<br />
H$/M'A()L*A0)%'AL,/M)'RA'/%$-*A*'()*%)L0*&0)"B%$H*&),/)b+&02H;'A'/*$%c)&,H;'A*A*,%)'#,-#'()<br />
3D#$%()$%")+&0H$-'%(''8)=>>?46)]')'R;-*&*A-B)-,,M)$A)A0')'CC'&A),C)%'AL,/M)'RA'/%$-*A*'(),%).TU)<br />
&,H;'A*A*,%)$%")*%A/,"2&')A0');,((*J*-*AB),C)C*/H()%,A)L$%A*%G)A0'*/)A'&0%,-,GB)A,)J');/,;/*'A$/B)<br />
L0*&0)*()%,A)/'&,G%*X'")JB)A0')A/$"*A*,%$-).TU)-*A'/$A2/'6)<br />
)<br />
]*A0*%)A0')A,;*&),C).TU8)A0'/')0$()J''%)(,H')L,/M)"'#,A'")A,)-*&'%(*%G6)K0*()-*A'/$A2/')A$M'()$()<br />
A0')(A$/A*%G);,*%A),%'),/)H,/')C*/H()0$#*%G)$);$A'%A6)S*&'%(*%G)*()A0'%)$)H'$%(),C)"*(('H*%$A*%G)<br />
$%) *%%,#$A*,%6) !H,%G) A0') *%&'%A*#'() C,/) -*&'%(*%G) $/') ;/,"2&A) H$/M'A) &,H;'A*A*,%) L0*&0)<br />
&/'$A'() *%&'%A*#'() C,/) H$%$G'/() L0,) L,2-") ,A0'/L*(') 'R;-,*A)A0'*/) H,%,;,-B) ;,(*A*,%(8) &,(A)<br />
($#*%G() A,) A0') -*&'%(''() L0*&0) &,2-") J') $;;/,;/*$A'") JB) A0')-*&'%(*%G) C*/H8) $%") -,L'/*%G)<br />
/*#$-(P)*%&'%A*#'()A,)*%#'%A)$/,2%")A0')*%%,#$A*,%6)F$AX)$%")+0$;*/,)3?`d[4)-,,M)$A)A0')<br />
*%&'%A*#'()A,)'%G$G')*%)-*&'%(*%G),%&')$%)*%%,#$A*,%)*()"'#'-,;'")*%)$)L,/-"),C);'/C'&A);$A'%A(6))<br />
)<br />
154<br />
!"#$%&'())@)+,&*'AB)C,/)+&*'%&')$%")D"2&$A*,%8)E%*A'")F*%G",H)<br />
)<br />
?I:)
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
HANS W. GOTTINGER<br />
!"##$%&'()*+,*-./012,*3'#4"(5*67"%"8$'9*:"(*;-! 5#,4#:-;! *B! 4! ,:5:4,?"! ?:>#+@:5!#*!'+55:J+>4#:!+>B*,J4#+*>!J4;!C:!#**!':!G.HH3I!<br />
:($!#"+5!B45"+*>!?4>!+>?,:45:!$,*B+#5!C;!<br />
-+@+>-!?*>5LJ:,5!4>!4'':'!!"#$"%!&$'#*!+>?L,!#":!5:#L$!?*5#!*B!$L,?"45+>-!#":!J4+>!?*J$*>:>#)!!<br />
9":!,:5L!*$:>!5;5#:J!+5!!+#!<br />
+5!J*,:!:!#":!5"4,:! *B!!<br />
#":!J4+>!?*J$*>:>#!+>!#":!#*#4! ,:@:,5:'! ?*J$4,:'! #*! #":! #,4'+#+*>45+>-! 5#:4'! *B! 5+>-! C:+>-! L5:'! 45! 4! J:4>5! *B! '+BBL5+>-! 4>! +>>*@4#+*>! ()%$*' +#! "45!<br />
*??L,,:'A!+>!*L,!J*':5+>-!4?#L4:,! G.H3TA! .H38I! 4>'! D4#E! 4>'! F"4$+,*! G.H3TA.H38I! 4,:! #":! $+*>::,+>-!K*,=5!<br />
:($-! #":! +J$5! *B! >:#K*,=! :(#:,>4! +>'L5#,+4+E4#+*>)! S4,,:'! 5+':! 4>'! 5"*K! #"4#! #":! :(+5#:>?:! *B! >:#K*,=!<br />
:(#:,>4!$,*C'!#"L5!#*!4!JL! *B! 4! >:K! 5#4>'4,'! J+-"#! C:! #**! 5'!F4:,M5!K*,=5!C;!+>?-! C#)!9":;!L5:!#":+,!J*':B*,?:!#":!+J$*,#4>?:!*B!?*>5LJ:,!<br />
:($:?#4#+*>5!+>!J4,=:#5!K+#"!>:#K*,=!:(#:,>4-!?*J$4#+C+'!K":#":,!5+':!$4;J:>#5!4,:!B:45+C'!F"4$+,*!G.H38I!5"*K!<br />
#"4#! +>! #":! $,:5:>?:! *B! >:#K*,=! :(#:,>45!L>':,!K"+?"!B+,J5!J4;!L5:!?*J$4#+C+-!<br />
?*J$:#+#+*>)!<br />
!<br />
W"+:#K*,=! :(#:,>4:K! $,*'L?#! K"+?"! :JC*'+:5! #:?">*!:UL+:K! $,*'L?#! +5! +>#,*'L?:'! #**! 5**>! ?*J$4,:'! #*! #":!5*?+45!<br />
C:?4L5:! #":! 5**>:,! #":! $,*'L?#! +5! +>#,*'L?:'A! #":! 5**>:,! #":! B+,J! ?4>! 5#4,#! CL+-!+#5!C:>:B+#5)!["*+!G.HH0I!!4!<br />
J4,=:#!K+#"!>:#K*,=:(#:,>4
HANS W. GOTTINGER<br />
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
!"#$%&'()*%)+,&*$-)+&*'%&'().'('$/&0)1,2/%$-)3!++.14)<br />
5,-678)9((2'):)!;/*-?:)<br />
)<br />
K')(A2"*'()A0')*%&'%A*#'),C)A0')H,%,;,-*(A)A,)*%A/,"2&')$%)*%&,H;$A*L-')*H;/,#'");/,"2&A)*%)<br />
A0') ;/'('%&') ,C) %'AM,/N) 'OA'/%$-*A*'(6) F/*(A*$%('%) 3?JJI4) (A2"*'() A0') &,%('P2'%&'() ,C) %'AM,/N)<br />
'OA'/%$-*A*'() ,%) A0') /*(N*%'(() ,C) .QR) ;/,S'&A() &0,('%) LB) $%) '%A/$%A) $%") $%) *%&2HL'%A6) K')<br />
(0,M() A0$A) A0') *%&2HL'%A) &0,,('() $) A,,) /*(NB) ;/,S'&A) A0$A) A,,) ,CA'%) -'A() $) %'M) C*/H)M*A0) $%)<br />
*%&,H;$A*L-')A'&0%,-,GB)'%A'/)$()&,H;$/'")A,)A0')(,&*$-),;A*H2H6)9%)$""*A*,%8)A0')'%A/$%A)<br />
0$() $%) *%&'%A*#') A,) &0,,(') H,/') &'/A$*%) ;/,S'&A() A0$%) $/') (,&*$--B) ,;A*H$-) $%") A0'(')<br />
(A/'%GA0'%) A0') ;,((*L*-*AB) ,C) $",;A*,%) ,C) $%) *%&,H;$A*L-') A'&0%,-,GB6) .'G*L'$2) $%") .,&N'AA)<br />
3?JJT4)$%$-BU')&,H;$A*L*-*AB)&0,*&'(),C)AM,)C*/H()M0*&0)H2(A)$-(,)"'&*"')M0'%)A,)*%A/,"2&')<br />
A0'*/) G,,"() *%) $) H$/N'A) &0$/$&A'/*U'") LB) %'AM,/N) 'OA'/%$-*A*'(6) V0'B) (0,M) A0$A) A0') C*/H(W)<br />
*%&'%A*#'() A,) $&0*'#') &,H;$A*L*-*AB) "';'%") &/2&*$--B) ,%) A0') A*H') $A) M0*&0) A0') "'G/'') ,C)<br />
&,H;$A*L*-*AB) H2(A) L') &0,('%6) V0') &2//'%A) ;$;'/) *() $-(,) /'-$A'") A,) A0') -*A'/$A2/') ,%)<br />
*%#'(AH'%A)$()$)H'$%(),C)'%A/B)"'A'//'%&'6)X*/H()&,H;'A')%,A),%-B)M*A0)'O*(A*%G)C*/H(8)$()*()<br />
A/$%(;$/'%A8)L2A)$-(,)M*A0);,A'%A*$-)'%A/$%A(6)9%)$%)'OA'%(*,%),C)A0')+;'%&')H,"'-)R*O*A)3?JT>4)<br />
(0,M() A0$A) $%) *%&2HL'%A) C*/H) H$B) H$N') *//'#,&$L-') &,HH*AH'%A) ,C) *%#'(AH'%A) *%) ,/"'/) A,)<br />
$-A'/)A0')*%*A*$-)&,%"*A*,%(),C)A0');,(A)2%-*N')A0')'%A/$%A)3OD()(>4*([')&,%(*"'/)$)AM,)(A$G')G$H')*%)$)H$/N'A)<br />
M*A0)%'AM,/N)'OA'/%$-*A*'(6)V0');/,"2&A(),C)A0')C*/H()$/')*%&,H;$A*L-')3*6'68)A0'B)0$#')(';$/$A')<br />
%'AM,/N(46)<br />
)<br />
9%) A0') C*/(A) (A$G'8) C*/H() &0$/G') $) -*&'%(') C'') 3+4) A,) ,A0'/) ",M%(A/'$H) C*/H() C,/) 2(') ,C) A0'*/)<br />
;/,"2&A)3C,/)'O$H;-'8)\*&/,(,CA)-*&'%('()A0'*/)[*%",M(),;'/$A*%G)(B(A'H)A,),A0'/)&,H;$%*'()<br />
A,)"'#'-,;)$;;-*&$A*,%()(,CAM$/'46)V0')%2HL'/),C)",M%(A/'$H)C*/H()M*--*%G)A,)M,/N),%)$)C*/HW()<br />
;/,"2&A)3H*4)"';'%"(),%)A0')-*&'%(')C'')&0$/G'")$%")$-(,)A0')*%*A*$-)%'AM,/N),C)A0')C*/H)3A0')<br />
&,%(2H'/)L$('),%)C,/)C*/H)*46)<br />
(<br />
-(%)(.(/(,%(0(+%8)(.(1(2( 3?4)<br />
)<br />
.(%3($)H'$(2/'),C)A0');,A'%A*$-)H$/N'A)(*U'),C)",M%(A/'$H)C*/H()*%"';'%"'%A),C)A0')-*&'%(')C'')<br />
$%")*%*A*$-)%'AM,/N6)<br />
)<br />
V0')(H$--'/)A0')-*&'%(')C'')&0$/G'"8)A0')H,/')C/''-B)A0')C*/H)"*(A/*L2A'()*A()A'&0%,-,GB)$%")*%),2/)<br />
A'/H*%,-,GB8) A0') H,/') ],;'%^) A0') C*/H) *(6) [') $((2H') %,) &,H;'A*A*,%) *%) G'AA*%G)<br />
",M%(A/'$H) C*/H() A,) L2B) A0') C*/H(W) -*&'%('(6) -%) "'&/'$('() $() +*) *%&/'$('() L'&$2(') $() A0')<br />
-*&'%('() L'&,H') H,/') 'O;'%(*#') C'M'/) ",M%(A/'$H) C*/H() L2B) A0'H6) V0*() *() S2(A) A0')<br />
(A$%"$/") $/G2H'%A) C,/) $) ",M%M$/") (-,;*%G) "'H$%") *%) ;/*&'6) !() C,/) A0') /'-$A*,%(0*;)<br />
L'AM''%( -%) $%") ,%(4) (*%&') A0') ",M%(A/'$H) C*/H() $/') "'#'-,;*%G) ;/,"2&A() A0$A) &,2-") L')<br />
2('")*%)&,%S2%&A*,%)M*A0)A0')C*/HW();/,"2&A)3-*N')&,H;-'H'%A(48)A0')-$/G'/)A0')*%(A$--'")<br />
L$('),C)&2(A,H'/()A0$A)A0')C*/H)0$(8)A0')-$/G'/)*()A0');,A'%A*$-)"'H$%")C,/)A0')",M%(A/'$H)<br />
C*/H(W);/,"2&A6))V02(8)A0')H,/')A0')",M%(A/'$H)C*/H()$/')M*--*%G)A,);$B)C,/)A0')-*&'%('6)<br />
)<br />
_'A)2()"'C*%'`))9C)))+% ( !a)+5 ( !A0'%)C*/H)*)&0,,('()A,)L')H,/'),;'%)A0$%)C*/H)S6)3!-A'/%$A*#'-B(C*/H)S)<br />
&0,,('()A,)L')-'((),;'%)A0$%)C*/H)*4*(<br />
)<br />
)<br />
156<br />
!"#$%&'())@)+,&*'AB)C,/)+&*'%&')$%")D"2&$A*,%8)E%*A'")F*%G",H)<br />
)<br />
?IJ)
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
HANS W. GOTTINGER<br />
!"##$%&'()*+,*-./012,*3'#4"(5*67"%"8$'9*:"(*;##*! 5:@@! *?#$?#)! N:9;:! D:! ,:5#,+;#! #":!<br />
:9#,49#!#*!;"**5+9-!4!@+;:95:!*?,9*#!-4=:)!B"?5H!#":!X'*D95#,:4=!
HANS W. GOTTINGER<br />
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
!"#$%&'()*%)+,&*$-)+&*'%&'().'('$/&0)1,2/%$-)3!++.14)<br />
5,-678)9((2'):)!;/*-?:)<br />
)<br />
!()*%),A0'/)AI,
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
HANS W. GOTTINGER<br />
!"##$%&'()*+,*-./012,*3'#4"(5*67"%"8$'9*:"(*;!'9:+5+*;!*?!?+,@!A!4;'!#"9!9BC+D+E,+C@!D+:9;59!<br />
?995!?*,!#"9!#!+;!#"9!H*C,;*#!5CE-4@9G!#*!#"9!'+??9,9;#!D+:9;59!<br />
?995!#"4#!#"9!+;:C@E9;#!:4;!:"4,-9)!I*,!$,**?5!$D9459!,9?9,!#*!#"9!4$$9;'+()!89!:4;!5"*!*C#!*?!#"9!@4,J9#!4;'!?*,!"!##>##" ! = #+#!+5!#*!9;#9,)!K"9,9!9(+5#5!"! ? ##>##" ! = !<br />
5C:"!#"4#!+;!#"9!+;#9,L4D!M"!# = #6#"! ? N##"9!E95#! ,95$*;59! *?! #"9! 9;#,4;#! +5! #*! :"4,-9! #"9! D4,-95#!<br />
4::9$#4ED9! D+:9;59! ?99! #"4#! @4J95!*$#+@4D! (A! 9(4:#D>! /G! 4;'! ?*,! ƒO! P! ƒ! ? ! +#! +5! *$#+@4D! ?*,! #"9!<br />
9;#,4;#!#*!9;#9,!4;'!:"4,-9!+#5!C;:*;5#,4+;9'!*$#+@4D!D+:9;59!?99!,95CD#+;-!+;!4!$*5+#+L9#<br />
@A#<br />
%<br />
&'()*(+%,%&'(%-(.(-/%)0%12 ) 3%4%567%8%7(9(*+:6(%9'(%(69*;%7(+%5/%0)--)@/$%<br />
.) A;#,>!+5!ED*:J9'!+?!Q.!R!7!S!(O! * !"!T!R!7J)!<br />
7) I*,!T!R!7J!"!Q.!R7!S!(O! * !"!TRJ&7G!9;#,>!+5!,95#,+:#9'!)!#"9!9;#,4;#!<br />
9;#9,5!#"9!@4,J9#)!Z*!?995!4;'!@4J9!+#5!@*;*$*D>!$,*?+#5!! "+-"! #"9;! #"9! +;:C@E9;#!:4;! ,95#,+:#!<br />
9;#,>!4;'!@4J9!@*;*$*D>!$,*?+#5)!O?!+#5!4'L4;#4-9!+5!"*!$,*?+#5)!K"C5!<br />
+;!#"+5!:459!#"9!+;:C@E9;#!+5!4ED9!#*!4:#!45!4!@*;*$*D+5#!+;!#"9!*C#$C#!@4,J9#!EC#!;*#!+;!D+:9;59!<br />
?995)!<br />
!<br />
O?!#"9!+;+#+4D!;9#
HANS W. GOTTINGER<br />
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
!"#$%&'()*%)+,&*$-)+&*'%&'().'('$/&0)1,2/%$-)3!++.14)<br />
5,-678)9((2'):)!;/*-?:)<br />
)<br />
!()&,/,--$/*'()A,)J/,;,(*A*,%)=)K')G'A)A0')/'(2-A)A0$A)A0'/')*()%,)'%A/B)"'A'//'%&')$%")K')$-(,)G'A)$)<br />
&,H;$/*(,%) L'AK''%) A0') H,%,;,-B) -*&'%(') C''() ,C) A0') *%&2HL'%A) $%") *A() -*&'%(')C''() K0'%) $)<br />
;,A'%A*$-)'%A/$%A)'M*(A(6)<br />
)<br />
!"#"$$%#&'(')*+#+',-'."'+./#&'0+/+##+.1+2'<br />
!A)'!"# $# 'A0')'%A/$%A);/,"2&'()%,A0*%G)$%")A0')*%&2HL'%A)N2(A)H$O'()A0');/,C*A()*A)K,2-")$()$)<br />
H,%,;,-*(A) *%) A0') ('&,%") (A$G') ,2A;2A) G$H'6) +*%&') !"# $# ' *() A0') -$/G'(A) -*&'%(') C'') C,/) A0')<br />
*%&2HL'%A)A0$A)"'A'/()'%A/B8)*C)#!"# %# !#!"# $# 'A0'%)A0'/')*()L-,&O'")'%A/B6)9C)!" % #&#!"# $# '0,K'#'/8)A0'%)<br />
'%A/B)*()%,A)L-,&O'"6)P2A)*A)$-(,)H'$%()A0$A)A0')*%&2HL'%AQ();/,C*A()H2(A)L')*%&/'$(*%G)$A)!"# $# '<br />
3(*%&')A0')H,%,;,-B);/,C*A()$/')&,%&$#')*%)!" # '$%")C/,H)R'HH$)!6=)K')O%,K)A0$A)*%)A0')*%A'/#$-)S!"#<br />
$ ''!"# (# ]#'#A0')*%&2HL'%A)H$O'()H,%,;,-B);/,C*A()46)T'%&')A0')*%&2HL'%A)K*--)%,A)C*%")*A);/,C*A$L-')<br />
A,)&0$/G')!"# $# ',/)-'((6)U02()'%A/B)*()%,A)"'A'//'"6))<br />
)<br />
!()$)&,/,--$/B)A,)U0',/'H)=)K')G'A)$)&,H;$/*(,%)L'AK''%)A0')H,%,;,-B)-*&'%(')C''(),C)A0')<br />
*%&2HL'%A)$%")*A()-*&'%(')C''()K0'%)$);,A'%A*$-)'%A/$%A)'M*(A(6)<br />
)<br />
!"#"$$%#&'3'4/'/*+'+56,$,7#,68'06"9"$&'$,1+.-+':++':;
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
HANS W. GOTTINGER<br />
!"##$%&'()*+,*-./012,*3'#4"(5*67"%"8$'9*:"(*;9:!#":!;:#A*,C!#:,=H!A:!5::!#"4#!#":!+;9>=D:;#!9"**5:5!#*!D:!:L:;!=*,:!*$:;!A"+8:!<br />
#":!:;#,4;#!9"**5:5!#*!D:!8:55!*$:;)!!<br />
!<br />
O,*=!#":!P$$:;'+(!QP!)./@!A:!5::!#"4#!5#4,#+;-!9#+*;!+;!!%"# +;!<br />
*,':,!#*!D,+;-!#":!=4,-+;48!:8#!+#!+5!<br />
,:!#":!9:;#,48!+':4!*#>,:!<br />
,:5:4,9"!:':'! #":! #,4'+#+*;48! M#:9";*8*-?! 5#:48+;-M! :=:,5H!;4=:8?!#"4#!9*;5>=:,5!A+88!$:,9:+L:!#":+,!<br />
$,*'>9#! #*! D:! ;*#! 45! 4##,49#+L:! 45! #":! +;9>=D:;#J5H! :L:;! +48+#?)! E#!<br />
A*>8'!5#+88!D:!+;#:,:5#+;-!#*!5::%#":!,:5>8#!*-"!#*!4!8:55:,!:(#:;#)!I":!<br />
:R>+8+D,+>=!;*A!;::'5!#*!5#,+C:!4!D484;9:!D:#A::;!#":!#A*!*$$*5+;-!:8'! :($:9#! #":! ,:5>8#! #"4#! #":! +;9>=D:;#! -*:5! =*,:!<br />
*$:;!#*!D:!,:+;9:!9*=$:#+#+*;!+;!#":!
HANS W. GOTTINGER<br />
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
!"#$%&'()*%)+,&*$-)+&*'%&'().'('$/&0)1,2/%$-)3!++.14)<br />
5,-678)9((2'):)!;/*-?:)<br />
)<br />
$%) *"'%A*&$-) $%") *%"';'%"'%A) "'H$%") C,/) *A() -*&'%('(6) J0*() &,2-") K') L2(A*C*'") *C) A0'/') $/')<br />
%2H'/,2() ",M%(A/'$H) C*/H() -,,N*%G) C,/) ;/,(;'&A*#') A'&0%,-,G*'() A,) "'#'-,;6) !) H,/')<br />
/'$-*(A*&)('AA*%G)H$B)K'),%')M0'/')K,A0)A0')*%&2HK'%A)$%")A0')'%A/$%A)$/')&,H;'A*%G)*%)<br />
-*&'%(') C''() C,/) ",M%(A/'$H) C*/H() A,) K2B) A0'*/) -*&'%('(6) D$&0) ",M%(A/'$H) C*/H) K2B() $A)<br />
H,(A),%')2%*A),C)A0')-*&'%('6))<br />
)<br />
O,M'#'/8)A0*()(0,2-")%,A)&0$%G'),2/)/'(2-A()"/$(A*&$--B)$%")M')(0,2-")(A*--)('')P,;'%P)$()$%)<br />
'Q2*-*K/*2H)(A/$A'GB6))<br />
)<br />
9%)C$&A8)M*A0)&,H;'A*A*,%8)K,A0)A0')*%&2HK'%A)$%")A0')'%A/$%A)$/')-*N'-B)A,)G,)H,/'),;'%)<br />
K'&$2(')%,M)A0'B)%''")A,)&,H;'A')A,)('--)A0'*/)-*&'%('(6)<br />
)<br />
R')&,2-")*%A/,"2&')&,(A()*%A,)A0')A'&0%,-,GB),C)A0')C*/H(6)9%),2/)&2//'%A)H,"'-8)M')0$#')<br />
%,/H$-*S'")K,A0)C*/H(T)&,(A()A,)S'/,6)U*CC'/'%&'()*%)A0')A'&0%,-,G*'(),C)A0')C*/H()H$B)(M$B)<br />
,2/)/'(2-A()'*A0'/)M$B)"';'%"*%G),%)M0,)0$()A0')K'AA'/)3-'(()&,(A-B4)A'&0%,-,GB6)9A)M,2-")<br />
$-(,)K')*%A'/'(A*%G)A,)('')0,M)A0')&,2;-*%G),C)A0')PA'&0%,-,GB)(A'$-*%GP)'CC'&A)$%");,(*A*#')<br />
&,(A()H$B)M,/N6)V,/)'W$H;-'8)*C)A0')*%&2HK'%A)C/''-B)"*(A/*K2A'()*A()A'&0%,-,GB)A0'%)A0')<br />
&,(A() C,/) A0') '%A/$%A) H$B) C$--) K'&$2(') *A) &$%) %,M) P(A'$-T) A0') A'&0%,-,GB) ,C) A0')<br />
*%&2HK'%A6)<br />
)<br />
9%),2/)H,"'-8)M')",%TA)0$#')'%A/B)"'A'//'%&')K2A)M')0$#')/'(A/*&A'")'%A/B)*%)M0*&0)A0')<br />
'%A/$%A)('--()-*&'%('()*%)A0')C*/(A)(A$G'8)K2A);/,"2&'()%,A0*%G)*%)A0')('&,%")(A$G'6)J0*()<br />
H$B) (''H) $) K*A) 2%/'$-*(A*&6) R*A0) ;,(*A*#') C*W'") &,(A(8) M') (0,2-") K') $K-') A,) 0$#') A0')<br />
(A$%"$/")'%A/B)"'A'//'%&')/'(2-A6)R')M,2-")(A*--)'W;'&A)/'(A/*&A'")'%A/B)K2A),%')M0'/')A0')<br />
'%A/$%A)*()T/'(A/*&A'"P)A,)(,H');,(*A*#'),2A;2A)-'#'-6)<br />
)<br />
!%,A0'/);,((*K-')'WA'%(*,%)*()H$N*%G)*%%,#$A*,%)(A,&0$(A*&8)*6'6)/';-$&*%G)A0')-*%N)!"#$%#&#$#'(%)#<br />
M*A0) $%#) X)$'(%*+%)# *#M0'/') + % #*() (A,&0$(A*&6) R') $/G2') A0$A) A0*() $G$*%) (0,2-") %,A) &0$%G') ,2/)<br />
/'(2-A() Q2$-*A$A*#'-B) (,) -,%G) $() A0') /'(2-A*%G) "*(A/*K2A*,%() C,/) Q2$-*A*'() 'W0*K*A) ('&,%") ,/"'/)<br />
(A,&0$(A*&) ",H*%$%&')
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
HANS W. GOTTINGER<br />
!"##$%&'()*+,*-./012,*3'#4"(5*67"%"8$'9*:"(*;/:N0/ONN2!<br />
EK*F*I+'B5)!A)!=.>>8?!@S"B!EK*F*I+K5!*U!AB#C*,D5M:!-)0#()*01&)*+$%&'()*+$&,$-).'/0(1*+$5(6*)17*01&)$.0:!133O3//!<br />
S"B!EK*F*I+5#!=7//048$9:)$5;#)3N?!@EK*F*I+K!V55YB5!+F!Z#4F'4,'+`4#+*FM:!+F!;4IB5!\+GGB,:!B'):!!#+#3&44')13*01&)/$*).$<br />
2G'10HI$J&+13H$K#/#*(3"$-//'#/:!A*,#"!a*GG4F':!ABC!b*,D:!.8NO.13$<br />
^4,,BGG:!;!4F'!Z4G*FB,:!_)!=.>38?!@VF5#4GGB'![45B!4F'!9*I$4#+J+G+#H%!VFF*W4#+*F:!3N?!@LF!#"B!c+KBF5+F-!*U!VFF*W4#+*F5M)!K*).$%&'()*+$&,$23&)&413/$.8:!N/0ON7/!<br />
R4#`:!\)!4F'!Z"4$+,*:!9)=.>38?!@>7?!!d8?!@TeP!+F!\4,DB#5!C+#"!AB#C*,D!E(#B,F4G+#+B5M:!-)0#()*01&)*+$%&'()*+$&,$-).'/0(1*+$5(6*)17*01&)$<br />
.0:!N2.ON01!<br />
cB,FB,:;)!4F'!;)!S+,*GB!=7//0?:!S"B!EK*F*I+K5!*U!SBK"F*G*-H!Z"4,+F-%!L$BF!Z*Y,KB!4F'![BH*F'M:!a4,W4,'![Y5+FB55!ZK"**G!]*,D+F-!<br />
HANS W. GOTTINGER<br />
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
!"#$%&'()*%)+,&*$-)+&*'%&'().'('$/&0)1,2/%$-)3!++.14)<br />
5,-678)9((2'):)!;/*-?:)<br />
)<br />
*++,-./01<br />
+%""21"213(4"%45161<br />
I0'%)A0')H,%,;,-*(A)0$()$);/,"2&A),C)J2$-*AB)!"#$%")&0$/G'()$);/*&'),C);9)A0'),%-B)&,%(2H'/()<br />
K0,) L2B) C/,H) *A) $/') A0') ,%'() C,/) K0,H) $# %# !"# %# !# &" # ' # (# )"# # !) >6) M0'/'C,/') "'H$%") C,/) A0')<br />
H,%,;,-*(AN();/,"2&A)'J2$-()*#%#!"#%#!#&" #' #(#)"#+##9AN()/'#'%2'3.4)'J2$-())"#,*#%#!"#%#!#&" #' #(#)"-+6)M0')<br />
,;A*H$-);/*&')C,/)('--*%G)*A(),2A;2A)*()A02()G*#'%)LBO)<br />
)<br />
" .P" ; 9 ) Q )!)R)J9)R)#)S9 ), ))<br />
)<br />
,/))))))))))); 9 T QS9 T Q3)!)R)J9)R)#)S9 ), 4P)=))))))))))))))))))))))))))))3!?4)<br />
)<br />
M0')H,%,;,-*(AN();/,C*A)*()A0'/'C,/')<br />
$9 T) Q)[)3)!)R)J9)R)#)S9 ), 4P)=] )= )R)C9)3U9)R)S9 ), )
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
HANS W. GOTTINGER<br />
!"##$%&'()*+,*-./012,*3'#4"(5*67"%"8$'9*:"(*;B,>45>5!#">!<br />
$,*=+#!=,*:!#">!=+,5#!#>,:!D5+;B>!=,*:!#">!E*9,;*#!-4:>A!"#!+5!'>B,>45+;-!+;!ƒ+$%!8#!#">!*$#+:4?!ƒ+!<br />
#"+5! :95#! ! >(4B#?C! *==5>#! !>F9+?+,>45!#">!=+,:!+H5!,>@>;9>!<br />
=,*:!?+B>;5+;-!+5!9;4==>B#>'!;!4!5:4??!+;B,>45>!+;!ƒ+!=,*:!+#5!<br />
$,>@+*95!*$#+:4?!?>@>?!+;B,>45>5!#">!,>@>;9>!=,*:!#">!?+B>;5>!=>>5!!'>B,>45>!+;!,>@>;9>!=,*:!'+,>B#!54?>5!+5!?4,->,!5+;B>!"#!+5!?4,->,)!I"95!G+#"!<br />
#">!?4,->,!ƒ6!$,*=+#5!4,>!'>B,>45+;-!+;!+H5!?+B>;5>!=>>5!4#!#">!$,>@+*95?C!*$#+:4?!ƒ+%%!K+@>;!#">!<br />
B*;B4@+#C!*=!#">!$,*=+#!=9;B#+*;!+;!4!=+,:H5!*G;!?+B>;5>!=>>!#"+5!,>F9+,>5!4!?*G>,+;-!*=!ƒ+!+;!*,'>,!#*!<br />
,>4B"!>F9+?+;B>!#">!,>4B#+*;!=9;B#+*;!+5!'*G;G4,'!5?*$+;-)!<br />
!<br />
L>::4! 8.)! I">,>! >(+5#5! =M N ! =*,! #">! +;B9:;#! 59B"! #"4#! =*,! 4??! =M! O! =M N !!#">! *$#+:4?!<br />
,>5$*;5>!*=!#">!>;#,4;#!+5!#*!5#4C!*9#!*=!#">!:4,P>#!4;'!=*,!=M!Q!=M N! +#!+5!#*!>;#>,)!=M N !+5!#"95!<br />
#">!?4,->5#!?+B>;5>!=>>!=*,!#">!+;B9:;#!#"4#!'>#>,5!>;#,C)!<br />
!<br />
R,**=&%L>#!!'( ) !!#">!?+B>;5>!=>>!=*,!G"+B"!'*!+!/!+5!#">!5:4??>5#!?+B>;5>!=>>!#"4#!:4P>5!"*!,'*&!'( ) $!S!/)!<br />
T">;!'*!+!-%!"*!S!/A!4;'!'(!+!'( ) !#">;!!*!S!/!4;'!"!*!/!"ƒ*!+!/!U!P!Q!/)!I"95!,>'9B+;-!'*!G+??!<br />
*;?C!:4P>!!*!!;>-4#+@>!G"+?>!4!,!'*!+5!;*#!4;!4BB>$#4!?+B>;5>!=>>)!V*!>;#,C!+5!'>#>,,>'!4#!<br />
'( ) %!<br />
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!<br />
I*!B4?B9?4#>!'( ) !G>!5>#!!"*!>F94?!#*!W>,*!4;'!'*!+!-!G+#"!+#)!<br />
T>!#">;!->#!<br />
!<br />
8U7DP!!!ƒXY!!!DP!U!(M! 0 !!!!=M N Y!!!"(M! * !S!/!!!!!!!!!!!!!!D8ZY!<br />
!<br />
I">,>=*,>A!=M N !S!D.!U!"Y!(M! !!!8!!!P!!!!!!!!!!!!!!!!!!!!!!!!!!!D81Y!<br />
0<br />
!<br />
X;#,C!+5!'>#>,,>'!4#!4;C!ƒ(!.!'( ) !!#**A!5+;B>!#">;!#">!5:4??>5#!'*!!#"4#!:4P>5!"*!,'*&!'($!S!/)!:95#!!<br />
;>-4#+@>!A5>>!D8ZY)![*,!'(!/!'( ) !#">!5:4??>5#!!' * !#"4#!:4P>5!"*!,'*&!'($!S!/)!+5!$*5+#+@>A!#"95!#">,>!'*!<br />
>(+5#! 4BB>$#4! ?+B>;5>! =>>5! #"4#! -+@>! #">! >;#,4;#! $*5+#+@>! $,*=+#5)! J>;B>! #">! >;#,4;#! G+??!<br />
>;#>,!4;'!5*!#">,>!+5!;*!>;#,C!'>#>,,>;B>)!I"95!'( ) !#0!+;'>>'!#">!?4,->5#!?+B>;5>!=>>!#"4#!#">!<br />
+;B9:;#!B4;!B"4,->!4;'!5#+??!'>#>,!>;#,C)!<br />
!<br />
L>::4!87&%%I">,>!>(+5#5!=M \ !Q!=M N !, 59B"!#"4#!+;!#">!+;#>,@4?![!=M N A!=M \ ]!!+#!+5!*$#+:4?!=*,!=+,:!X!#*!<br />
B"4,->! #">! ?4,->5#! 4BB>$#4! ?+B>;5>! =>>! =X!!#"4#! :4P>5! (X! S! /! 4;'! =*,! =M! Q! =M \ ! #">! *$#+:4?!<br />
,>5$*;5>!=*,!=+,:!X!+;'9B>5!+#!#*!$,*'9B>!4!$*5+#+@>!4:*9;#)!<br />
!<br />
R,**=&%[,*:!>F94#+*;!D8]Y!G>!P;*G!#"4#!#">!5#!,>5$*;5>!=*,!#">!>;#,4;#!#*!=M \ +5!-+@>;!;!=M!S!=M \ A!!= \ X!!:95#!!45!-+@>;!!(H5)!#">;!G">;!=+,:!M!B"*5>!'( 1 !4;'!=+,:!X!B"*5>!+#5!?+B>;5>!<br />
=>>!*$#+:4??CA!! +#5! ?+B>;5>! =>>! *$#+:4??CA! #">;! +;! #">!<br />
59F9>;#!*9#$9#!-4:>!=+,:!X!G*9?'!>;'!9$!5>??+;-!4!$*5+#+@>!4:*9;#)!M=!!"*G>@>,!=+,:!M!<br />
B"*5>!'(!;!+;!#">!<br />
59F9>;#!*9#$9#!-4:>!=+,:!X!G*9?'!!,>F9+,>'!#*!5>??!4!;>-4#+@>!4:*9;#)!V+;B>!#"4#!+5!<br />
!<br />
!"#$%"##$%&&'()'*+)*,-&./).0123&455,6)01)72.7)% .13!<br />
!<br />
165
HANS W. GOTTINGER<br />
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
!"#$%&'()*%)+,&*$-)+&*'%&'().'('$/&0)1,2/%$-)3!++.14)<br />
166<br />
5,-678)9((2'):)!;/*-?:)<br />
)<br />
%,A);,((*J-')D)K2(A)('A()*A()-*&'%(')C'')$A)A0')-'#'-)A0$A)K2(A)H$L'()*A(),2A;2A)*%)A0')%'MA)(A$G')<br />
N'/,6)O'%&') P0'%) A0'/') *() %,) '%A/B) J-,&L8) *%) A0') *%A'/#$-)[!"# $ %!"# & ]!!C*/H) D) &0$/G'() A0') -$/G'(A)<br />
$&&';A$J-')-*&'%(')C'')"'!A0$A)H$L'()('!)!*+!Q*/H)9)H$L'();/,C*A()$&&,/"*%G)A,)A0')H,%,;,-*(AR()<br />
;/,C*A)C2%&A*,%6)S0*()*()P0'/')A0')*%'T2$-*AB)M*)!)>)C,/)A0')'%A/$%A)*()J*%"*%G6)<br />
)<br />
Q,/)"#!,!"# & !A0')'%A/$%A)K2(A);-$B()*A()2%&,%(A/$*%'")J'(A)/'(;,%(')*%)A0')+A$&L'-J'/G)G$H'*+!"# & )*()<br />
A0') (H$--'(A) -*&'%(') C'') C,/) A0') *%&2HJ'%A) $J,#') P0*&0) A0') '%A/$%A) ;-$B() *A() 2%&,%(A/$*%'")<br />
,;A*H$-)C''()$%"),2A;2A)*%)'T2*-*J/*2H*+<br />
+<br />
,%""-+"-+.(/"%/0+1+<br />
U');/,#')A0')A0',/'H)*%)$)C'P)(A';(6)<br />
)<br />
-.//0!12+!'3456!78!9:;!7"!?@!A!B!C!(# ; "!1!A!B=+!<br />
)<br />
U')C*%")A0')(H$--'(A)"'!A0$A)G*#'()2()('!?"'%!"# D C!V)>!<br />
Q/,H)3!W4)P')G'AX)<br />
)<br />
!)Y)=3L")ƒD4)")3L)Y)M9 > )")C9 Z 4)")#M9 , )V)>)<br />
)<br />
+2J(A*A2A*%G)A0')'M;/'((*,%)C,/)"# D !f C/,H)'T2$A*,%)3!=4!*%A,)A0')$J,#')G*#'()2(X)<br />
)))))))))))))))))))))))))))))))))))))))))))))))))))<br />
ƒD $) V)!)Y=L)")3?)Y)=#4M9 , [\)<br />
))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))<br />
ƒD $ !78!A0')-$/G'(A)$&&';A$J-')-*&'%(')C'')C,/)A0')'%A/$%A)P0'%)A0')*%&2HJ'%A)&0$/G'()"# D )<br />
U0'%)ƒD $ %)>)8)#'!)$A)ƒD $ )*()%,%)>6!<br />
)<br />
]*#'%)A0')&,%&$#*AB),C)#'!)*%)ƒD)))H'$%()A0$A)*C)!ƒD %)>)8)<br />
$<br />
A0'/') *() %,) $&&';A$J-') -*&'%(') C'') C,/) A0') '%A/$%A) A0$A) G*#'() *A) $) ;,(*A*#') ;/,C*A6)O'%&') '%A/B) *()<br />
J-,&L'")*C)!ƒD %)>)8)*6'6)*C)<br />
$<br />
)<br />
!Y=L^)3?Y=#4M9 ,)<br />
)<br />
ƒD V) >) *() A0') -$/G'(A) $ ƒD ) C,/) P0*&0) '%A/B) *() J-,&L'") (*%&') *C) $ ƒD ) >) >) A0'%) A0'/') ",'() 'M*(A) $%)<br />
$<br />
$&&';A$J-')-*&'%(')C'')C,/)A0')'%A/$%A)A0$A)G*#'()*A);,(*A*#');/,C*A6)<br />
!<br />
-.//0!1EFG;5!1AB="!?@!A!B!C!(# ; !"!1!A!=HB%!.3456!78!5.8457!I74J!4J.!73
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
HANS W. GOTTINGER<br />
!"##$%&'()*+,*-./012,*3'#4"(5*67"%"8$'9*:"(*;.=H#?!(9! * !<br />
!<br />
IE##+F-!#*-D#"D,!#"+5!B*F'+#+*F!A+#"!#"4#!8*,!F*!JK*B.!=#?!(9 * "!@T!<br />
&!" . /"!!"#/#567#[8#9#:+# #7&5#9:$-#1"# ; #]#<br />
!<br />
!<br />
!" . !+5!'DB,D45+F-!+F!!!"##*4#!"#/#ƒ" ; !+8!<br />
.!=7#?!(9 * !<br />
!<br />
["E5!8*,!7@&P!=!3# ?;@# 7@&P=
HANS W. GOTTINGER<br />
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
!"#$%&'()*%)+,&*$-)+&*'%&'().'('$/&0)1,2/%$-)3!++.14)<br />
)<br />
5,-678)9((2'):)!;/*-?:)<br />
ƒ! . "#+*%&')!! $ )*()&,%&$#'8)A0'),;A*H$-)-*&'%(')C'')C,/)A0')*%&2HJ'%A)H2(A)J')G*#'%)JB)A0'),%')<br />
A0$A)H$K*H*L'()!! $ "#)M$--)A0*()-*&'%(')ƒ! + "#+'AA*%G)!!! $ /")%!# N)>)C/,H)3!?>4)O')G'A)<br />
)))))))))))))<br />
ƒ9 + )N)[PQ)")I!)")3:)R)S=#4)K9 , ])T?I))))))))))))))))))))3!??4)<br />
)<br />
U*#'%),2/)"'C*%*A*,%),C#ƒ! . &#%!# #ƒ! . )*(#,;A*H$-)C,/)'()V)>6))<br />
)<br />
W02()*C))#*#+,)VP-.#*+$/#'! 0 8)A0'%)A0')*%&2HJ'%A)&0$/G'(#ƒ! $ &#A0')'%A/$%A)&0$/G'()%(#G*#'%)JB)*A()<br />
/'$&A*,%)C2%&A*,%8)$%")J,A0)C*/H();/,"2&');,(*A*#')$H,2%A(6)<br />
)<br />
*%""+,"+,!"%"--.%$,/,<br />
X/,H) Y'HH$() ,+#$%") !S) O') *%C'/) A0$A) C,/) ,#*#)1+) $-.#*+$/# '!# 0 8) A0') *%&2HJ'%A) &0$/G'()*A()<br />
,;A*H$-)H,%,;,-B)-*&'%(')C''8)ƒ! 2 "##Z'%&')A0')(A$A'H'%A),C)A0')&,/,--$/B)*()A/2')*%)A0*()&$('6)X/,H)<br />
Y'HH$)!78)O0'%)=!TP)R))13#%#-.#*+$/#'!#%4,#*#)1+"#A0'),;A*H$-)%!#56#ƒ! . # "#[2A)*%)A0*()(&'%$/*,)<br />
H,%,;,-B);/,C*A()$/')*%&/'$(*%G)*%)A0')*%A'/#$-)[#%! 7 &#%! 8 ]#$()O')($O)*%)Y'HH$),9#6)W02()*A)H2(A)<br />
J')A0')&$(')A0$A)%!# 2 >ƒ! . )N)%! & ##"#X*%$--B8)O0'%)))#*#+,)VP-.#*+$/#'! 0 )&#A0'),;A*H$-)-*&'%(')C'')C,/)<br />
A0')*%&2HJ'%A)*()G*#'%)JB)ƒ! + )*%)'\2$A*,%)3!??46)<br />
)<br />
ƒ9 +) ")ƒ9 ] )N))")[=!)R)Q)R)?S)3?)R)=#4)K9 , ])T?I))6)<br />
W02()O0'%))#*#+,#:3-.#*+$/#'! , 8)ƒ9) ] >ƒ9 + )N)ƒ9 % )<br />
,<br />
*%""+,"+,0(1"%12,3,<br />
9A) &$%) J') &0'&Q'") A0$A) *%) A0') &$('() O0'%) A0'/') *() 2%/'(A/*&A'") '%A/B) $%") O0'%) A0'/') *()<br />
/'(A/*&A'") '%A/B) O*A0,2A) H,%,;,-B) ,2A;2A8) %!#
2.3 VERTICAL COMPETITION AND OUTSOURCING<br />
HANS W. GOTTINGER<br />
!"#$%"&#'("&)*+(,%"&)*(-*.,/'"$//*&"0*12("(3'2/*4$/$&%25*<br />
"#$%&!'()*+!,$%-,""!<br />
./0123456!781285!95:5;05*&!GHHI+!","E-F%)J!(K81285*! !<br />
6$%#'2&)*7(38$#'#'("*&"0*9,#/(,%2'":*'"*&*;,88)5CM=!N5:48717CM=!O:787;2:3=!HNPQNOR!R783/1>28C=!S/82:4=!T55:U:VML477B:7;!<br />
4)%5+*-%*6+$%".*+51-3%<br />
WL83!XB!T7>>28C52>278!L86!K/>37/L8:53! 28! ;L8/[L:>/2783! 7[!<br />
17:]-283!:56!0M!2887\L>2\5!3/??125551! L86! S2:! L3! 3/??1257!<br />
653]>7?! ;L8/[L:>/278-0L356! 17:]-283B! G86556=! >453! L8! 5\71\28C!<br />
?7A5/3!;5;052>278! L3!<br />
L665C2:L11M!0M!.755Z>!<br />
23! 58\23LC56! 28! A42:4=! L>! L8M! C2\58! ?728>! 28! >2;5! 7[! >45!<br />
278342?=! 07>4! >45! 0/M545! 3/??125278! 32;/1>L857/31MB! X5! 4L>! >45!<br />
?278! [7;58>3! 28! 2887\L>278! 0M!<br />
;5;0545!3/??1M!:4L28!23!>7!28:4522L1!<br />
72\5! :4L8851! ?7A545! 3/??1M! :4L28B! Q137! 28!L!<br />
32>/L>278!A4545!0/M545!0/M52\51M! ?/45! :278! 7[! L! 3/03>2>/>5! >5:48717CM!<br />
0M! 28\53>28C! 28! 2887\L>278B! N45! ?278! [745!<br />
0/M5423!:L35!A7/16!05!>7!512;28L>5!>45!>5:48717CM!17:]-<br />
28!L86!05:7;5!2865?58658>B!<br />
X5! :L8! 177]! L>! 47A! >42:L1!28>5C278+!<br />
$*!H>L?153!28!>45!7[[2:5!?3!286/3>23>5?=!<br />
42C4-:73>! \L1/5! :4L28B! N45M! 58\2327856! L! [/>/7;556! 0M! 3475828C! >45! :4L28! >7! :L?>/423! \L1/5! ?278! A2>4! L! ;7651! >4L>!<br />
3/??756! 2>3! L21! 3>74! 2>3! 7A8! :58>278! :58>557! 512;28L>5!<br />
;26615;58!L86!28:
HANS W. GOTTINGER<br />
2.3 VERTICAL COMPETITION AND OUTSOURCING<br />
! G8>5278L1!f7/3!>4L>!>45!:L!L3!L8!5ZL;?15!<br />
7[! i>45! 65:7;?732>278! 7[! >45! \52:L11M! 28>5C56! 0/328533!<br />
;7651j![45!785!>4L>!/356!>7!05!\55C56!L3!?L!7[!<br />
L!342[>!28!L!\L1/5!85>B!G8!3/:4!:2L8:53=![2545!<br />
:5CM! 7/C4>! >7! 05!28! ?1L:5! >7!<br />
65[586! >45! L0212>M! >7! L??5! \L1/5! 0M!<br />
583/4L>! >45! 3/??12545! ! >45! [2! >7! 7A8! L! L015! >7! ?/>! >45;351\53! 28! L!<br />
?732>278!>7!15\57!>45![2=! ?4L>! 2>! 37/45! ;2:3>L8628C!3/::533[/1!?!6532C8!0/>!L!623L3>51!28!>45!<br />
3/??1M! :4L28! [756! 0/>! :4L1158C56! 0M!<br />
Q6\L8:56!S2:2783!>4L>!<br />
7::/423! 286/3>58>! 65[2856! 0M! >423!<br />
/?3>7!<br />
05AL45!bG8>51!G83265c!3M8658628C! >423! L! >7! >45! /?3>2783B!9/45! $d)#3=! >45! ?2:5! 7[! 35:786! 37/! :L8! L:>! L3! L! 35:786! 37/4L>!37;5!3715!<br />
3/??125756c!>4527;545M! 4L6! 17:]56! 28! >452! 6532C83! >7! >4L>!7[!>45!<br />
3/??125/45! 286/3>45!5\71/>278!7[!;L!3>/>2\5! 3>5CM! [757! ?1M! A4215! 7[[52:L1! >7! >4L>! 7[! >45! 2887\L>28C! [2423!<br />
4/!:7/16!/35!2>3!C7!<br />
3/??7!L!1L4!L!\25A!>7!65\517?28C! 2>3!<br />
7A8!85Z>!C585278!?3B!^756!<br />
L3!L!35:786!37/54L8!4L1[!L!67e58!:7;?L8253!A578=!"##$*B!<br />
Q8! 5ZL;?15! :L8! 05! [7/86! 28! >45! 3/??1M! :4L28! 7[!<br />
L/>7;702153! (R1L7=! $dd$=! R4L?3B%=! )*B! Q/-<br />
>7;70215! ?278! 05C283! A2>4! 6532C8=! A42:4! :78323>3! 7[!<br />
>43B! N45![2!515;58>!23! >45!:78:5?>!2>351[=!<br />
>45545!6532C8!7[!>45! \542:15!:L8!05! /35[/11M! 62\2656!<br />
28>7!>45!;L:45!65\517?;58>!7[!>45!0L32:!:4L3323=!<br />
3/0-L335;01M! L86! :7;?7858>! 3?5:2[2:L>278*! L86! >45! ;2:45! 65\517?;58>=! >7! LC2783=! 7[! >45!<br />
\542:15l3!:783>2>/58>!:7;?7858>3*B!N45![2!>A7!515;58>3!7[!<br />
6532C8! >586! >7! 05! /865L]58! 0M! >45! :L2:/1L45!L335;015L]53!:4L!7278B!<br />
G8! >45! [L:5! 7[! 28>5835! :7;?5>2>278=! 47A5\545! :73>3!<br />
L337:2L>56! A2>4! 65\517?28C! 85A! \542:15! ?7>M?53! 4L\5!<br />
28:3! (>45! ;2:7!<br />
C170L1! 3>L86LL]5! [/11! M! [745! 6532C8! 7[! 3/0-<br />
L335;01253!L86![745!:77278!7[!>45!35:786!L86!>4225!;L8/[L:>/!:78>5!>7!>45!?B!<br />
N45! 3>278! 7[! >4535! 3/0-L335;015278B! I7! 328C15!<br />
3/??125M?53! 7[! 3/0-L335;01MB! ^745! A77;7>2\5! 5_/2?;58>!<br />
;L8/[L:>/3! 5[[73! 78! 3>L55;3=! 3?L28C!L86!A2863:278!23!>4L>!>45!3/??1M!7[!<br />
?L2:/1L5;3! 4L3! 05:7;5! :78:58>56!<br />
L;78C3>!D/3>!L!4L86[/1!7[!;L8/[L:>/5:48717C2:L1!:7;?5>2>278!23!/356!>7!5ZL;285!>45!:7;?5>2>278!<br />
05>A558!>45!0/M545!3/??1252CL>278!7[!>423!>M?5!<br />
7[! 0/M5278! A7/16! 15L6! >7! L! 05>>5L8628C! 7[! >45! 6M8L;2:3! 7[! :711L07278! L;78C!<br />
3/??1M!:4L28!?L85528C! L3?5:>! 7[! >45! ?45! [L:>! >4L>! >45!<br />
3/??125!>L]5!28>7!L::7/8>!>45!2845!28:58>2\53![745!<br />
0/M57! 65\517?! L! b0L:]3>7?c! >5:48717CM! A42:4! :L8! 05!<br />
3/03>2>/>56![745!3/??1256!:7;?7858>B!N45!3/??1254!>45!<br />
]87A156C5! 7[! >423! 28>58>! 7[! >45! 0/M53! 3/:4! >4L>! >45!<br />
?3! L278! 7[! 3/03>2>/>5!<br />
>5:48717CM! 0M! >45! 0/M52;5! A458! 3/:4! 2887\L>278!<br />
;L>5!:L8!05!L[[5:>56!0M!Pk9!5[[73B!<br />
N45! ;7651! [7278! 23! 56! >7! L! :L35! 28! >45!<br />
5:787;2:3! 7[! 5Z4L/3>2015!
2.3 VERTICAL COMPETITION AND OUTSOURCING<br />
HANS W. GOTTINGER<br />
,$F! WL83!XB!T7>>28C52>278!L86!K/>37/4L>!23!355]28C!>7!<br />
28\58>! L! 3/03>2>/>5! >5:48717CMB! N45M! L66278! 28! >45! 5:787;2:3! 7[! 5Z4L/3>2015! 5:48717C253! >7! 3/03>2>/>5! [75! 8L>/56! >7! 623:7\57?! >5:48717C253! 87>! 781M! >7!<br />
7\545!?4L>!!L137!<br />
>7! 5:48717C2:L1!<br />
:7;?5>2>278! 23! ?56! 28! H5:>278! "B! G8! H5:>278! ,! >45!<br />
62[[52L1! CL;5! 23! L8L1Me56! L86! >45! SL1"*+)#%<br />
N45! 87>L>2783! L86! 3M;0713! /356! 28! >45! ;7651! L![745!0/M55!>2;5!475C253B!<br />
>+!Q8!283>L8>!7[!>2;5!28!>45!6M8L;2:!CL;5!35>/?B!<br />
/ a (>*+!a/M5278!5[[7!(Pk9!5Z?5862>/2;5*B!<br />
/ H (>*+! H/??125/!>7!05!28:/!>7!05!28:/!28:/4L>! A7/16! 05! L! ?5!<br />
3/03>2>/>5![745!3/??125278B!<br />
!+!L8>!7[!>2;5!L>!A42:4!>45!15\51!7[!2887\L>278!<br />
(04278*!23!*+!>45!?M!>4L>!>45!0/M28C![245!2887\L>278!?!0M!>2;5!>=!>4L>!23!.*!p!^(>*B!<br />
] #$ +![/8:>278!45!3A2>:428C!5B!<br />
Z(>*+!3>L>5!6M8L;2:3!5Z>!7[!0/M54L>!23!0L356!78!<br />
2887\L>278! L86! _/L12>M! ?L5B! Q! 785-! >7-! 785!<br />
:745!0/M57!>45![28L1!:783/;5!15L3>!785!:7;?7858>!23!<br />
! :7;?5>2>278! 28! >45!<br />
[28L1!:783/;5B!<br />
G>! 23! L33556! >4L>! L! 785! q! >7! -! 785! :745!L8L1M323!28!>A7!623>28:>! ALM3B!^2=!<br />
>423!L117A3!5Z?12:2>!5ZL;28L>278!7[!65?58658:5!05>A558!>45!<br />
>A7! :711L0728C! ?L8545! 5\71/>278! 7[! ?7A5/278B! H5:786=! >423! 15863! L8L1M>2:L1!<br />
32;?12[2:L>278! [745! ?! [745!<br />
:73>!7[!?278!23!: H (/ H *!m!#!?5B!H/??1255!7[!<br />
?278! 23! L! [/8:>278! 7[! ?278! 5[[7! / H !L86!23!<br />
L33/;56!>7!05!"/ H B!^72:L1!32;?12:2>M!"!23!L33/;56!>7!<br />
05!5_/L1!>7!$B!N45!5!7[!?278!23!65[2856!L3=!/ H +!(#=#*!<br />
$! [#=! / H ]! N45! 45! 3>L>5! \L45! :78>5Z>! 7[! >45! ;7651! 3?5:2[2:L>278! L86! >45!<br />
/8652783=!>45!2:L1!<br />
[745!0/M545!3/??125278!7[!>45!3/??12545!5!7[!:4L8C5!7[!*!p!-!"!/ H !(Z(>**!p!%!/ H !(Z(>**&!Z(#*!p!Z 7B!<br />
Q3! >45! ?278! 5! 28:45! 278! 5! 7[! >45!<br />
3/??1257!:5!L!3/03>2>/>5!>5:48717CM!<br />
A42:4! 23! L33/;56! >7! 58L015! ?278! 7[! >45! ?!<br />
:7;?7858>!L>!L!/82>!:73>!: a !r!: H B!G8!>45!5\58>!7[!L!3/::533[/1!<br />
:278!7[!3/:4!L!>5:48717CM!>45B! N45! 28\53>;58>! ?712:M! 28! 2887\L>278!<br />
:47358! 0M! a! 23! 6587>56! 0M! / a +! (#=! #*! $! s#=! #*B! N423!<br />
28\53>;58>! 45! 0/M5278! 7[! >45! *B! aM! 28\53>28C! 28!<br />
2887\L>278=!>45!0/M545!>2;5!?L>4![745!0/M5*!23!C2\58!0M=!<br />
] • (>*!p!/ a !(Z(>**=!](#*!p!#B!<br />
N45!7/>:7;5!7[!2887\L>278!28\53>;58>3!23!:7832657!05!<br />
/8:5L28B! N45! >45! Pk9! 28\53>;58>3!<br />
678l>!?545!6L>5!7[!3/::533!7[!2887\L>278B!Q!<br />
8L>/!ALM!>7!278!<br />
23!>7!3/??735!>4L>!>45!283>L8>L857/3!?M!7[!>45!0/M528C=! :7862>278L1! 78! 87>! 4L\28C! 37! [L278! 7[! 2>3! :/!23!L33/;56!>7!<br />
>L]5! L8! 5Z?7858>2L1! [7! 7[! >45!<br />
L::/;/1L>56!]87A156C5!15\51=!](>*B!<br />
^(>*!@![$!q!^(>*]!p!/ a !(Z(>**!!!!!!!!!!!!!!!!!!!!!!!($*!<br />
A45*!23!>45!?M!>4L>!>45!0/M545!2887\L>278!?!0M!>2;5!>&!>4L>!23!.*!p!^(>*B!<br />
`5>! u(/ H *! p! s?(/ H *! q: H (/ H *v! w/ H ! 05! >45! 3/??125L8>L857/3! ?! [/8:>278B! N423! 283>L8>L857/3! ?!<br />
[/8:>278! L:>3! L3! >45! 70D5:>2\5! [/8:>278L1! [745! 3/??12545!?5Z>=!>45!0/M5!:7;?5>2>278B!N45!<br />
0/M53! 28! 2887\L>278! >7! 05:7;5! 2865?58658>! 7[! >45!<br />
171
HANS W. GOTTINGER<br />
2.3 VERTICAL COMPETITION AND OUTSOURCING<br />
! G8>5278L1!f7/! 7[! ?278! 5[[7! : H (/ H *! 23! :78>28/7/31M!<br />
62[[52L015!78!s#=( / H *+!: H (#*!p!#!78!s#=( / H *!L86!:l H (M*!23!L!<br />
:783>L8>B! N45! :73>! 7[! 2887\L>278! 5[[7! : n (/ a *! 23! >A2:5!<br />
:78>28/7/31M!62[[52L015!78!s#=!#*&!: n !(#*!p!:) n !(#*!!p!#&!<br />
:+ n !m!#!78!s#=!#*&!L86!:) n ! (M*! $! #! L3!M!$! #B! N423!<br />
L33/;?>278! 3/CC53>3! >4L>! :73>3! 28:!L137!3/CC53>3!>4L>!e52;L1!L86![745!3>5CM!/ a! A5!:L8!:78[285!7/5!<br />
>7!28>5423!;L!35>>28C=!2>!23!L33/;56!>4L>!>45!?!58D7M3!<br />
L! \5278!<br />
5! 7[! >45! 3/??12545! _/L8>2>M! 65;L8656! ?5!<br />
>2;5!28!>45!;LB!N45!28\5278!74545!?278!7[!/82>3!?2;5B!<br />
?(/ H (Z(>***pL%0"/ H (>*pLq0/ H (>*&!Lm: H m#=!0m#!LL8>3!("*!<br />
N45!\L!7[!?278!23!L33/;56!>7!05!1285L278!7[!?278!5[[7!/ H B!<br />
: H !=!: $ ![/ H]=!: $ !m!#!23!L!?M!:783>L8>!!!!!!!!!!!(,*!<br />
!"#$%&'()*7("(8()9*:%(0,2#'("*&"0*:%'2$6*<br />
N45! 3>5C253! / H (B*! L86! / a (B*! :783>2>/>5! L8! 5_/212045!<br />
?278!5[[7!/ H (B*!;LZ2;2e53!>45!3/??12545! 2887\L>278! 5[[7! / a (B*! ;LZ2;2e53! >45! 0/M55C253B!N45![7117A28C!L33/;?>2783!L45! ?!<br />
[/8:>278!L86!>45!:73>!7[!5[[7![/8:>278B!<br />
(2*! N45! 28\5278! ?(/ H *! 23! >A2:5!<br />
:78>28/7/31M!62[[52L015!78!s#=( / H ]+!<br />
A45!7[!5[[7![/8:>278![745!0/M57!05!<br />
3>1M!:78\5Z!L86!28:278!5[[7!/ a B!<br />
: n !=!: " ![/ a] " = ! : " !m!#!23!L!?M!:783>L8>!!!!!!!!!!('*!<br />
I7>5!>4L>!>423!?45!3L;5!3>L>5!6M8L;2:3![745!<br />
0/M545!3/??1252\5![/8:>278L1!7[!3/??12545!5Z?5:>56!?!/8>21!>45!L8>!7[!>2;5!!=!<br />
f H (Z*!=!O /a(B*! # ! !5 - !s?(/ H (Z(>***!q: H (/ H *vw/ H !Z(>**]6>!<br />
p!O /a(B*! # ! !5 - ![L!-!0/ H (Z(>**!%!: $ (/ H *(Z(>**/ H (!Z(>**]!6>!!!(%* !<br />
I7>5!>4L>!>45!3/??125![7!m!!!(328:5!<br />
65;L86!\L823453*B!N45!70D5:>2\5![/8:>278L1!7[![245!<br />
5Z?5:>56!CL28![278!5[[73=!C2\58!0M!<br />
f a (Z=!/ a *!=!O /a(B* !! # ! !5 - s&(/ H (Z(>***!q!: n (/ a *(Z(>***]6>!+!! ! # !5 - !' a !(/ a *6>!<br />
p!O /a(B*! # ! !5 - !s&(/ H (Z(>***!q: " (/ a *(Z(>***] " 6>!+!! ! # !5 - !' a !(/ a *6>!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!()*!<br />
&(/ H (Z(>***!=!! # /H !s?(M*6M!q/ H! (Z(>**?(/ H! *!=!! # /H ![L!-!0M]6M!%/ H (Z(>**![L!-!0/ H (Z(>**]!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!(F*!<br />
23!>45!3/
2.3 VERTICAL COMPETITION AND OUTSOURCING<br />
HANS W. GOTTINGER<br />
,$d! WL83!XB!T7>>28C52>278!L86!K/>37/***!q: H (/ H *v/ H Z(>**]6>!<br />
=!O /a(B*! # ! !5 - ![L!-!0/ H (Z(>**!%!: $ (/ H *(Z(>**]/ H (!Z(>**6>!<br />
3/0D5:>!>7!Z • (>*!p!%!/ H !(Z(>**&!Z(#*!p!Z 7 !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!(E*!<br />
f a (Z=/ a *!=!O /a(B*! # ! !5 - !s&(/ H (Z(>***q: n (/ a *(Z(>***]6>!+!! ! # !5 - !' a !(/ a *6>!<br />
N45!62[[52L1!CL;5!65[2856!0M!(E*!L86!(d*!05178C3!>7!>45!<br />
:1L33! 7[! ?25:5A235! 65>52:! CL;53! (97:]85! L1B=!<br />
"###*+! N45! CL;5! 4L3! >A7! ;7653! 781M+! ;765! #! 23! L:>2\5!<br />
05[7278!L86!;765!$!05:7;53!<br />
L:>2\5!2[![245!3/03>2>/>5!>5:48717CMB!<br />
N4/3=! >45! ;73>! 785! 3A2>:4! 7[! ;765B! N45!<br />
3A2>:428C!>2;5!23!>45!278!7[!!!23!^=!L3!C2\58!28!5_/L>278!($*B!<br />
C'%7#"1:$+$%)A%*6-%D)2-1%<br />
N45!;7651!23!L8!5Z>583278!7[!L!?>28C5***] " 6>!+!! ! # !5 - !' a !(/ a *6>!<br />
3/0D5:>!>7!Z • (>*!p!%!/ H !(Z(>**&!Z(#*!p!Z 7 !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!(d*!<br />
3>L>278L278L! n H! (4=!Z*!23!L!<br />
;L??28C!n H +!S!,(#=!#*!$!s#=!( / H v=!A4545!<br />
35>! 7[! ;7653B! a/M5L>278L278! L??2783! 7[! >45! 3/??1252\51M![745!65[282>278!L86!>4577! ?25:5A235! 65>52:! 62[[52L1! CL;53!<br />
?45!Q??5862Z=!>4535!\L1/5![/8:>2783!L7!05!07/8656!L86!:78>28/7/31M!62[[52L015!L86![74L>! 3L>23[M! >45! WL;21>78-fL:702-a511;L8! (Wfa*!<br />
5_/L>2783B!<br />
3! 7[! / H-! s#=! ( / H v!L86!/ a-! s#=#*! >4L>!<br />
;LZ2;2e5!>45!2\5!\L1/5![/8:>2783!7[!>45!3/??12545!0/M5278!($#*!L86!($"*!L56!L3!0 H (Z*!L86!<br />
HANS W. GOTTINGER<br />
2.3 VERTICAL COMPETITION AND OUTSOURCING<br />
! G8>5278L1!f7/45! :73>! 7[! 5[[7! [/8:>278! [2783!("*!L86!5_/L>278!('*=!2\51MB!<br />
T H (e*!=!;LZ!{L!%!0/ H !%!: $ !/ H !]!/ H !%!/ H !e.#!1!/ H!1!( / H}!!($d*!<br />
T a (e*!=!;LZ!{e/ a !%!: " (/ a * " .#!1!/ a !}!!!!!!!!!!!!!("#*!<br />
n H (#=Z**!=!L5!>45!?732>2\5!5_/212045!3/??125458!!<br />
! "# $<br />
%&"'&<br />
"&#()*+$' , - . *%/ 0 $)' . -1&1% . +' !!!!!!!!!!!!!!!!!!!("'*!<br />
! 2# $ 134 & 51$1%6 7 111&1 "&#()*+$' ,- . *%/ 0 $)' . -1&1% . +<br />
!!!!!!!!("%*!<br />
"&*+$' , -<br />
;5$(%$3*=6%G[!n H 2 !L86!n a 2 !6587>5!>45!?732>2\5!5_/21202783!7[!>45!3/??125458!<br />
8 9 # $ :;! < 9 *=>?-1<br />
3*@$4 A -1<br />
8 B # *=> ?- $ 6C 1&16 C *=>?-<br />
$ 1:<br />
3*@$4 A - !!!!!!!!!!!!!!!!!!!!!!!!!!!(")*!<br />
34 3<br />
$ "D5*@E4 A -3 F6 C E54 3 G;: 3@<br />
3"3*@$4 A -"4 3<br />
!!!!!!!!!!("F*!<br />
;5$(%$3*>6!Q>!5_/212045!!>2;5!>!<br />
:L8!05!5Z?$$+)#%<br />
?@/$%A&#'("* ! 5_/212045! 3/??125278!5B!<br />
N45! 7035278! :L8! 05! 5Z?1L2856! 0M! 5Z?1745!<br />
;787?71M!?278!5!7[!>45!3/??12545!C2\58!35>>28CB!<br />
N45!;787?71M!?!7[!>45!3/??12545![2!77!>45!;787?71M!?278!5!C2\58!0M!L!/ H; !=!L/!"(0+: $ *B!<br />
N45!3/??1257!05!L!;787?71MB!G8!>45!C2\58! 35>/?! >45! 3/??125278!5!>7!CL28!L3!<br />
;/:4!7[!;787?71M!?3!L3!?7332015!05[7!7[!>45!;LB!Q8!211/3>278!7[!3/:4!L!;787?71M!?278!23!<br />
?
2.3 VERTICAL COMPETITION AND OUTSOURCING<br />
HANS W. GOTTINGER<br />
,"$! WL83!XB!T7>>28C52>278!L86!K/>37/45!3/??1253!L!85CL>2\5!5[[5:>!78!>45!<br />
0/M5278!5B!<br />
N45! 7035278! 3>L>56! 28! "(2*! 23! 28>/2>2\5! L86! :L8! 05!<br />
703545! 5Z?5! 0M! >45! 0/M556B! X2>4! 37;5! L1C502783! >45!<br />
5_/2120278!5!7[!>45!0/M545! L07\5! 5Z?! :L8! 05! 87>56! >4L>! >45!<br />
3/??1252\5!5[[5:>!<br />
78! >45! 0/M5278! 5[[73B! K035278! $! L86! "!<br />
:4L545! 8L>/45!C2\58!35>>28C!>453!87!<br />
;7>2\L>278![745!>A7!?1LM57!:711L075!L86!5L:4!L:>3!28!<br />
2>3!053>!28>5B!N45!3/??12545!;787?71M!<br />
?7A545!0/M528/53!28\53>;58>3!28!2887\L>278!>7!<br />
:5!3/03>2>/>5!>5:48717CMB!<br />
W7A5\52783B! I7>5=! >4L>! 28! 5_/L>278! ("d*! A5! 4L>!A2>4!L!67A8AL278B!N45!28\5
HANS W. GOTTINGER<br />
2.3 VERTICAL COMPETITION AND OUTSOURCING<br />
! G8>5278L1!f7/7!<br />
Q86!<br />
( 63 %<br />
7 8 "9 8<br />
":#; $ 8 +%? @ $>28C52>278j=!<br />
!"#$%"&#'("&)* +(,%"&)* (-* .,/'"$//* &"0! 12("(3'2/* 4$/$&%25!<br />
'("*=!)F-F$B!<br />
sdv! T7>>28C5
2.4 SUPPLY-CHAIN COOPETITION<br />
HANS W. GOTTINGER<br />
!"#$%"&#'("&)*+(,%"&)*(-*.,/'"$//*&"0*12("(3'2/*4$/$&%25*<br />
"#$%&!'(")*!+,-,$!<br />
./0123456!781285!9:;!"#$%!(4??@*AABBBC3
HANS W. GOTTINGER<br />
2.4 SUPPLY-CHAIN COOPETITION<br />
+G! V:83!WC!S7??28D5;*! ! I/@@1M-R4:28!R77@5?2?278! !<br />
!<br />
:8:1M_53! ?45! /`!:=!28Y!{?!>!#*!g? /I>/`!=!M!}CN45!Y2;3?!42??28D!?2E5!?7!?45!<br />
@728?!M!/865;!?45!3@5/`!C)!<br />
@(_)! *b@;2
2.4 SUPPLY-CHAIN COOPETITION<br />
HANS W. GOTTINGER<br />
! H8?5;8:?278:1!e7/;8:1!7Y!`/328533!:86!O@CG#]! :86! /?212_28D! ?45! ;53/1?3!<br />
Y;7E!(')!:86!(%)!D2Z53 !<br />
6!g? /I>/`!=!g? /I>/`[E(/ I (?)>!(/`(?))6?!+!/ I (?)# I !6W I (?)!!!/`(?)#`!<br />
6W`(?)]!!!!!!!!!!!!(+)!<br />
B45;5!?45!Y/8
HANS W. GOTTINGER<br />
2.4 SUPPLY-CHAIN COOPETITION<br />
,#! V:83!WC!S7??28D5;*! ! I/@@1M-R4:28!R77@5?2?278! !<br />
!<br />
c / (_)!=!3/@ /I !28Y /`!O _ !(5 !,' 0 /I>/`)!=!28Y /`!3/@ /I !O _ !(5 !,' 0 /I>/`)!>!Y7;!<br />
_!! !f![%`"!($!!!- " )!±!.p!]!/!"![%`"!($!o!- " !!!"!&!-)!o!"!,!($!!!& " )]!($")!<br />
:86!?45!:337!<br />
/`/(_)!=![%`/#`]![($!!&-!)0 + !!!$]!/!(!($!!!& " )(0 + ) " !!!$)!]!($=)!<br />
(")!R:35!`!<br />
N457;5E!"C!HY!?45!65D;55!7Y!E:;[5?!:6Z:8?:D5!7Y!:!3/@@125;!<br />
7Z5;!:!0/M5;!-!3:?23Y253!?45!!-!>!- + !B45;5!- ! !>- + !<br />
:;5!:3!65Y2856!28!(a)>!?458!?45!Z:1/5!7Y!?45!623$$+)#%<br />
W5!
2.4 SUPPLY-CHAIN COOPETITION<br />
HANS W. GOTTINGER<br />
! H8?5;8:?278:1!e7/;8:1!7Y!`/328533!:86!O!:!3/@@125;!B7/16!D:28!0M!E:[28D!<br />
28Z53?E58?3! 28! 2887Z:?278! ?4:?! :;5! 2865@58658?! 7Y! ?45!<br />
28Z53?E58?3!E:65!0M!?45!0/M5;C!c/;?45;E7;5>!:!3/@@125;!B7/16!<br />
:137! D:28! 0M! E:[28D! 28Z53?E58?3! 28! 2887Z:?2783! ?4:?! :;5!<br />
@7332015!3/03?2?/?53!7Y!?45!0/M5;r3!@;76/
HANS W. GOTTINGER<br />
2.4 SUPPLY-CHAIN COOPETITION<br />
182
2.5 A TALE ON CHINESE INNOVATION AND COMPETITION<br />
HANS W. GOTTINGER<br />
A Tale <strong>on</strong> Chinese Innovati<strong>on</strong> <strong>and</strong> Competiti<strong>on</strong>: A Case Study<br />
of the Chinese Telecommunicati<strong>on</strong>s Equipment Industry<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
STRATEC Munich Germany<br />
www. stratec-c<strong>on</strong>.net<br />
Extended Abstract.<br />
We examine the four largest Chinese suppliers of stored program c<strong>on</strong>trol<br />
(SPC) switches: Datang Telecom <strong>Technology</strong> (DTT), the Great Drag<strong>on</strong><br />
Informati<strong>on</strong> <strong>Technology</strong> (GDT), Huawei Technologies (Huawei), <strong>and</strong><br />
Zh<strong>on</strong>xing Telecommunicati<strong>on</strong> Equipment (ZTE). The tale shows how<br />
these four Chinese suppliers competed against the two largest SPC switch<br />
manufacturers that had foreign joint venture partners :Shanghai Bell an<br />
Alcatel joint venture, <strong>and</strong> Beijing Internati<strong>on</strong>al Switching System (BISC),<br />
a Siemens joint venture , <strong>and</strong> further how they competed globally against<br />
well established suppliers of telecommunicati<strong>on</strong>s equipment (TE)such as<br />
Alcatel, Cisco, Ericss<strong>on</strong>, Fuijitsu, NEC, Nokia, Nortel, <strong>and</strong> Siemens.<br />
First, the fast growth in the period after 1985 of two suppliers, Huawei<br />
<strong>and</strong> ZTE, requires appropriate attenti<strong>on</strong>. Their success threatens well<br />
established incumbents’ plans to dominate the global <strong>and</strong> China markets<br />
for TE. ZTE started operati<strong>on</strong>s in 1985, <strong>and</strong> Huawei in 1988. In slightly<br />
over ten years, Huawei became the number <strong>on</strong>e supplier of TE in China.<br />
In 1998. Huawei’s annual revenues exceeded those of the top two TE<br />
suppliers that had foreign joint venture partners: Shanghai Bell <strong>and</strong> BISC,<br />
<strong>and</strong> since then had the gap growing. Still, in the mid eighties, the<br />
possibility that any of the four Chinese suppliers could pose a serious<br />
threat to established global suppliers seemed very improbable. Today,<br />
Chinese suppliers compete aggressively against these well established<br />
global suppliers <strong>and</strong> their joint ventures in China.<br />
For example, reputable market share tracking firms ranked Huawei the<br />
number <strong>on</strong>e supplier in the global market for new extended switching<br />
equipment in 2003, the number <strong>on</strong>e supplier in the global market for new<br />
generati<strong>on</strong> networks in 2004, the number two supplier in the global<br />
market for digital subscriber line (DSL)access multiplexers in 2003, the<br />
number three supplier in the global market for l<strong>on</strong>g distance wavelength<br />
divisi<strong>on</strong> multiplexers, <strong>and</strong> the number four supplier in the global market<br />
for optical transmissi<strong>on</strong>. Huawei <strong>and</strong> ZTE have been ranked the number<br />
183
HANS W. GOTTINGER<br />
2.5 A TALE ON CHINESE INNOVATION AND COMPETITION<br />
three <strong>and</strong> eight suppliers in the global market for integrated access<br />
networks. As the Wall Street Journal reported repeatedly in 2003 to 2005,<br />
executives of North American <strong>and</strong> European vendors of TE have become<br />
increasingly c<strong>on</strong>cerned about the head-to-head competiti<strong>on</strong> from Chinese<br />
suppliers.<br />
On the basis of previous models of competitive racing (<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, 2006,<br />
2009) we trace empirically the various stages of competitive strength in<br />
this strategic industry.<br />
In the first stage we observe that the four startups targeted the basic need<br />
for infrastructural development in telecommunicati<strong>on</strong>s in Western <strong>and</strong><br />
rural China in supplying low cost telecommunicati<strong>on</strong>s gear to those areas<br />
which were less lucrative for foreign vendors <strong>and</strong> joint ventures <strong>and</strong> were<br />
heavily encouraged by the Chinese nati<strong>on</strong>al government in a sort of<br />
infant industry protecti<strong>on</strong>. This kind of asymmetic competiti<strong>on</strong> separated<br />
the startups from the established players in the Chinese market, call it the<br />
separati<strong>on</strong> stage.<br />
In the course of this stage the’four horsemen’ underwent technological<br />
learning either through indigenous innovati<strong>on</strong> or imitati<strong>on</strong> of some sort,<br />
even industrial espi<strong>on</strong>age in China,therefore gaining competitive strength<br />
<strong>and</strong> competing against foreigners <strong>on</strong> large scale projects in the Chinese<br />
market. This is termed the c<strong>on</strong>vergence stage. When asymmetric<br />
competiti<strong>on</strong> turns symmetric we observe competitive c<strong>on</strong>vergence , in<br />
which each technology’s development is directed at exp<strong>and</strong>ing its appeal<br />
not <strong>on</strong>ly in its own home market but in its rival’s as well.<br />
While the Chinese companies with the implicit support of the Chinese<br />
government c<strong>on</strong>tinued to gain market share against foreign competitors<br />
<strong>and</strong> as their technological learning adavanced product quality at lower<br />
cost they exp<strong>and</strong>ed in actively seeking to bid successfully for<br />
telecommunicati<strong>on</strong>s projects in developing <strong>and</strong> emerging ec<strong>on</strong>omies<br />
where they gained further strength by competing <strong>on</strong> given product quality<br />
<strong>and</strong> lower prices. This is the globalizati<strong>on</strong> stage.<br />
However, whether Chinese companies keep <strong>on</strong> growing outside their<br />
home market will largely depend <strong>on</strong> whether they turn into genuine<br />
sustainable innovati<strong>on</strong> leaders rather than followers. So far they have not<br />
gained a notable footage in advanced markets for smart networks.<br />
184
2.5 A TALE ON CHINESE INNOVATION AND COMPETITION<br />
HANS W. GOTTINGER<br />
Disruptive Innovati<strong>on</strong><br />
If we c<strong>on</strong>sider the competitive positi<strong>on</strong>ing of high technology firms as a<br />
technology race in which falling behind, getting ahead <strong>and</strong> catching-up in<br />
industry leadership is the name of the game. We may come across<br />
specific situati<strong>on</strong>s that would be c<strong>on</strong>nected with ’disruptive innovati<strong>on</strong>s’<br />
that could lead to a dramatic paradigm shift of that race. The management<br />
of disruptive technologies originated with the substantial works by C.<br />
Christensen et al.(2004) <strong>and</strong> C. Christensen (2005).<br />
Disruptive innovati<strong>on</strong>s introduce a new kind of product or service that is<br />
actually worse initially, as judged by the performance metrics that main<br />
stream customers value. In terms of the Chinese market where firms<br />
came up with simpler appropriate performance level of network gear at a<br />
much lower price for use in rural China those could be identified as<br />
disruptive. They also c<strong>on</strong>tributed to a market separati<strong>on</strong> that<br />
distinguished two different tiers of market segmentati<strong>on</strong> applying to rural<br />
<strong>and</strong> urbanized areas. In an industrial organizati<strong>on</strong> c<strong>on</strong>text incumbents<br />
have a high probability of beating entrant attackers when the competiti<strong>on</strong><br />
is about incremental innovati<strong>on</strong>s, but almost always lose to attackers<br />
armed with disruptive innovati<strong>on</strong>s (Christensen et al, 2004).. At the<br />
bottom line a cumulati<strong>on</strong> of incremental innovati<strong>on</strong>s <strong>on</strong> a particular<br />
design may end up in a disruptive innovati<strong>on</strong> as will <strong>on</strong> the upper end<br />
technologically radical innovati<strong>on</strong>s . This may lead to market situati<strong>on</strong>s<br />
<strong>and</strong> revenue streams of ’winner-take-all’ or ’winner-take-most’.<br />
We may categorize two aspects of disruptive innovati<strong>on</strong>s that relate to<br />
(1) new markets <strong>and</strong> (2) quality characteristics.<br />
(1) Products offered are too expensive, too complicated, too difficult too<br />
maintain.<br />
(2) Product quality characteristics are appropriate,’ good enough’ with<br />
significantly lower pricing. New entrants compete profitably , learn <strong>and</strong><br />
invest in R&D while pricing at deep discounts.<br />
Competitive Phases<br />
In adapting Christensen’s theory of innovati<strong>on</strong> to the Chinese telecom<br />
industry we may distinguish clearly competitive phases such as<br />
separati<strong>on</strong>( ’isolati<strong>on</strong>’ as Christensen’s term), c<strong>on</strong>vergence <strong>and</strong><br />
disrupti<strong>on</strong>. These phases seem to be more technologically motivated<br />
through innovati<strong>on</strong> but may be supported by strategic directi<strong>on</strong> or other<br />
factors.<br />
In a competitive separati<strong>on</strong> (isolati<strong>on</strong>) regime technologies do not interact<br />
(in terms of being replaceable) in the course of their development.<br />
185
HANS W. GOTTINGER<br />
2.5 A TALE ON CHINESE INNOVATION AND COMPETITION<br />
SUMMARIZING REMARKS<br />
This case study is of interest as a general attempt in identifying Chinese<br />
efforts <strong>on</strong> innovati<strong>on</strong> at the interface of culture, politics, technology <strong>and</strong><br />
industry. The subject is so complex that it is unlikely to find a c<strong>on</strong>sistent<br />
interrelated interpretati<strong>on</strong> of innovati<strong>on</strong> with 'Chinese Characteristics'.<br />
(The well known Japanese growth ec<strong>on</strong>omist Michio Morishima in the<br />
1980s did this for Japan (CUP) <strong>and</strong> ended up with a tax<strong>on</strong>omy which yet<br />
failed to provide a c<strong>on</strong>sistent interpretati<strong>on</strong> of Japanese innovati<strong>on</strong>) In my<br />
brief comments I limit myself <strong>on</strong> a case study based Chinese high<br />
technology industry<br />
To substantiate some of the claims <strong>on</strong> industrial innovati<strong>on</strong> <strong>on</strong>e needs a<br />
set of industry specific case studies such as the <strong>on</strong>e submitted (A Tale of<br />
Innovati<strong>on</strong> <strong>and</strong> Competiti<strong>on</strong> in the Chinese Telecom Industry) . A few<br />
remarks are in order.<br />
Chinese advanced technology <strong>and</strong> science based industries tend to be<br />
biased toward" product innovati<strong>on</strong> through commercializati<strong>on</strong>" --- quick<br />
market access <strong>and</strong> then incrementally improving product performance<br />
(quality) in subsequent releases.Indigeneous innovati<strong>on</strong> could also<br />
involve 'rapid prototyping' in intenti<strong>on</strong>ally or inadvertently bypassing<br />
IPRs in favour of quick market introducti<strong>on</strong>. Indigeneous innovati<strong>on</strong><br />
could also be a camouflage word for nurturing Chinese startups (where<br />
state directed subsidizati<strong>on</strong> acts as the equivalence of private sector<br />
venture capital or angel investment), <strong>and</strong> shielding them from foreign<br />
competiti<strong>on</strong> (at least in the domestic market for a while) in support of<br />
'infant industry' protecti<strong>on</strong> <strong>and</strong> expansi<strong>on</strong>. This is coupled with state<br />
sp<strong>on</strong>sored private entrepreneurship with encouragement of<br />
Schumpeterian entrepreneurship as examples in the Chinese network<br />
equipment industry clearly show. Competiti<strong>on</strong> in the<br />
Chinese domestic market am<strong>on</strong>g indigeneous firms could be fierce <strong>and</strong><br />
adopts characteristics of a 'technology race' but since the domestic<br />
market is so large <strong>and</strong> with rising incomes<br />
dem<strong>and</strong> growing in leaps there sometimes is a lack of incentives to go for<br />
global markets (with some notable excepti<strong>on</strong>s outlined by industrial <strong>and</strong><br />
development policies).<br />
Chinese innovati<strong>on</strong> resembles a 'learning-by-doing' (Arrow) effect as the<br />
Japanese have d<strong>on</strong>e over several decades whereas the Chinese have<br />
accelerated this process by choice <strong>and</strong> through competiti<strong>on</strong>. Indigeneous<br />
innovati<strong>on</strong> through domestic competiti<strong>on</strong> may limit market risk <strong>and</strong><br />
therefore protect the company against total financial failure (bankruptcy).<br />
By industrial policy targeted companies can experiment <strong>on</strong> product<br />
development without severe downside risk. You d<strong>on</strong>'t bet the company if<br />
the product fails in the marketplace, therefore you can come up with<br />
186
2.5 A TALE ON CHINESE INNOVATION AND COMPETITION<br />
HANS W. GOTTINGER<br />
incremental changes, whether true innovati<strong>on</strong> or not, <strong>and</strong> let the market<br />
decide in the Chinese seller's market without fearing drastic financial<br />
penalties. No Western companies can take these (financial) risks in their<br />
domestic markets besides being blocked by Chinese entry<br />
barriers(reinforced by Chinese regulati<strong>on</strong>)<br />
Christensen,C., Anth<strong>on</strong>y,S.D. <strong>and</strong> Roth,E.A. (2004),’Seeing what’s next:<br />
using the theories of innovati<strong>on</strong> to predict industry change’,<br />
Bost<strong>on</strong>:Harvard Business School Press<br />
Christensen,C. (2005), The Innovator’s Dilemma, New York,<br />
HarperCollins<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>,H.W.(2006), Innovati<strong>on</strong>, <strong>Technology</strong> <strong>and</strong> Hypercompetiti<strong>on</strong>,<br />
Routledge, L<strong>on</strong>d<strong>on</strong><br />
---- (2009), Strategic Ec<strong>on</strong>omics of Network Industries, NovaScience,<br />
New York<br />
---------------------------------------------------<br />
187
HANS W. GOTTINGER<br />
2.5 A TALE ON CHINESE INNOVATION AND COMPETITION<br />
188
2.6 CHINESE INNOVATION AND COMPETITION<br />
HANS W. GOTTINGER<br />
!<br />
!"#$%"&#'("&)*+,-'"$--*&".*/0("(1'0-*2$-$&%03*<br />
!"#$%&!'!()&!!*$!& &<br />
+,-./0123&45./52&6,52&#"7&!"#$&'1889)::;;;>?4,9O!DP-/5>4*2Q
HANS W. GOTTINGER<br />
2.6 CHINESE INNOVATION AND COMPETITION<br />
& T582?5M8/45M.&^,0/5200&M53&I=454@/=0&G202M?=1&!"#$%&!'!()&!!*$!& !$&<br />
&<br />
@,.8/9.2\2?0& /5& !""$7& 812& X,@-2?& F1?22& 0,99./2?& /5& 812& U4?2/>5& =4@928/84?0& M53& M0& 812/?& 82=154.4>/=M.& .2M?5/5>&<br />
>.4-M.& @M?Z28& U4?& .45>& 3/08M5=2& ;MQ2.25>81& 3/Q/0/45& M3QM5=23& 9?43,=8& V,M./8N& M8& .4;2?& =408& 812N& 2\9M5323& /5&<br />
@,.8/9.2\2?07& M53& 812& X,@-2?& b4,?& 0,99./2?& /5& 812& >.4-M.& M=8/Q2.N&022Z/5>&84&-/3&0,==200U,..N&U4?&82.2=4@@,5/=M8/450&<br />
@M?Z28&U4?&498/=M.&8?M50@/00/45&'I=454@/087&E298/5>&2=454@/20&;12?2&812N&<br />
`FI[0& 32Q2.49@258& 4Q2?M..& ;M0& ?2.M8/Q2.N& @4?2& @43208& >M/523& U,?812?& 08?25>81& -N& =4@928/5>& 45& >/Q25& 9?43,=8&<br />
'I=454@/087&d=8&<br />
I/>18& 0,99./2?0& /5& 812& >.4-M.& @M?Z28& U4?& /582>?M823& M==200& 0,08M/5M-.N&4,80/32&812/?& 14@2& @M?Z28& ;/..&.M?>2.N&329253&<br />
528;4?Z025,/52&/554QM8/45&.2M32?0&?M812?&<br />
!"##&R,M;2/&;M0&./0823&X,@-2?&d52&M53&`FI&M@45>&812& 81M5&U4..4;2?0M/523&M&548M-.2&U448M>2&<br />
849&F25& 812/?& @M?Z28&<br />
9M?8/=,.M?7& 1M0& -225& 0/5>.23& 4,8& M0& 812& 3M?Z& 14?02& /5& 812& 940/8/45&/5&M&09M8/M.&i-4M?3&>M@2[N[&7&/8&=4,.3&-2&<br />
'W2?N& ?2=258.N7& -2=M,02& 4U& M..2>23& 02=,?/8N& =45=2?50& 7& @M8=123& M53& 8?M=23& 84& E=488& ^44?@M5& '#YBY(& 45& @/./8M?N&<br />
R,M;2/& 1M0& -225& -M??23& U?4@& -/33/5>& 45& 9,-./=& 08?M82>N& M53& /5Q/820& M& @M8=1/5>& /5& 82=154.4>N& ?M=20& M53&<br />
/5U?M08?,=8,?M.& 9?4A2=80& /5& H,08?M./M& M53& U?4@& =4@9M5N& -,0/5200&;M?0&'L488/5>2?7&!""B7&!""Y(N[&'XMZM/&M53&FM5MZM7&<br />
#YY!%&L488/5>2?7&!""B(& &<br />
!"#"(& -,8& @4?2& 092=/U/=M..N& 812N& -481& M?2& 2@-23323& /5&<br />
T5&Q/2;&4U&32Q2.49@258&4U&3N5M@/=&=4@928/8/45&;2&8?M=2& S2/[=1/20&4U&=4@928/8/Q2&08?25>81&/5&81/0&<br />
08?M82>/=&/53,08?N2& ;2& 4-02?Q2& 81M8& 812& 81?22& 08M?8,90&<br />
?'%9-:%@)#5-A*$%)B%@)0A-*+*+)#%B).%<br />
8M?>2823& 812& -M0/=& 5223& U4?& /5U?M08?,=8,?M.& 32Q2.49@258& /5& C-*;).D%=#2>$*.+-$%<br />
82.2=4@@,5/=M8/450&/5&S2082?5&M53&?,?M.&J1/5M&/5&0,99.N/5>&<br />
.4;& =408& 82.2=4@@,5/=M8/450& >2M?& 84& 81402& M?2M0& ;1/=1&<br />
d52&4U&4U&812&@408&/@94?8M58&?20,.80&U?4@&812&/53,08?/M.&<br />
?2V,/?23& .200& 0491/08/=M823& 2V,/9@258& M53& ;2?2& .200&<br />
4?>M5/]M8/45& ./82?M8,?2& /0& 812& 0,>>208/45& 4U& M5& M5M.N8/=M.&<br />
.,=?M8/Q2&U4?&U4?2/>5&Q2534?0&M53&A4/58&Q258,?20&&<br />
3/08/5=8/45& -28;225& 9?43,=8& @M?Z28& =4@928/8/45& M53&<br />
?2>/45M.& 3/09M?/8N7& 258?N& /5& 81402& @M?Z280& ;2?2& 12MQ/.N&<br />
=4@928/8/45&/5&/554QM8/45M?3& ;2& U4..4;& @MA4?&<br />
25=4,?M>23& -N& 812& J1/5202& 5M8/45M.& >4Q2?5@258& M.04& /5&<br />
=45=2980&/5&WM?/M5&M53&E1M9/?4&'#YYD(&M53&6&9?43,=8/45&9?4=20020&M53&=4080(254,0& /554QM8/45& 81?4,>1& Gg_7812?2U4?2& >M/5/5>&<br />
M?2& =450/32?23& ?2.M8/Q2.N& [08M8/=[7& 9?43,=8& @M?Z28&<br />
=4@928/8/Q2& 08?25>81& M53& /5=?2M0/5>.N& =4@928/5>& M>M/508&<br />
=4@928/8/45& /0& 812& @M/5& =1M552.& 81?4,>1& ;1/=1& 812&<br />
U4?2/>52?0& 45& .M?>2& 0=M.2& 9?4A2=80& /5& 812& 258/?2& J1/5202&<br />
=4@928/8/Q2&9?4=200&8MZ20&9.M=2225=27& /5& ;1/=1& 2M=1& 82=154.4>N[0&<br />
9?43,=8/45& 9?4=20020%& 81/0& =4@928/8/45& /0& 4U825& M004=/M823&<br />
32Q2.49@258&/0&3/?2=823&M8&2\9M53/5>&/80&M992M.&548&45.N&/5&<br />
;/81&812&/32M0&4U&M&=4@928/8/Q2&81?2M8&U?4@&/554QM8/45&<br />
S1/.2&812&J1/5202&=4@9M5/20&;/81&812&8M=/8&0,994?8&4U&812&<br />
',5=44?3/5M823(& /554QM8/Q2& M=8/Q/8N& 81M8& =M5& -2& /3258/U/23&<br />
J1/5202&>4Q2?5@258&=458/5,23&84&>M/5&@M?Z28&01M?2&M>M/508&<br />
;/81&812&<br />
! "#!<br />
190
2.6 CHINESE INNOVATION AND COMPETITION<br />
HANS W. GOTTINGER<br />
!C&<br />
RM50&S2?)&J1/5202&T554QM8/45&M53&J4@928/8/45)H5&T53,08?N&JM02&4U&812&L.4-M.& &<br />
F2.2=4@@,5/=M8/450&IV,/9@258&T53,08?N&<br />
9?4092=8&4U&/58?43,=8/45&4U&9?43,=80&4?&02?Q/=20&81M8&;/..&<br />
=4@9282& /5& 812& U,8,?21& M& 9?4=200& 4U&<br />
.2M?5/5>*-N*34/5>& /5&<br />
/554QM8/457&;2&2\92=8&812/?&32=/0/450&?2>M?3/5>&/554QM8/Q2&<br />
08?M82>N&84&-2&/5U.,25=23&<br />
0/>5/U/=M58.N& -N& 812& /554QM8/Q2& 08?M82>/20& 2@9.4N23& -N&<br />
?/QM.&U/?@0&j&UM/.,?2&84&/554QM82&0,==200U,..N&;/..&.2M3&548&<br />
A,08& 84& .400& 4U& 948258/M.& 9?4U/8& -,8& M.04& ?/0Z& UM../5>& -21/53&<br />
/554QM8/Q2&?/QM.0N&<br />
4U&[08M8/=[&M53&[3N5M@/=[&=4@928/8/45&812&0M@2&./520&<br />
M0& M-4Q27& 812& 08M8/=& 3/@250/45& 4U& =4@928/8/45& /0& ,0,M..N&<br />
?2.M823&84&9?/=2&=4@928/8/457&;1/=1&8MZ20&9.M=2&>/Q25&812&028&<br />
4U&9?43,=80&4?&02?Q/=20&81M8&M?2&@M?Z2823&-N&U/?@0,/3M5=2&U4?&=4@928/8/45&94./=N&9?M=8/=2&/554QM8/45&M53&@M?Z28&3N5M@/=0&/584&<br />
M==4,58& U4?& 812& 9,?94020& 4U& =4@928/8/45& 94./=N1& M5M.N0/0& 4U& =4@928/8/45& ;/81/5& M& ?2.2QM58& @M?Z282&/5&812&=45=2980&4U&=4@928/8/45&<br />
,023&U4?&M00200@258,/3M5=2&32Q2.4923&<br />
/5& 0,-02V,258& 02=8/450& /0& 84& ?2.M82& 81202& 3/@250/450& 4U&<br />
=4@928/8/45&84&9?M=8/=M.&=45=2980&,023&/5&=4@928/8/45&94./=N&<br />
M00200@2587&/5&9M?8/=,.M?&84&@M?Z28&32U/5/8/45&M5M.N0/0M5/]M8/45&@432.07&M53&812&?2Q/2;&4U&08,3/20&<br />
4U&528;4?Z&/53,08?/207&0,>>208&81M8&=4@928/8/Q2&/582?M=8/45&/5&<br />
/554QM8/Q2&M=8/Q/8/20&@MN&8MZ2&3/UU2?258&U4?@0&M==4?3/5>&84&<br />
812&08?,=8,?2&4U&9MN4UU0&4U&/554QM8/Q2&M=8/Q/8/20&84&[;/552?0[&<br />
M53&[.402?0[M5/]M8/45&@432.07&M53&812&?2Q/2;&4U&08,3/20&4U&<br />
528;4?Z& /53,08?/207& 0,>>208& 81M8& =4@928/8/Q2& /582?M=8/45& /5&<br />
/554QM8/Q2&M=8/Q/8/20&@MN&8MZ2&3/UU2?258&U4?@0&M==4?3/5>&84&<br />
812&08?,=8,?2&4U&9MN4UU0&4U&/554QM8/Q2&M=8/Q/8/20&84&[;/552?0[&<br />
M53&[.402?0[&-252U/80&84&812&@M?Z280&<br />
U4?&0,99.N&4U&=4@9M8/-.2&04U8;M?27&-N&2\9M53/5>&812&@M?Z28&<br />
0,=1&81M8&M&>?2M82?&QM?/28N&4U&04U8;M?2&/0&4UU2?23&4?&812&,5/8&<br />
9?/=2&32=?2M020&'0/5=2&U/\23&=4080&4U&04U8;M?2&32Q2.49@258&<br />
=M5&-2&09?2M3&4Q2?&M&>?2M82?&5,@-2?&4U&948258/M.&-,N2?0(&0N082@&@4?2&QM.,M-.27&M53&81,0&/5=?2M02&32@M53&<br />
U4?&/8/=&<br />
?2094502& 4U& 04U8;M?2& 0,99./2?0& 84& 812& /5=?2M02& /5& 812&<br />
=450,@98/45&4U&492?M8/5>&0N082@0
HANS W. GOTTINGER<br />
2.6 CHINESE INNOVATION AND COMPETITION<br />
& T582?5M8/45M.&^,0/5200&M53&I=454@/=0&G202M?=1&!"#$%&!'!()&!!*$!& !O&<br />
&<br />
-2882?&81M5&2\/08/5>&9?43,=80&45.N&-N&04@2&,02?0&'4?&U4?&04@2&<br />
,020(&M0/32&9?/=2&=450/32?M8/4507&/0&<br />
=450/32?23&-2882?&81M5&2\/08/5>&9?43,=80&-N&M..&,02?0&'U4?&M..&<br />
,020(%&125=2&812&/32M&81M8&9?43,=80&=M5&-2&?M5Z23&M==4?3/5>&<br />
84&M&lV,M./8N&.M332?m?4,90&81M8&<br />
=M5&=42\/08&;/81&4.32?&9?43,=8&>?4,90252?M..N&<br />
M004=/M823&;/81&M&9?4=200&;12?2-N&52;&M53&-2882?&9?43,=80&<br />
3/09.M=2&4.32?&4-04.282&9?43,=80&U?4@&812&@M?Z28&<br />
lV,M./8N&;/534;m&2\/0807&4,80/32&;1/=1&U/?@0&=M5548&0,?Q/Q2&<br />
/5& 812& @M?Z28(,/32& 45& 812&<br />
5M8,?2& 4U& @M?Z28& 3N5M@/=0& 81M8& @MN& MUU2=8& M& 9M?8/=,.M?&<br />
/53,08?N>208& 81M8& /5=,@-258& U/?@0& @MN&<br />
1MQ2&3/UU2?258&/5=258/Q20&84&/554QM82&81M5&258?M580/Q25&812&3/UU/=,.8/20&81M8&/5=,@-258&.2M3/5>&U/?@0&@MN&1MQ2&<br />
84&32M.&;/81&3?M08/=&=1M5>20&o&M&U4?@&4U&4U&/58?M*U/?@&<br />
0N082@/=&/52?8/M&o&812&4994?8,5/8/20&U4?&0829*;/02&/554QM8/45&<br />
@MN&?2.M8/Q2.N&UMQ4?&/5=,@-258&U/?@0252?M.&8124?N252?/=& .2Q2.7&<br />
>/Q25& 812& =,??258& .2Q2.& 4U& ,532?08M53/5>& 4U& 81202& 2UU2=80&812&./5Z&-28;225&9?43,=8&@M?Z28&<br />
=4@928/8/45&M53&/554QM8/45,/3M5=2&U?4@&812&./82?M8,?2&45&81/0&/00,27&;1/=1&<br />
/0&92?1M90&-208&M33?20023&45&M&=M02*-N*=M02&-M0/01& eKFE& ;2?2& 0Z/@@23& 4UU& U?4@&<br />
M58/=/9M8/5>&@M?Z28&9M?8/=/9M5807&45.N&84&Z54;&.M82?&81M8&?/QM.&<br />
08M53M?30& -M023& 45& 52;& 82=154.4>/20& 'S/*b/(& -N9M00/5>&<br />
eKFE7& ;4,.3& =?2M82& =4@9.282.N& 52;& @M?Z280& 548& -2/5>&<br />
M58/=/9M823
2.6 CHINESE INNOVATION AND COMPETITION<br />
HANS W. GOTTINGER<br />
!B&<br />
RM50&S2?)&J1/5202&T554QM8/45&M53&J4@928/8/45)H5&T53,08?N&JM02&4U&812&L.4-M.& &<br />
F2.2=4@@,5/=M8/450&IV,/9@258&T53,08?N&<br />
@M?Z280&812&?M5>2&<br />
4U& 9400/-.2& 0,-08/8,8202& @MN&<br />
MUU2=8&812&08?,=8,?2&4U&812&@M?Z28&M0&32U/5237&U4?&2\M@9.27&-N&<br />
@MZ/5>& 2=454@/20& 4U& 0=M.2& @4?2& 4?& .200& /@94?8M5820&@MN&-2&3,2&84&/554QM8/45&,532?8MZ25&-N&812&U/?@0&/5&<br />
812& @M?Z287& 4?& 3,2& 84& UM=84?0& 4,80/32& 812& @M?Z28& 0,=1& M0&<br />
2\82?5M.&82=154.4>/=M.&32Q2.49@25807&4?&01/U80&/5&=450,@2?&<br />
8M0820&M==4,58&4U&81202&UM=84?0&/0&548&0/@9.N&M&V,208/45&4U&<br />
2\M@/5/5>& 258?N& 9400/-/./8/2020&<br />
4Q2?&8/@22?7&812&8;4&@2?>/5>&U/?@0&@MN&<br />
=,??258.N& -2& 492?M8/5>& /5& =4@9.282.N& 029M?M82& ?2.2QM58&<br />
@M?Z2807& 81,0& ?M/0/5>& 54& =,??258& =4@928/8/45& =45=2?52&/5&812&U,8,?27&0,=1&81M8&81202&U/?@0&M?2&2\92=823&84&<br />
-2&=4@928/84?0&/5&812&U,8,?22?&<br />
014,.3& -2& -.4=Z23%& 0/@9.N& 81M8& =4@928/8/45& M5M.N0/0& @MN&<br />
5223&84&-2&U4?;M?3*.44Z/5>&M0&84&14;&=,??258& @M?Z280&M?2&<br />
./Z2.N&84&2Q4.Q2&/5&U,8,?21./>1823&M-4Q27&M&U/?@&@MN&548&-2&M=8/Q2&45&M5N&=,??258&<br />
@M?Z280&-,8&@MN&-2&32Q2.49/5>&9?43,=80&81M8&;4,.3&=4@9282&<br />
/5&812&U,8,?2M8/45& 45& 812& -M0/0& 4U& 04@2& 9?2*32U/523& /53/=/M>208&812&-252U/80&4U&812&<br />
U,?812?&M5M.N0/0&9?494023&/5&81/0&08,3N20& /5& ?2.M8/Q2&<br />
9?/=20&%&81202&U2M8,?20&M?2&3/0=,0023&-?/2U.N&-2.4;&4U&9M82580&/0&329253258&45&04@2&U4?@&4U&<br />
54Q2.8N7& /554QM8/45& /0& 52=200M?N& 'M.814,>1& 812& ?2.M8/Q2&<br />
/@94?8M5=2&4U&/554QM8/45&/5&812&=4@928/8/Q2&9?4=200&/0&548&<br />
U,..N&208M-./0123(/207&=1M?M=82?/08/=0&M53&9?/=20
HANS W. GOTTINGER<br />
2.6 CHINESE INNOVATION AND COMPETITION<br />
& T582?5M8/45M.&^,0/5200&M53&I=454@/=0&G202M?=1&!"#$%&!'!()&!!*$!& !q&<br />
&<br />
7"C"$HA+63B(.$35$)& &<br />
!! /58?43,=8/45& 4U& 9?43,=80& 81M8& 2@-43N& M& 52;&<br />
82=154.4>N%& &<br />
!! 014?8&9?43,=8&./U2*=N=.20%& &<br />
!! 1/>1&9?494?8/45&4U&@M?Z28&8,?54Q2?&M==4,5823&U4?&-N&<br />
52;& 9?43,=80& '/$*.:%"#2%<br />
G".D-*%F:#"0+5$% %<br />
M53&R,M;2/&81402&7&/5&812&=4,?02&4U&32Q2.49@2587&<br />
?2M..N& 45.N& 81?22& _FF7& `FI& M53& R,M;2/& 1MQ2& -2=4@2&<br />
/582?5M8/45M..N&4U&=45=2?57&M53&812N&M?2&812&34@/5M58&U4=,0&<br />
4U&81/0&08,3N& M53& T>2.& '!""#(& 08,3/23& 81202& U4,?& J1/5202&<br />
0,99./2?0& ;/81& 8;4& A4/58& Q258,?207& ^2/A/5>& T582?5M8/45M.&<br />
E;/8=1/5>& EN082@& '^TEJ(7& M& E/2@250& A4/58& Q258,?2& M53&<br />
E1M5>1M/&^2..7&M53&H.=M82.*a,=258&A4/58&Q258,?24Q2?5@258& M00/08M5=2& ;2?2&<br />
/@94?8M58&84&J1/5202&0,99./2?0[&08?M82>/20&;1/.2&U4?2/>5&A4/58&<br />
Q258,?20&?2./23&45&U4?2/>5&=4@9M5/204Q2?5@258(4Q2?5@258&1M0&8?/23&84&32=?2M02&812&/@94?8&<br />
4U&E+J&0;/8=120&23&82.291452&492?M84?0&84&,02&<br />
9?43,=80&U?4@&5M8/45M.&@M5,UM=8,?2?0&84&`1M5>&M53&T>2.&'!""#(&7&/5&812&2M?.N&#YY"0&<br />
J1/5202*4;523& U/?@0& 12.3& 45.N& M-4,8& #& 9=& 4U& 812& @M?Z28&<br />
01M?2&4U&=258?M.&4UU/=2&2\=1M5>20M/523&<br />
@4?2& 81M5& 1M.U& 4U& 812& 5M8/45M.& @M?Z28& 0/5=2& #YYq&<br />
08?45>& =1M..25>20& U?4@& 34@208/=& U/?@0& M53& .4;2?& 9?4U/80&<br />
04@2&.2M3/5>&6M9M5202&U/?@07&0,=1&M0&b,A/80,&M53&XIJ7&1MQ2&<br />
;/813?M;5& U?4@& J1/5M[0& E+J& @M?Z287& M53& 04@2& .2M3/5>&<br />
S2082?5& E+J& 9?43,=2?0& 1MQ2& -225& =450/32?/5>& .2MQ/5>& M0&<br />
;2..M/508&A4/58&Q258,?20&M53&U4?2/>5&=4@928/8/45&/5&<br />
?,?M.& M53& ?2@482& J1/5202& @M?Z280& ;1/=1& ;2& =M..& 812&<br />
029M?M8/45&91M02&4U&=4@928/8/45&812N&1MQ2&-225&.2M?5/5>&45[&<br />
=4,?027& =45Q/=8/45& M53& =M9M-/./8N[& 'T^K& ^,0/5200& T508/8,82&<br />
45&J4@9M5N&L?4;81(&81?4,>1&=M8=1,9&3?/Q25&/554QM8/45&84&<br />
812& 2\8258& 81M8& 812N& =4,.3& =4@928282& 12M3& 45& ;/81& @MA4?&<br />
U4?2/>5&0,99./2?0&4?&812/?&A4/58&Q258,?20&/5&@MA4?&,?-M5&M?2M0&<br />
M.45>& 812& =4M08M.& =?20=258& 4U& J1/5M7& 08?28=1/5>& U?4@& Rf7&<br />
E125]257& E1M5>1M/& M53& ^2/A/5>25=2&91M02& 4U& 1/>1&<br />
82=154.4>N& U/?@0& M0& M& 82=154.4>N& ?M=2& /5& ;1/=1& UM../5>&<br />
-21/537&>288/5>&M12M3&M53&=M8=1/5>*,9&84&/53,08?N&.2M32?01/9&<br />
/0&812&5M@2&4U&812&>M@2&'L488/5>2?7!""B7J1M9.2& U/\23& ./52& =M??/23& M& O7"""& G2@/@-/&<br />
'GK^(&=408&7&M53&4U825&M54812?&O7"""&GK^&;M0&522323&M0&<br />
M5&2\923/8/5>&U22&'_2S40Z/57!""#(&/5U?M08?,=8,?M.&?2U4?@&M53&>?4;81&/5&<br />
J1/5M&/8&;M0&32=/323&84&,02&82.2=4@@,5/=M8/450&2V,/9@258&<br />
0,99./2?0&84&9?4@482&812&32Q2.49@258&4U&545*A4/58&Q258,?2&<br />
=M9M-/./8N& /5& J1/5M& M0& 2M?.N& M0& @M?Z28& 8?M50/8/45/5>& ;M0&<br />
8M?>2823&'X4,>18457!""q7&J1M9&528;4?Z&2V,/9@258&@M?Z28&4?/>/5M..N&<br />
U4,?& =4@9M5/20& ;2?2& /3258/U/23& M0& Z2N& 2\9452580& 4U& 81/0&<br />
08?M82>N)& 6,.45>& 'L_F(7& _M8M5>& '_FF(7& `145>\/5>& '`FI(&<br />
52;& 9?43,=8& 32Q2.49@258& /5[M99?49/M82[& 82=154.4>N& 81M8&<br />
! "'!<br />
194
2.6 CHINESE INNOVATION AND COMPETITION<br />
HANS W. GOTTINGER<br />
!D&<br />
RM50&S2?)&J1/5202&T554QM8/45&M53&J4@928/8/45)H5&T53,08?N&JM02&4U&812&L.4-M.& &<br />
F2.2=4@@,5/=M8/450&IV,/9@258&T53,08?N&<br />
=4,.3&.2M3&84&M&9M?M3/>@&01/U8&4U&81M8&?M=22524,0&U/?@0&=M@2&,9&<br />
;/81&0/@9.2?7&0,UU/=/258&92?U4?@M5=2&.2Q2.&4U&528;4?Z&>2M?&M8&<br />
M&@,=1&.4;2?&9?/=2&U4?&,02&/5&?,?M.&J1/5M& ;1/=1&=4,.3&-2&<br />
/3258/U/23& M0& M99?49?/M82& 82=154.4>N@258M8/45&M99.N/5>&84&?,?M.&M53&,?-M5:@28?494./8M5&M?2M0.N7& 812& 0M@2& 844.0& 4U& @M?Z28& 92528?M8/45& ;2?2&<br />
M99./23&-N&J1/5202&U/?@0&.M82?&45&45&/5&4812?&2@2?>/5>&M53&<br />
8?M50/8/45M.&2=454@/20& &<br />
Q/?8,M..N& M..& M?4,53& 812& ;4?.32823& M8& ?,?M.& M53& ?2@482& M?2M0& 4U& J1/5M7& =.2M?.N&/5&M&<br />
32Q2.49/5>&2=454@N&7&=4,.3&-2&82?@23&[?2Q2?02&/554QM8/45[&<br />
/5&812& @2M5/5>&4U&L4Q/53M?MAM5&M53&F?/@-.2&'!"#!(&0/5=2&<br />
812N& 4?/>/5M823& /5& 812& 52230& M53& ?2V,/?2@2580& 4U& 81/0&<br />
/5U?M08?,=8,?M.& 02=84?& 81M8& ;M0& 029M?M82& 84& 812& M3QM5=23&<br />
,?-M5&M?2M0&/5&J1/5M4?/]2&8;4&M092=80&4U&81202&/554QM8/450&81M8&<br />
?2.M82&84&812&3/=1484@N&4U&812&82.2=4@&@M?Z28&/5&J1/5M&/5&812&<br />
2M?.N&#YY"0)&<br />
'#(&52;&@M?Z280&M53&'!(&V,M./8N&=1M?M=82?/08/=0443&<br />
254,>1[&;/81&<br />
0/>5/U/=M58.N& .4;2?& 9?/=/5>& M8& 3229&<br />
3/0=4,580"#"$E3B@(+&+&=($GI,.(.$ $<br />
@M?Z280&0/>5M..23&812&08M?8&4U&812&=4@928/8/Q2&=45Q2?>25=2&<br />
9M815&0,99./2?0&=4@92823&12M3*84*12M3&<br />
U4?8&12&,?-M5&M53&?/?M.&J1/5202&@M?Z28&M53&8125&/5&812&>.4-M.&<br />
@M?Z28?2M82?&<br />
5,@-2?0& 4U& 3/>/8M.& 0,-0=?/-2?& ./520& '_Ea(& /5& J1/5M& 81M5&<br />
U4?2/>5& 0,99./2?0& /5=.,3/5>& 812/?& A4/58& Q258,?20&<br />
3/3'SE67!""C%_/88-2?52?7!""C(&=45=2?523&M-4,8&812&<br />
U/2?=2& =4@928/8/45& U?4@& J1/5202& 0,99./2?0& /5& 812& >.4-M.&<br />
@M?Z2805M..23&M.?2M3N&M&8?M50/8/45&U?4@&812&=4@928/8/Q2&<br />
=45Q2?>25=2& 9M81& 84& 812& =4@928/8Q2& >.4-M./]M8/45& 9M81& U4?&<br />
J1/5202& 0,99./2?025=2& 9M81&<br />
82=154.4>/20& 2Q4.Q2& 84& =4@9282& 12M3*45& U4?& 812& 0M@2&<br />
=,084@2?05&<br />
0,99./2?0& 04.3& 1/>1& =M9M=/8N& E+J& 0;/8=120& 84& 0M8/0UN& 812&<br />
52230&4U&.M?>2&7&UM08&>?4;/5>&=/8/202823& 812& /5U?M08?,=8,?M.7&<br />
545*=450,@2?&@M?*Z28&M53&M3323&QM.,2&84&UM=/./8M82&>?4;81&<br />
9?4092=80&/5&,532?32Q2.4923&J1/5202&?2>/450&=2582?0'.2Q2.O(1M/&<br />
^2..[0& 9?43,=80& ;2?2& 04.3& 84& .M?>2& 3/08?/=8& =2582?& '.2Q2.&<br />
#(79?4Q/5=/M.&0;/8=1/5>&=2582?0&'.2Q2.&!(&M53&=/8N&0;/8=1/5>&<br />
=2582?0&'.2Q2.&$(4&U?4@&.M?>2&84&0@M..&<br />
>24>?M91/=&,5/805& 0,99./2?0&<br />
M53&812/?&A4/58&Q25,?20&M==4,5823&U4?&qD9=&'KTT7&!""C(& & &<br />
H54812?& ?2M045& U4?& 812& 029M?M8/45& -28;225& J1/5202&<br />
0,99./2?0& M53& 208M-./0123& 0,99./2?0& ;M0& 81M8& ?2>,.M84?N&<br />
?2V,/?2@2580&-N&812&+408&M53&F2.2=4@@,5/=M8/45&^,?2M,0&<br />
'+F^0(& /5& ?,?M.& M?2M0& ;2?2& V,/82& 3/UU2?258& 81M5& 812&<br />
?2V,/?2@2580& 4U& 812& 9F^0& /5& 812& =/8/20&<br />
F1,07&/5&2\9.M5/5>&812&91M020&4U&@M?Z28&32Q2.49@258&U4?&<br />
J1/5202&U/?@07&<br />
/5&82?@0&4U&1/084?N&;2&@MN&/3258/UN&812&029M?M8/45&91M02&<br />
?4,>1.N& ;/81& 812& 92?/43& 4U& @/3& #YD"0& 84& @/3& #YY"07& 812&<br />
=45Q2?>25=2& 91M02& U?4@& @/3& #YY"0& 84& 2M?.N& !""0& M53&812&<br />
>.4-M./]M8/45&91M02&U4?&812&?208&,58/.&54;/20&'./Z2&E+J& 0;/8=120(&<br />
-,8& M.04& 84& -?4M3& 9?43,=8& >?4,90& M0& 528;4?Z& ?4,82?0& M53&<br />
02?Q2?0& M0& ;2..& M0& 84& =450,@2?& 253& 9?43,=80& 0,=1& M0&<br />
0@M?8914520&M53&8M-.28&32Q/=20&814,>1&81402&@M?Z280&1MQ2&<br />
2@2?>23& 45.N& ?2=258.N/=M..N& @48/QM823& 81?4,>1& /554QM8/45& -,8& @MN& -2&<br />
0,994?823& -N& 08?M82>/=& 3/?2=8/45& 4?& 4812?& UM=84?0& 0,=1& M0&<br />
>4Q2?5@258&3/?2=823&/53,08?/M.&94./=N/20&34&<br />
548&/582?M=8&'/5&82?@0&4U&-2/5>&?29.M=2M-.2(&/5&812&=4,?02&4U&<br />
812/?&32Q2.49@25825=2&91M02&7&82=154.4>/20&<br />
2Q4.Q2& 84& =4@9282& 12M3*45& U4?& 812& 0M@2& =,084@2?0.4-M./]M8/45&9M81&=4@928/8/45&2Q4.Q20&81?4,>1&<br />
82=154.4>N& 92?U4?@M5=2& M53& .4=M.& @M?Z28& =1M?M=82?/08/=0&<br />
0,=1& M0& ?2>,.M84?N& ?2V,/?2@2580& 4U& 32Q2.49/5>& 2=454@/207&<br />
U4?2/>5& >4Q2?5@258& 0,994?8& M53& 0,-0/3/]M8/45& /=.,3/5>&<br />
948258/M.&=4??,98/45&0=12@20&/5&82532?/5>1&=M9M=/8N&E+J&<br />
0;/8=120&/5&#YYO&U4?&0M.2&/584&812&J1/5202&,?-M5&M53&?,?M.&<br />
9.M8U4?@0'I3V,/087&!""$(
HANS W. GOTTINGER<br />
2.6 CHINESE INNOVATION AND COMPETITION<br />
& T582?5M8/45M.&^,0/5200&M53&I=454@/=0&G202M?=1&!"#$%&!'!()&!!*$!& !Y&<br />
&<br />
>"!"$G)3:A6+$F(=(/3@B('+$J$0,'A5,6+A)&'K$E,@,L&/&+D$,':$<br />
7&!""!(&M53&a/,7&!""$(2& 9M?80& 4U& 812/?&<br />
?2Q25,2&/5&Gg_&?2V,/?23&84&32Q2.49&1/>1&=M9M=/8N&0;/8=120252?M823& 4Q2?& 812&<br />
9?2Q/4,0& U4,?& N2M?0&84&32Q2.49&/80&RJ& 0;/8=10&84&32Q2.49&/80&RJ&0;/8=11& =M9M=/8N& 0;/8=120& ;M0& =450/32?23& M5& /@94?8M58&<br />
-?2MZ*81?4,>1& 84& 8?M50/8& 84& 812& =45Q2?>25=2& 91M02>2?&528;4?Z&M3QM58M>2&/0&./Z2.N&84&M88?M=8&@4?2&=,084@2?0&<br />
/5&812&M-025=2&4U&/554QM8/451&9?4-M-/./8N7&8125&MU82?&<br />
?2M./]M8/45& 81/0& U/?@& ;/..& >28& M12M3& 4U& /80& ?/QM.0& M53& >M/5&<br />
@M?Z28&01M?2&45&/80&-?M53&/5329253258&4U&9?/=/5>
2.6 CHINESE INNOVATION AND COMPETITION<br />
HANS W. GOTTINGER<br />
$"&<br />
RM50&S2?)&J1/5202&T554QM8/45&M53&J4@928/8/45)H5&T53,08?N&JM02&4U&812&L.4-M.& &<br />
F2.2=4@@,5/=M8/450&IV,/9@258&T53,08?N&<br />
.2M32?01/9&940/8/45& &<br />
812N&1MQ2&548&45.N&84&=M8=1&,9&-,8&M.04&84&.2M9U?4>&812/?&<br />
?/QM.0N&U?458/2?&/8&/0&M&@,=1&@4?2&3/UU/=,.8&81/5>&84&342M?7&<br />
J/0=4& EN082@07& =450/32?0& R,M;2/& M0& /80& 84,>1208& ?/QM.&<br />
'SE67!"#!3(/=& 940/8/45/5>& 4U& 812&<br />
/53/>2524,0& J1/5202& U/?@01& 812/?& =4@928/8/Q2& 91M020&<br />
U4..4;23& 812& ?,.20& 4U& L4& 4?& S2/[=1/7& M& 82??/84?/M.& >M@2&<br />
320/>523& 84& 0,-3,2& N4,?& ?/QM.& -N& 09M8/M..N& 25=/?=.2& M53&<br />
=4?52?&1/@&M53&=M98,?2&1/0&940/8/45&-N&34@/5M8/5>&1/@&4Q2?&<br />
812&-4M?3M?37&812&J1/5202&@M?Z28&.44Z0&./Z2&M&L4&<br />
-4M?3& ;1/=1& ;125& 2\9M5323& 84& 2@2?>/5>& M53& 8?M50/8/45M.&<br />
2=454@/20& 258M/.0& 812& >.4-M.& @M?Z28M@2&/0&548&4Q2?&N28N&/0&=4@92../5>&-,8&52230&<br />
U,?812?&?202M?=1&812&849&#""&J1/5202&2.2=8?45/=0:TF&=4@9M5/20/5>&<br />
M53&8?M50/8/45M.&2=454@/20&-,8&M.04&02Q2?M.&dIJ_&=4,58?/202&=M9M=/8N&E+J&2\=1M5>2&M53&M==200&<br />
528;4?Z&2V,/9@258&/5&#YY$&M53&#YYC7&?2092=8/Q2.N/=&9?/4?/8/20)&'#(&_2Q2.49&U?4@&-2/5>&M&<br />
0/5>.2& 0;/8=1/5>& 9?43,=8& =4@9M5N& 84& M& @,.8/*9?43,=8&<br />
=4@9M5N7&'!(&M33?200&812&52230&4U&812&,?-M5&@M?Z28&M0&;2..&<br />
M0& 81402& 4U& 812& ?,?M.& @M?Z28& M53& '$(& M33?200& 812& 52230& 4U&<br />
/582?5M8/45M.& &<br />
KM?Z280&M53&M==200&9?43,=80M>23& /5& ?202M?=1& 9M?852?01/90& 'G6W0(&<br />
;/81&.M?>27&;2..*208M-./0123&U/?@0&0,=1&M0&F2\M0&T508?,@25807&<br />
K484?4.M&M53&H>2?2&EN082@0&'`FI[0&M55,M.&?294?8&4U&!""$(N[7&X2;&h4?Z)&<br />
d\U4?3&e5/Q*^,0/5200& S22Z'!"#!(7[`FI& M/@0& 84& 9M00& ?/QM.&<br />
R,M;2/&/5&0@M?891452&0M.20&/5&J1/5M[7&6M57_
HANS W. GOTTINGER<br />
2.6 CHINESE INNOVATION AND COMPETITION<br />
& T582?5M8/45M.&^,0/5200&M53&I=454@/=0&G202M?=1&!"#$%&!'!()&!!*$!& $#&<br />
&<br />
^2/A/5>)J4582@94?M?N&J1/5M&+?200& &<br />
sBt! J1/5M& F49& #""& I.2=8?45/=0& J4@9M5N& a/08'!""$(7&<br />
;;;M?& &<br />
s##t! _2S40Z/576/8M.& 0;/8=1/5>& 08M8,0& M53&<br />
U4?2=M08[7&;;;N& M53&<br />
RN92?=4@928/8/457&a45345)&G4,8.23>2& &<br />
s#Ot! L488/5>2?7R2)&KTF&+?200& &<br />
s!qt! +4?82?7K1& F2=154.4>N&<br />
J4@928/8/457&JM@-?/3>2)&RM?QM?3&e5/Q?200&<br />
=4,.3&0V,22]2&-/>&9.MN2?0[7&KM?=1&q& &<br />
s$Ct! F12& SM..& E8?228& 64,?5M.& '!"#!-(7[R,M;2/& M/@0& U4?& 1/>12?&<br />
9?4U/.2&/5&0@M?8914520[7&KM?=1&D& &<br />
s$Ot! F12& SM..& E8?228& 64,?5M.& '!"#!=(7[R,M;2/& M/@0& 84& -2=4@2&<br />
@4?2&@M8,?2[7&H0/M&F2=154.4>N7KM?=1&#D& &<br />
s$Bt! F12& SM..& E8?228& 64,?5M.'!"#!3(7[J/0=4& JId& 0220& J1/5M[0&<br />
R,M;2/&M0&/80&84,>1208&?/QM.[7&H9?/.&B& &<br />
s$qt! F12& SM..& E8?228& 64,?5M.& '!"#!2(7[J1/5M& SM?50& R,M;2/&<br />
G294?8&=4,.3&1M?@&e
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
HANS W. GOTTINGER<br />
!"#$%&$%'()("*(%+,",-(.("#/%012$%3/%41$%3/%3333% 1<br />
!"#$%&'&'(%)$"*'&'"+'+,)&-."/,-)*&.0&%,'1)022'0+1%)<br />
3".#0&'"+4).%('%5)0+6)*7+&-%*'*)<br />
!"#$%&'(#'()*+,,-#.'()<br />
Institute of Ec<strong>on</strong>omic Analysis,<br />
University of Osaka (KGU),<br />
Osaka, Japan<br />
E-mail: hansgottinger@aol.com<br />
!"#$%&'$( This paper explores competitive positi<strong>on</strong>ing through network<br />
competiti<strong>on</strong> <strong>on</strong> the basis of alliance formati<strong>on</strong> (strategic alliances, joint<br />
ventures). From a strategic perspective, technological competiti<strong>on</strong> will be<br />
refined <strong>and</strong> exp<strong>and</strong>ed into new markets, or new markets will be created through<br />
alliance formati<strong>on</strong>. Alliance formati<strong>on</strong> could speed up competitive positi<strong>on</strong>ing<br />
<strong>and</strong> technological leadership in strategically important, though geographically<br />
diverse, markets. Revenue management tools can help in linking alliance<br />
formati<strong>on</strong> to better competitive outcomes, thereby improving strategic<br />
directi<strong>on</strong>s.<br />
)*+,-%.#( competiti<strong>on</strong>; corporate governance; managerial ec<strong>on</strong>omics;<br />
networks; revenue management; strategic alliance.<br />
/*0*%*1'* to this paper should be made as follows: <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H-W. (xxxx)<br />
‘Competitive positi<strong>on</strong>ing through strategic alliance formati<strong>on</strong>: review <strong>and</strong><br />
synthesis’, !"#$%&$%'()("*(%+,",-(.("#, Vol. x, No. x, pp.xx–xx.<br />
23"43-5%&673'8 1-$*#( Dr. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g> is a Professor of Ec<strong>on</strong>omics at the<br />
University of Osaka (KGU), <strong>and</strong> the Director of the Institute of Management<br />
Science, Maastricht, NL. He has internati<strong>on</strong>ally taught <strong>and</strong> widely published in<br />
the areas of industrial, energy <strong>and</strong> envir<strong>on</strong>mental ec<strong>on</strong>omics. In these areas he<br />
has also advised internati<strong>on</strong>al organisati<strong>on</strong>s. His most recent forthcoming book<br />
titles, !""1),#51"/%6(78"121-9%,":%;9#=*7#*=(, illuminated the impact of the rise of the<br />
railway <strong>and</strong> related industries during the 19th century <strong>on</strong> the development of the modem<br />
capitalist corporati<strong>on</strong>. During the Industrial Revoluti<strong>on</strong>, fixed assets produced value<br />
through the creati<strong>on</strong> <strong>and</strong> distributi<strong>on</strong> of hard goods, often across l<strong>on</strong>g distances. The wide<br />
geographic, temporal <strong>and</strong> financial requirements of managing <strong>and</strong> operating a railway<br />
compelled the inventi<strong>on</strong> <strong>and</strong> evoluti<strong>on</strong> of substantially more sophisticated <strong>and</strong> structured<br />
corporate organisati<strong>on</strong>s. Earlier organisati<strong>on</strong>al forms <strong>and</strong> business practices were<br />
incapable of effectively managing in this new envir<strong>on</strong>ment. As such, the railway<br />
compelled the evoluti<strong>on</strong> of the modern global corporati<strong>on</strong>. Later, the breadth <strong>and</strong> depth of<br />
modern corporati<strong>on</strong>s encouraged the evoluti<strong>on</strong> of ever more complex functi<strong>on</strong>al<br />
Copyright © 200x Inderscience Enterprises Ltd.<br />
199
HANS W. GOTTINGER<br />
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
2% ;?@$%A1##5"-(=%<br />
hierarchies to coordinate disparate resources within expansive, multi-divisi<strong>on</strong>al firms<br />
(Ch<strong>and</strong>ler, 1962, 1990).<br />
We witness a similar phenomen<strong>on</strong> as we enter the 21st century. In c<strong>on</strong>trast to the<br />
Industrial Age, the emerging ec<strong>on</strong>omy primarily generates value through the creati<strong>on</strong>,<br />
disseminati<strong>on</strong> <strong>and</strong> applicati<strong>on</strong> of knowledge. Since the 1980s, networking technologies<br />
have created a dynamic similar to that of the railway in the 19th century, influencing the<br />
opti<strong>on</strong>s for corporate structures, relati<strong>on</strong>ships <strong>and</strong> competiti<strong>on</strong>. The most important <strong>and</strong><br />
far-reaching impacts occur as a result of how people <strong>and</strong> firms use these tools to create<br />
value. The c<strong>on</strong>flict between proprietary ownership as a necessary means for profit <strong>and</strong> the<br />
social nature of knowledge comprises the fundamental dynamic compelling the<br />
transformati<strong>on</strong> of organisati<strong>on</strong>al forms during the present period.<br />
In partial resp<strong>on</strong>se to this challenge, building <strong>and</strong> maintaining corporate alliances has<br />
become an increasingly important capability for the pursuit of both operati<strong>on</strong>al<br />
efficiencies <strong>and</strong> competitive advantage. Alliances have been transforming competiti<strong>on</strong><br />
<strong>and</strong>, as such, corporate strategy. The primary objective of this paper is to develop <strong>and</strong><br />
apply a new c<strong>on</strong>ceptual framework providing insight into the impact of alliances <strong>on</strong><br />
strategy (Secti<strong>on</strong> 5).<br />
Before we will deal with some pertinent issues of strategic envir<strong>on</strong>ments that<br />
facilitate or c<strong>on</strong>strain alliance formati<strong>on</strong>,<br />
Secti<strong>on</strong> 2 shows how managerial strategy drives competiti<strong>on</strong> but in turn is driven by<br />
competiti<strong>on</strong>. Secti<strong>on</strong> 3 shows how Transacti<strong>on</strong> Cost Ec<strong>on</strong>omics (TCE) <strong>on</strong> <strong>on</strong>e side limits<br />
the strategic manoeuvrability but, <strong>on</strong> the other h<strong>and</strong>, provides new opportunities to get<br />
ahead of rivals.<br />
The role of corporate governance in view of achieving superior competitive<br />
performance of the firm is discussed in Secti<strong>on</strong> 4. Secti<strong>on</strong> 5 builds the case for alliance<br />
formati<strong>on</strong> to strengthen competitive performance. Secti<strong>on</strong> 6 pinpoints the dimensi<strong>on</strong>s of<br />
network ec<strong>on</strong>omics in achieving leading market positi<strong>on</strong>s. Secti<strong>on</strong> 7 adds the time<br />
dimensi<strong>on</strong> as the facilitator of change management. Secti<strong>on</strong> 8 outlines the scope of<br />
applicati<strong>on</strong>s of network formati<strong>on</strong> to various high-tech industries.<br />
>&8-08#$%&$*5+(8'->>3$>*1$?8;1'*%$&31$+8&1.8'7&15*8<br />
C<strong>on</strong>structing <strong>and</strong> executing strategy, proactively or passively, requires predicti<strong>on</strong>s, yet<br />
the future always presents uncertainty. Herein lies the dilemma of the strategist –<br />
committing resources to an uncertain future. As market <strong>and</strong> technology change<br />
accelerates, uncertainty becomes of increasing c<strong>on</strong>cern. In a most basic challenge to<br />
strategic planning, in the view of Porter (1980), how can firms effectively positi<strong>on</strong><br />
themselves <strong>and</strong> their offerings if marketplace positi<strong>on</strong>s keep changing? Alternatively,<br />
executing <strong>on</strong> strategy in the real world requires firms to commit to directi<strong>on</strong>s they believe<br />
will provide sustainable profits. Commitment theory, as advanced by P. Ghemawat,<br />
argues that strategy requires resource investments that might be difficult or even<br />
impossible to recover from in the future, if too many strategic decisi<strong>on</strong>s turn out to have<br />
been wr<strong>on</strong>g (Ghemawat, 1991). Strategically phrased, you cannot arrive anywhere in<br />
particular if you do not commit to a directi<strong>on</strong> – <strong>and</strong> you cannot know in advance if you<br />
have chosen a good destinati<strong>on</strong>. Developing advantageous competency <strong>and</strong> resource<br />
combinati<strong>on</strong>s require substantial time <strong>and</strong> effort, so re-directi<strong>on</strong> can be costly.<br />
200
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
HANS W. GOTTINGER<br />
% B1.
HANS W. GOTTINGER<br />
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
4% ;?@$%A1##5"-(=%<br />
substantially decreased the uncertainty regarding the marketplace availability of inputs or<br />
capabilities to accomplish ec<strong>on</strong>omic objectives. If <strong>on</strong>e needs something made or<br />
accomplished, <strong>on</strong>e can probably (out) source it.<br />
Certainly, the c<strong>on</strong>trol over resources provided by vertical integrati<strong>on</strong> still provides a<br />
compelling rati<strong>on</strong>ale for the formati<strong>on</strong> <strong>and</strong> boundaries of firms; however, the uncertainty<br />
of another sort increasingly dominates as we move to the future. Rapid technology <strong>and</strong><br />
marketplace change threatens firms by challenging the relevance of every product,<br />
service <strong>and</strong> business model. This represents the uncertainty regarding what to execute or,<br />
by extensi<strong>on</strong>, what, exactly, to integrate. Which competencies <strong>and</strong> other resources will<br />
provide a competitive advantage over the l<strong>on</strong>g term, most effectively integrated into the<br />
firm? While this issue presents the most significant challenge in dynamic markets, even<br />
mature industries can transform as a result of many factors, such as changing regulatory<br />
regimes, new entrants, technologies or business models, or a significant re-directi<strong>on</strong> by<br />
an established player. Moreover, smaller firms typically do not have the capability or<br />
resources necessary to achieve a broadly integrated firm, even if they have the<br />
philosophical wherewithal.<br />
The greater the uncertainty regarding the future, the more firms must c<strong>on</strong>struct a<br />
portfolio of strategic opti<strong>on</strong>s, without compromising the focus necessary to successfully<br />
compete. So effective strategy requires commitments towards an uncertain future, while<br />
marketplace change threatens to undermine or even destroy a firm’s value creating<br />
capabilities.<br />
<strong>Markets</strong> characterised by rapid technology change exhibit high levels of alliance<br />
formati<strong>on</strong> compared with mature industries, which tend to exhibit c<strong>on</strong>solidati<strong>on</strong> <strong>and</strong> even<br />
decline (Hagedoorn, 1993). The high frequency of alliance formati<strong>on</strong> in emerging<br />
industries will become evident when we look at the biotechnology <strong>and</strong> pharmaceutical<br />
industries, which evidenced a ninefold increase in operative alliances between 1993 <strong>and</strong><br />
2000 (<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, 2004). More mature industries encountering re-organisati<strong>on</strong> or<br />
transformati<strong>on</strong> also experience increases in alliances, as occurred when the global<br />
chemical industry experienced severe overcapacity in the 1980s. Alliances <strong>and</strong> mergers<br />
allowed the industry to rati<strong>on</strong>alise producti<strong>on</strong> in a manner that competing firms could not<br />
have accomplished individually. These alliances represented an industry-wide alternative<br />
to c<strong>on</strong>solidati<strong>on</strong> through Mergers <strong>and</strong> Acquisiti<strong>on</strong>s (M&As), which occurred as well<br />
during this period.<br />
So marketplace actors apply network strategies in the periods of significant change<br />
<strong>and</strong> uncertainty. This seems to c<strong>on</strong>tradict the noti<strong>on</strong> that uncertainty engenders vertical<br />
integrati<strong>on</strong>. We are dealing with differing types of uncertainty, but the resp<strong>on</strong>ses are<br />
more similar than might appear. Vertical integrati<strong>on</strong> <strong>and</strong> alliances both integrate activities<br />
more closely than in the open market. To an extent, alliances represent integrati<strong>on</strong> by<br />
other means. N<strong>on</strong>etheless, alliances present significant strategic <strong>and</strong> managerial c<strong>on</strong>trasts<br />
to integrated firms, not the least of which is the opportunity for firms to interact with a<br />
broader, more diverse <strong>and</strong> potentially more productive universe of knowledge creating<br />
opportunities than is possible within a single firm. Let us first examine its precedents in<br />
the established organisati<strong>on</strong>al ec<strong>on</strong>omics <strong>and</strong> strategy literature.<br />
202
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
HANS W. GOTTINGER<br />
% B1.64*D In the early 1990s, Eastman Kodak Corporati<strong>on</strong> became a pi<strong>on</strong>eer in<br />
outsourcing Informati<strong>on</strong> <strong>Technology</strong> (IT) services. While many corporati<strong>on</strong>s outsourced<br />
various aspects of their IT operati<strong>on</strong>s, Kodak was <strong>on</strong>e of the first large, multinati<strong>on</strong>al<br />
corporati<strong>on</strong>s to outsource the majority of its corporate IT functi<strong>on</strong>. After a l<strong>on</strong>g period of<br />
planning, Kodak outsourced to three firms: Digital Equipment Corporati<strong>on</strong> (DEC),<br />
BusinessL<strong>and</strong> <strong>and</strong> IBM. Although a Harvard Business School case <strong>on</strong> the outsourcing<br />
strategy (Applegate <strong>and</strong> M<strong>on</strong>tealegre, 1995) reports overall success of the outsourcing<br />
venture, the process of building the integrati<strong>on</strong> <strong>and</strong> management processes between these<br />
four firms proved more challenging than Kodak’s CIO originally imagined. Kodak did<br />
achieve cost savings as a result; however, the overall cost savings were lower than<br />
anticipated. Kodak found that by outsourcing they saved the costs from shifting IT<br />
pers<strong>on</strong>nel <strong>and</strong> infrastructure to the firms whose core competencies were IT services.<br />
N<strong>on</strong>etheless, outsourcing introduced new costs involving developing staff <strong>and</strong> processes<br />
203
HANS W. GOTTINGER<br />
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
6% ;?@$%A1##5"-(=%<br />
to manage, or rather, coordinate the inter-firm efforts. C<strong>on</strong>trol costs were swapped for<br />
coordinati<strong>on</strong> costs.<br />
The TCE perspective biases the discussi<strong>on</strong> of this example towards the costs of<br />
achieving business results. Other motivati<strong>on</strong>s for outsourcing should include higher<br />
quality <strong>and</strong> improved corporate focus <strong>on</strong> the part of all firms in outsourcing relati<strong>on</strong>ships.<br />
Effective outsourcing should provide equal or higher-quality service than would be<br />
possible in-house, or at least a satisfactory level of service, at a similar or lower cost.<br />
Thus, outsourcing should represent the net value creati<strong>on</strong> between the c<strong>on</strong>tracting<br />
entities, allowing each to focus <strong>on</strong> its core competencies. While the TCE perspective<br />
approaches a successful outsourcing arrangement as a more efficient governance<br />
c<strong>on</strong>figurati<strong>on</strong>, focusing <strong>on</strong> costs can neglect or de-emphasise the potential benefits of<br />
greater value creati<strong>on</strong>.<br />
The transacti<strong>on</strong> cost TOF presents transacti<strong>on</strong> costs reducti<strong>on</strong> as the organising principle<br />
of firms; as such, TCE should prescribe any organisati<strong>on</strong>al form that provides the lowest<br />
transacti<strong>on</strong> costs for the ec<strong>on</strong>omic value created. It is partly a result of the efficiency <strong>and</strong><br />
effectiveness of the modem corporate form that a transacti<strong>on</strong> cost approach has provided<br />
theoretical justificati<strong>on</strong> for the existence of firms. It is also partly historical. Firms, in the<br />
traditi<strong>on</strong>al definiti<strong>on</strong>, exist, so they therefore must provide transacti<strong>on</strong> cost advantages<br />
over market governance, given TCE’s foundati<strong>on</strong> in bounded rati<strong>on</strong>ality (Sim<strong>on</strong>, 1947;<br />
Williams<strong>on</strong>, 1985) <strong>and</strong> the path-dependent nature of organisati<strong>on</strong>al <strong>and</strong> instituti<strong>on</strong>al<br />
development (Nels<strong>on</strong> <strong>and</strong> Winter, 1982; Hannan <strong>and</strong> Freeman, 1989). Moreover, the<br />
original c<strong>on</strong>text within which Coase developed his transacti<strong>on</strong> cost insights to<br />
underst<strong>and</strong>ing firm existence <strong>and</strong> boundaries likely biased the foundati<strong>on</strong> of the<br />
approach. Coase extrapolated from his experience working at the Ford Motor Company<br />
in the 1930s, recognising the company’s vertical integrati<strong>on</strong> strategy.<br />
He postulated that vertical integrati<strong>on</strong> was justified to decrease the transacti<strong>on</strong> costs<br />
that the firm would have otherwise incurred purchasing inputs in the open market. In the<br />
1930s, the industrial producti<strong>on</strong> model defined by Ford strove almost exclusively for<br />
operati<strong>on</strong>al efficiency. TCE is fundamentally an efficiency-based approach to<br />
underst<strong>and</strong>ing firms. Industries <strong>and</strong> strategies defined by operati<strong>on</strong>al efficiencies seem<br />
best suited to TCE. A number of researchers have asserted over the past two decades that<br />
inter-organisati<strong>on</strong>al relati<strong>on</strong>ships are often not well explained by TCE, particularly when<br />
efficiency issues do not represent the primary decisi<strong>on</strong> factors (Mariti <strong>and</strong> Smiley, 1983;<br />
Zajac <strong>and</strong> Olsen, 1993; Hamel <strong>and</strong> Prahalad, 1994). Decisi<strong>on</strong>s to pursue new firm<br />
competencies, or to form l<strong>on</strong>g-term R&D alliances are unlikely to be atomisable into<br />
discrete transacti<strong>on</strong>s (Argyres <strong>and</strong> Silverman, 2004).<br />
E8<br />
F-%6-%&$*85-G*%1&1'*8<br />
Alliances can be a critical strategic mechanism for the l<strong>on</strong>g-term success of ventures;<br />
however, alliances include a broad range of types, characterised by a number of factors<br />
such as integrati<strong>on</strong>, co-ownership, trust <strong>and</strong> l<strong>on</strong>gevity (Inkpen, 2001). Varied forms of<br />
alliance c<strong>on</strong>stellati<strong>on</strong>s also create varied c<strong>on</strong>straints for participants, from the l<strong>on</strong>g-term<br />
commitments into which they enter to other potential partner opportunities which<br />
participants forego to build <strong>and</strong> maintain membership. For instance, membership in<br />
certain strategic blocks can preclude involvement with competing blocks, the clearest<br />
204
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
HANS W. GOTTINGER<br />
% B1.
HANS W. GOTTINGER<br />
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
8% ;?@$%A1##5"-(=%<br />
So when do network arrangements foster a competitive advantage, relative to firm or<br />
market organisati<strong>on</strong>? When are the assets or resources worth more as part of an alliance<br />
than within an individual firm or the open market? How do networks assist in the<br />
creati<strong>on</strong>, appropriati<strong>on</strong> <strong>and</strong> sustainability of value?<br />
H8<br />
=%3G31580-%'*#8"*731.8&443&1'*80-%>&$3-18<br />
After all, why do firms form alliances?<br />
The implicati<strong>on</strong>s of network strategies vary substantially, depending <strong>on</strong> the purposes<br />
for which a particular network of firms forms, as well as the purposes for which the<br />
network actually operates (which are not always identical). The creati<strong>on</strong> <strong>and</strong> management<br />
of alliance c<strong>on</strong>stellati<strong>on</strong>s must be understood in light of motivati<strong>on</strong>s for their creati<strong>on</strong>.<br />
While TCE presents a compelling approach to underst<strong>and</strong>ing firm boundaries<br />
(Milgrom <strong>and</strong> Roberts, 1992, Ch. 2), managers clearly engage in M&A, firm growth <strong>and</strong><br />
divestiture, re-organisati<strong>on</strong> <strong>and</strong> other boundary-shifting initiatives for motivati<strong>on</strong>s quite<br />
distinct from transacti<strong>on</strong> cost minimisati<strong>on</strong>. An extensive survey of decisi<strong>on</strong> makers<br />
involved in alliances in the early 1980s found that n<strong>on</strong>e of them cited decreasing<br />
transacti<strong>on</strong> costs as a primary rati<strong>on</strong>ale for their decisi<strong>on</strong>s (Mariti <strong>and</strong> Smiley, 1983).<br />
Marketplace competiti<strong>on</strong> encourages firms to minimise costs, but it also compels firms in<br />
other directi<strong>on</strong>s: to acquire firms to pre-empt or resp<strong>on</strong>d to competitors, to adjust<br />
strategic visi<strong>on</strong> in the face of disruptive technologies, to innovate to create value over the<br />
l<strong>on</strong>g term, to name a few. Innovati<strong>on</strong> by its nature requires investment in the creati<strong>on</strong> of<br />
new knowledge <strong>and</strong> capabilities, whether these capabilities are new to a particular firm or<br />
new to the marketplace. In the case of knowledge capital creati<strong>on</strong>, such as in R&D<br />
partnerships, ‘transacti<strong>on</strong>’ costs in the form of inter-firm coordinati<strong>on</strong> can in many cases<br />
be higher than an internally c<strong>on</strong>trolled effort; however, the overall value created by the<br />
combinati<strong>on</strong> of capabilities between firms can outweigh the increased costs of<br />
coordinati<strong>on</strong>. Zajac <strong>and</strong> Olsen’s noti<strong>on</strong> of ‘transacti<strong>on</strong>al value’ attempts to reflect the<br />
importance of value calculati<strong>on</strong>s in underst<strong>and</strong>ing inter-organisati<strong>on</strong>al arrangements<br />
(Zajac <strong>and</strong> Olsen, 1993).<br />
While value creati<strong>on</strong> <strong>and</strong> cost reducti<strong>on</strong> reflect the complementary noti<strong>on</strong>s for the<br />
factors that encourage firms to change their boundaries through M&A or interorganisati<strong>on</strong>al<br />
arrangements, these approaches do not assist much in elucidating the<br />
complexity of motivati<strong>on</strong>s influencing organisati<strong>on</strong>al decisi<strong>on</strong>s. In broad terms of the<br />
field of industrial organisati<strong>on</strong>, cost <strong>and</strong> value creati<strong>on</strong> provide appropriate<br />
generalisati<strong>on</strong>s. For the purposes of applied corporate strategy, cost <strong>and</strong> value paradigms<br />
are by themselves woefully limiting, particularly given an ever-uncertain future.<br />
Ec<strong>on</strong>omists attempt to reflect the behaviour of market actors based <strong>on</strong> preferences,<br />
available data <strong>and</strong> other decisi<strong>on</strong> factors, as well as provide prescripti<strong>on</strong>s for the most<br />
effective ec<strong>on</strong>omic acti<strong>on</strong>s as a result of the insights. The study of corporate strategy<br />
attempts to decipher why firms organise <strong>and</strong> act the way they do, how perhaps they<br />
should act regarding the development <strong>and</strong> executi<strong>on</strong> of effective strategies (but often do<br />
not) <strong>and</strong> the outcomes associated with various scenarios (Grant, 2001). The fact that<br />
much of the strategy literature seeks grounding in ec<strong>on</strong>omics reflects their<br />
complementary nature. Implicit in both of these disciplines is the noti<strong>on</strong> of motivati<strong>on</strong><br />
(Milgrom <strong>and</strong> Roberts, 1992, Ch. 16). Ec<strong>on</strong>omic behaviour, while observable in the<br />
marketplace, can <strong>on</strong>ly be modelled in the classical sense of optimising functi<strong>on</strong>s (even<br />
206
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
HANS W. GOTTINGER<br />
% B1.&$3-18<br />
What are the key building essentials the firms might need for network formati<strong>on</strong>? There<br />
are potentially limitless reas<strong>on</strong>s, but all rati<strong>on</strong>al, optimising motivati<strong>on</strong>s for forming<br />
alliances from a firm’s perspective’ can be categorised into three types, aside from purely<br />
financial motivati<strong>on</strong>s.<br />
1 4(#F1=G%(71"1.57C. A firm is attempting to compete in some manner under the<br />
c<strong>on</strong>diti<strong>on</strong>s influenced by network ec<strong>on</strong>omies. Network ec<strong>on</strong>omies refer to the<br />
situati<strong>on</strong>s in which network ec<strong>on</strong>omic phenomena significantly influence the<br />
dynamics of a particular industry or market (<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, 2003).<br />
2 B1.
HANS W. GOTTINGER<br />
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
10% ;?@$%A1##5"-(=%<br />
envir<strong>on</strong>mental, instituti<strong>on</strong>al <strong>and</strong> competitive factors present in a particular industry,<br />
market <strong>and</strong>/or ec<strong>on</strong>omy. The following list presents a summary of the examples<br />
collected:<br />
Factors favouring alliances<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
Knowledge sharing <strong>and</strong> creati<strong>on</strong><br />
Distributi<strong>on</strong><br />
Risk sharing<br />
R&D outsourcing<br />
Network effects<br />
Leveraging complementary competencies<br />
Manufacturing outsourcing<br />
Increasing focus <strong>on</strong> specific competencies<br />
Valuati<strong>on</strong><br />
News<br />
Entering a new product line or line of business<br />
Entering a new market, geographic or offering br<strong>and</strong> leverage<br />
Capital access investment<br />
Pre-emptive move<br />
Market spanning<br />
While such a list might never be complete, most factors for alliances can be dominantly<br />
<strong>and</strong> usefully categorised into <strong>on</strong>e, two or three of the types presented. The tax<strong>on</strong>omy need<br />
not be exhaustive to be effective. The point is that these three categories – Network<br />
Ec<strong>on</strong>omics, Competencies <strong>and</strong> Market Structure (NCM) – provide the dimensi<strong>on</strong>s that<br />
most powerfully present implicati<strong>on</strong>s for the creati<strong>on</strong> <strong>and</strong> prosecuti<strong>on</strong> of network<br />
strategy. Competencies refer to the internal decisi<strong>on</strong>s a firm makes regarding the<br />
competencies it develops in-house, which competencies it chooses to source through<br />
partners, <strong>and</strong> which competencies it might leave for the market to provide. Market<br />
structure refers to the envir<strong>on</strong>mental, instituti<strong>on</strong>al <strong>and</strong> competitive factors of a particular<br />
industry, market <strong>and</strong>/or ec<strong>on</strong>omy. Network ec<strong>on</strong>omics refers to whether network<br />
ec<strong>on</strong>omic phenomena significantly influence the dynamics of a particular industry or<br />
market under c<strong>on</strong>siderati<strong>on</strong>. Network ec<strong>on</strong>omics includes all of the issues traditi<strong>on</strong>ally<br />
associated with this field of ec<strong>on</strong>omics, such as dem<strong>and</strong>-side ec<strong>on</strong>omies of scale, network<br />
effects <strong>and</strong> positive <strong>and</strong> negative feedback loops (Shapiro <strong>and</strong> Varian, 1999; Varian,<br />
Farrell <strong>and</strong> Shapiro, 2004). In such envir<strong>on</strong>ments where fixed costs are high, variable<br />
costs low <strong>and</strong> decreasing <strong>on</strong> scale, dem<strong>and</strong> is volatile <strong>and</strong> competiti<strong>on</strong> is keen, profitable<br />
growth can be achieved <strong>on</strong>ly with sophisticated price discriminating strategies designed<br />
to help you provide the right product, at the right time, for the right customer, at the right<br />
price, that is <strong>on</strong> ‘mass customisati<strong>on</strong>’, as a major focus of revenue management.<br />
208
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
HANS W. GOTTINGER<br />
% B1.
HANS W. GOTTINGER<br />
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
12% ;?@$%A1##5"-(=%<br />
the firm’s effective m<strong>on</strong>opoly. Any government m<strong>and</strong>ated m<strong>on</strong>opoly falls into this<br />
category as well, but can also overlap with other categories, as in the case of nati<strong>on</strong>al<br />
telecommunicati<strong>on</strong>s m<strong>on</strong>opolies (network ec<strong>on</strong>omics) or nati<strong>on</strong>al c<strong>on</strong>trol of raw<br />
materials mining <strong>and</strong> export. In the case of an oligopoly, existing players might, officially<br />
or unofficially, ally to maintain the status quo, as in the case of a cartel. The opposite can<br />
also be true. In highly disaggregated markets, firms might decide to ally in order to create<br />
ec<strong>on</strong>omies of scale, lowering costs <strong>and</strong> developing bargaining power against suppliers.<br />
Ec<strong>on</strong>omists use metrics such as the Herfindahl Index or a top four-firm c<strong>on</strong>centrati<strong>on</strong><br />
ratio to represent market c<strong>on</strong>centrati<strong>on</strong>.<br />
A literature exists <strong>on</strong> the implicati<strong>on</strong>s of the level of market c<strong>on</strong>centrati<strong>on</strong> in the<br />
industrial organisati<strong>on</strong> literature. The purpose here is not to develop <strong>and</strong> test many<br />
possible hypotheses related to market c<strong>on</strong>centrati<strong>on</strong> <strong>and</strong> alliance formati<strong>on</strong>, but rather to<br />
underscore that market structure plays a role in motivating the creati<strong>on</strong> of networks of<br />
firms, which holds the implicati<strong>on</strong>s for network strategy.<br />
Market structure dimensi<strong>on</strong>s are not limited to m<strong>on</strong>opolistic or oligopolistic<br />
c<strong>on</strong>diti<strong>on</strong>s. As evident from the examples presented earlier, the ‘market structure’<br />
category includes instituti<strong>on</strong>al influences such as government regulati<strong>on</strong> <strong>and</strong> anti-trust<br />
issues that might motivate inter-organisati<strong>on</strong>al arrangements over acquisiti<strong>on</strong> or market<br />
relati<strong>on</strong>ships. Here we define ‘instituti<strong>on</strong>al’ in the terms developed by North in the<br />
ec<strong>on</strong>omic field of instituti<strong>on</strong>al analysis (North, 1990). These instituti<strong>on</strong>al issues relate to<br />
the market structure category of the tax<strong>on</strong>omy, given the fact that they influence the<br />
organisati<strong>on</strong>al decisi<strong>on</strong>s between firms (i.e. to integrate or ally) by impacting competitive<br />
market c<strong>on</strong>diti<strong>on</strong>s, providing rules of the game. In this sense, it is not necessary to<br />
separate market structure <strong>and</strong> instituti<strong>on</strong>al issues into separate categories. The insights<br />
turn out <strong>on</strong> close inspecti<strong>on</strong> to be similar.<br />
H$M% J=5)(=C%D1=%,225,"7(CK%71.
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
HANS W. GOTTINGER<br />
% B1.6-%&48.3>*1#3-1(8%*6%*#*1$3158'7&15*8-G*%8$3>*8<br />
Firm c<strong>on</strong>stellati<strong>on</strong>s change over time. Sometimes, this occurs as a natural process of<br />
growth, or a deepening of c<strong>on</strong>tact between firms. Other times, participant firm objectives<br />
diverge. By definiti<strong>on</strong>, the NCM tax<strong>on</strong>omy underscores the influences that motivate<br />
firms to form alliances. As those influences change <strong>on</strong> a given c<strong>on</strong>stellati<strong>on</strong>, so too will<br />
their nature <strong>and</strong> c<strong>on</strong>structi<strong>on</strong>. The ec<strong>on</strong>omics literature does address firm life cycles when<br />
211
HANS W. GOTTINGER<br />
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
14% ;?@$%A1##5"-(=%<br />
c<strong>on</strong>sidering l<strong>on</strong>g-term decisi<strong>on</strong>s. In these cases, ec<strong>on</strong>omists c<strong>on</strong>sider ‘entry <strong>and</strong> exit’" of<br />
firms over the course of the term in questi<strong>on</strong>.<br />
N<strong>on</strong>etheless, the assumpti<strong>on</strong>s of firm perpetuity, both by researchers <strong>and</strong><br />
practiti<strong>on</strong>ers, can be misleading. Rarely do firms last more than a few generati<strong>on</strong>s. Those<br />
that do, usually undergo significant change over time. While firms like Ford Motor<br />
Company c<strong>on</strong>tinue to create value based <strong>on</strong> the original purpose for the company<br />
(automobile producti<strong>on</strong>), other venerable firms compete in completely different markets<br />
to their predecessors. Nokia’s current dominance of the wireless h<strong>and</strong>set <strong>and</strong><br />
telecommunicati<strong>on</strong>s industries bears very little relati<strong>on</strong>ship with its past, which has<br />
included l<strong>on</strong>g periods in forest products, paper <strong>and</strong> pulp, <strong>and</strong> as a c<strong>on</strong>glomerate with<br />
many unrelated lines of business. Can the Nokia of 1865 , the year of its founding, be<br />
c<strong>on</strong>sidered the same firm as today? So, despite any legal c<strong>on</strong>tinuity (<strong>and</strong> a company’s<br />
legal status can change), firms change markedly over time. Given that firms are by no<br />
means perpetual, researchers should treat them as entities finite in time <strong>and</strong> seek to<br />
underst<strong>and</strong> firm lifespans <strong>and</strong> growth cycles (Jovanovic, 1982; Greiner, 1972).<br />
Similarly, hybrid forms should be treated as arrangements in flux, developed in<br />
resp<strong>on</strong>se to, <strong>and</strong> as a result of, participant firms’ motivati<strong>on</strong>s <strong>and</strong> changing marketplace<br />
c<strong>on</strong>diti<strong>on</strong>s. The tax<strong>on</strong>omy provides varied insights into the life cycles of firm<br />
c<strong>on</strong>stellati<strong>on</strong>s. C<strong>on</strong>sortia created to motivate st<strong>and</strong>ards development in an emerging<br />
industry, at the interface of network ec<strong>on</strong>omics <strong>and</strong> competencies dimensi<strong>on</strong>s, would be<br />
likely to migrate further to the network ec<strong>on</strong>omics after the st<strong>and</strong>ard has been developed<br />
<strong>and</strong> disseminated. Once the st<strong>and</strong>ards race has been w<strong>on</strong> (not necessarily by the<br />
c<strong>on</strong>sortia), the motivati<strong>on</strong> for the alliance recedes, <strong>and</strong> the c<strong>on</strong>sortia is likely to be<br />
disb<strong>and</strong>ed unless alternative benefits have been developed between member firms. By<br />
c<strong>on</strong>trast, alliances in the competencies dimensi<strong>on</strong> often take <strong>on</strong>e of the two predominant<br />
directi<strong>on</strong>s, assuming the competencies that motivated the alliances c<strong>on</strong>tinue to be<br />
perceived as useful by the participants. Either firms in a competency-based alliance build<br />
integrated, mutually dependent competencies, which may encourage alliance l<strong>on</strong>gevity,<br />
or <strong>on</strong>e or more of the participant firms decides to disengage from the alliance, perhaps<br />
having changed strategic priorities or having acquired the competency in house. Often,<br />
firms form alliances to develop a competency in-house, with the assistance of a partner.<br />
Sometimes firms do so avowedly, while in other cases knowledge appropriati<strong>on</strong> occurs in<br />
a more surreptitious manner. Firms can participate in an alliance until they have acquired<br />
necessary knowledge from their partner, <strong>and</strong> then defect from cooperati<strong>on</strong>. Microsoft has<br />
become notorious for this, particularly with small firms. Andy Groves’ famous<br />
‘paranoia’” (Groves, 1996) has been part of the reas<strong>on</strong> Intel’s partnership with Microsoft<br />
has endured successfully for so l<strong>on</strong>g. Intel has maintained a healthy suspici<strong>on</strong> regarding<br />
its alliance with Microsoft. If <strong>on</strong>e or more of the alliance members’ strategic interest, in<br />
particular competencies changes, competency alliances are most likely to be affected.<br />
Market structure provides the most obvious example of how the factors in each<br />
motivati<strong>on</strong> space can impact alliance evoluti<strong>on</strong>. Government regulati<strong>on</strong>s can very quickly<br />
upset the competitive balances between c<strong>on</strong>stellati<strong>on</strong>s of firms, as has occurred in most<br />
instances of de-regulati<strong>on</strong> or privatisati<strong>on</strong>. The massive, global airline alliances (such as<br />
One World <strong>and</strong> Star Alliance), fit in the dimensi<strong>on</strong>s Network Ec<strong>on</strong>omics <strong>and</strong> Market<br />
Structure. Airline de-regulati<strong>on</strong> in the USA <strong>and</strong> elsewhere exerted a significant impact<br />
during the 1980s <strong>and</strong> 1990s, leading to a proliferati<strong>on</strong> of airlines. Moreover, network<br />
ec<strong>on</strong>omics heavily impacts the airline industry, characterised by communicating<br />
networks of routes, hub-<strong>and</strong>-spoke methods though anti-trust regulati<strong>on</strong>s c<strong>on</strong>tinue to exist<br />
212
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
HANS W. GOTTINGER<br />
% B1.*1#3-18$-8G&%3-;#831.;#$%3*#8<br />
The network formati<strong>on</strong> dimensi<strong>on</strong> can be applied to high-tech industries to assist in<br />
underst<strong>and</strong>ing the evoluti<strong>on</strong> of c<strong>on</strong>stellati<strong>on</strong>s of firms <strong>and</strong> network strategy under varied<br />
c<strong>on</strong>diti<strong>on</strong>s. As an example, the dimensi<strong>on</strong> applies to underst<strong>and</strong> the evoluti<strong>on</strong> of alliances<br />
within the biotechnology <strong>and</strong> pharmaceutical industries. The analysis will provide a proof<br />
of principle of the efficacy of the dimensi<strong>on</strong>, as well as an example of its applicati<strong>on</strong>.<br />
A cursory examinati<strong>on</strong> of industry network strategy might suggest that specific<br />
dimensi<strong>on</strong> types dominate in given industries. Most clearly, c<strong>on</strong>stellati<strong>on</strong>s of firms in<br />
industries dominated by network ec<strong>on</strong>omics, such as telecommunicati<strong>on</strong>s equipment or<br />
the airlines industries, would likely coalesce within the network ec<strong>on</strong>omics space.<br />
Industries dominated by the creati<strong>on</strong> of new knowledge, such as those substantially<br />
involved in commercialising emerging technologies (e.g. biotechnology, optical<br />
computing), would favour the competencies dimensi<strong>on</strong> space.<br />
While these distincti<strong>on</strong>s make intuitive sense, <strong>and</strong> further validate the applicati<strong>on</strong> of<br />
the tax<strong>on</strong>omy, any purely static applicati<strong>on</strong> of the tax<strong>on</strong>omy neglects a good deal of its<br />
explanatory power. Networks of firms, <strong>and</strong> network strategy in particular, must be<br />
understood dynamically. Strategy aims to change or maintain market relati<strong>on</strong>ships to the<br />
advantage of a particular firm or group of firms, relative to competitors (<strong>and</strong> in many<br />
cases relative to partners as well). Transforming market relati<strong>on</strong>ships for competitive<br />
advantage, as well as attempting to sustain an advantage, requires an underst<strong>and</strong>ing of<br />
how <strong>and</strong> why the relati<strong>on</strong>ships between firms change over time. While a particular<br />
category of the tax<strong>on</strong>omy might dominate an industry during a given period, examinati<strong>on</strong><br />
over time will undoubtedly uncover shifting influences. A network strategist must<br />
become adept at m<strong>on</strong>itoring industry developments critical to alliance formati<strong>on</strong>,<br />
recognising these changing c<strong>on</strong>diti<strong>on</strong>s early <strong>on</strong>, <strong>and</strong> quickly developing resp<strong>on</strong>ses to<br />
maintain as advantageous a positi<strong>on</strong> as possible. An effective network strategy should<br />
maintain a vigilant eye for events substantially impacting any of the dimensi<strong>on</strong> spaces<br />
within or between industries, or throughout a sector or ec<strong>on</strong>omy-wide. Such events can<br />
transform the relati<strong>on</strong>ships between firms, <strong>and</strong> can provide alert firms with the<br />
opportunity to drive transformati<strong>on</strong>s to their own advantage, or at least to successfully<br />
resp<strong>on</strong>d.<br />
213
HANS W. GOTTINGER<br />
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
16% ;?@$%A1##5"-(=%<br />
Both biotechnology <strong>and</strong> large pharmaceutical firms compete in an industry<br />
characterised by rapid technology change; in particular, these firms depend <strong>on</strong> the<br />
creati<strong>on</strong> of new knowledge. Alliance competencies should be prevalent in any market<br />
characterised by fast-changing, intangible assets, given the difficulties inherent in trading<br />
intangibles; moreover, in industries with very high rates of technology change,<br />
technologies can be introduced that create new market segments, obsolesce existing<br />
product lines <strong>and</strong> create substantial competitors from previously little-known firms.<br />
Under such c<strong>on</strong>diti<strong>on</strong>s, few firms can afford to c<strong>on</strong>duct research in enough directi<strong>on</strong>s to<br />
build sufficient R&D opti<strong>on</strong>s. Alliances offer opportunities for firms, in essence, to<br />
outsource R&D efforts, creating opti<strong>on</strong>s <strong>on</strong> knowledge development, without requiring<br />
merger or acquisiti<strong>on</strong>. Biotechnology provides a critical example of such intangible asses<br />
that can be difficult to trade, <strong>and</strong> are, as well, not necessarily appropriate c<strong>and</strong>idates for<br />
acquisiti<strong>on</strong>. Additi<strong>on</strong>ally, under c<strong>on</strong>diti<strong>on</strong>s of fast change <strong>and</strong> high uncertainty, the<br />
network forms of governance provide preferred access to informati<strong>on</strong>, decreasing<br />
informati<strong>on</strong> asymmetries <strong>and</strong> allowing firms involved in a network to scan a broader<br />
envir<strong>on</strong>ment.<br />
/*0*%*1'*#8<br />
Applegate, L.M. <strong>and</strong> M<strong>on</strong>tealegre, R. (1995) N,C#.,"%O1:,G%B1$K%+,",-5"-%!"D1=.,#51"%>9C#(.C%<br />
#8=1*-8%>#=,#(-57%P225,"7(C. Bost<strong>on</strong>: Harvard Business School Press.<br />
Argyres, N.S. <strong>and</strong> Silverman, B.S. (2004) ‘R&D, organizati<strong>on</strong>al structure <strong>and</strong> the development of<br />
corporate technological knowledge’, >#=,#(-57%+,",-(.("#%&1*=",2, Vol. 25, pp.929–958.<br />
Besanko, D., Dranove, D. <strong>and</strong> Shanley, M. (2000) N71"1.57C% 1D% >#=,#(-9 (2nd ed.). New York:<br />
Wiley.<br />
Ch<strong>and</strong>ler, A.D. (1962) >#=,#(-9%,":%>#=*7#*=(. Cambridge, MA: MIT Press.<br />
Ch<strong>and</strong>ler, A.D. (1990) >7,2(% ,":% >71#=,#(-9. New York: Free Press.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H.W. (2003) N71"1.5(C%1D%4(#F1=G%!":*C#=5(C. L<strong>on</strong>d<strong>on</strong>: Routledge.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H.W. (2004) ‘Biotech <strong>and</strong> pharma industries: applicati<strong>on</strong> of network ec<strong>on</strong>omies’,<br />
R8,=.,S(*#5C78(%!":*C#=5(, Vol. 66, pp.1475–1478.<br />
Grant, R.M. (2001) B1"#(.#=,#(-9%P",29C5C. Oxford: Blackwell.<br />
Greiner, L. (1972) ‘Evoluti<strong>on</strong> <strong>and</strong> revoluti<strong>on</strong> as organizati<strong>on</strong>s grow’, ;,=),=:% T*C5"(CC% '()5(F,<br />
Vol. 50, pp.37–46.<br />
Groves, A. (1996) U"29%#8(%R,=,"15:%>*=)5)(. New York: Nort<strong>on</strong>.<br />
Gulati, R. (1999) ‘Network locati<strong>on</strong> <strong>and</strong> learning. The influence of network resources <strong>and</strong> firm<br />
capabilities <strong>on</strong> alliance formati<strong>on</strong>’, >#=,#(-57%+,",-(.("#%&1*=",2, Vol. 20, pp.397–420.<br />
214
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
HANS W. GOTTINGER<br />
% B1.#=,#(-57%<br />
+,",-(.("#%&1*=",2, Vol. 12, pp.105–124.<br />
North, D.C. (1990) ‘!"C#5#*#51"C/% !"C#5#*#51",2% B8,"-(% ,":% N71"1.57% R(=D1=."7(. Cambridge:<br />
Cambridge Univ. Press.<br />
Oxley, J.E. <strong>and</strong> Samps<strong>on</strong>, R.C. (2004) ‘The scope of governance of internati<strong>on</strong>al R&D alliances’,<br />
>#=,#(-57%+,",-(.("#%&1*=",2, Vol. 25, pp.723–749.<br />
Porter, M.E. (1980) B1.#=,#(-9. New York: Free Press.<br />
Shapiro, C. <strong>and</strong> Varian, H.R. (1999) !"D1=.,#51"% '*2(C. Bost<strong>on</strong>, MA: Harvard Business School<br />
Press.<br />
Sim<strong>on</strong>, H. (1947) P:.5"C#=,#5)(%T(8,)51=. New York: Macmillan.<br />
Stinchcombe, A.L. (1990) !"D1=.,#51"%,":%U=-,"5S,#51"C$ Berkeley, CA: University of California<br />
Press.<br />
Varian, H., Farrell, J. <strong>and</strong> Shapiro, C. (2004) 68(%N71"1.57C%1D%!"D1=.,#51" 6(78"121-9. Raffaele<br />
Mattioli Lectures, Cambridge: Cambridge University Press.<br />
Williams<strong>on</strong>, O.E. (1975) +,=G(#C%,":%;5(=,785(C/%P",29C5C%1D%P"#5#=*C#%!.
HANS W. GOTTINGER<br />
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
216
3.<br />
GROWTH<br />
AND<br />
DEVELOPMENT
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Supplementing: Strategies of Ec<strong>on</strong>omic Growth <strong>and</strong> Catch-Up . H.W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g> <strong>and</strong> M. Goosen,<br />
eds., NovaScience, New York, 2012<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, STRATEC Germany<br />
www.stratec-c<strong>on</strong>.net<br />
A Comprehensive Review of the Microec<strong>on</strong>omic Foundati<strong>on</strong>s of Ec<strong>on</strong>omic Growth<br />
This review provides a summary of the main findings of the literature review (which, as<br />
supporting document for the text “Strategies of Ec<strong>on</strong>omic Growth <strong>and</strong> Catchup”, is attached<br />
separately ). The secti<strong>on</strong> then provides a brief discussi<strong>on</strong> of key theories <strong>and</strong> c<strong>on</strong>cepts that we drew<br />
from the literature<br />
Summary Findings from the Literature<br />
Technological innovati<strong>on</strong> has an important role to play both as part of the competitive process <strong>and</strong><br />
as a driver of ec<strong>on</strong>omic growth <strong>and</strong> development. However, perhaps surprisingly, for a l<strong>on</strong>g time<br />
innovati<strong>on</strong> was c<strong>on</strong>sidered a field where ec<strong>on</strong>omists, as opposed to other scientists <strong>and</strong><br />
engineers, <strong>on</strong>ly viewed this as a ‘black box’.. The forces underlying the process of technological<br />
change were believed to be substantially independent of ec<strong>on</strong>omic incentives <strong>and</strong> mostly affected by<br />
the exogenous evoluti<strong>on</strong> of scientific knowledge <strong>and</strong> its technological applicati<strong>on</strong>s.<br />
It has been <strong>on</strong>ly recently, mainly over the last two decades, that the ec<strong>on</strong>omic forces behind<br />
technological innovati<strong>on</strong> have started to be investigated in more detail within mainstream<br />
ec<strong>on</strong>omics. As a result, a wide variety of theories <strong>and</strong> models, sometimes very diverse in spirit,<br />
describing the ec<strong>on</strong>omics of innovati<strong>on</strong> are now available. All these theories share the comm<strong>on</strong><br />
aim of providing a c<strong>on</strong>ceptual foundati<strong>on</strong> for underst<strong>and</strong>ing how innovati<strong>on</strong> affects the ec<strong>on</strong>omy,<br />
how ec<strong>on</strong>omic forces affect the emergence of technological changes, <strong>and</strong> the decisi<strong>on</strong>-making<br />
processes through which technological innovati<strong>on</strong> occurs.<br />
We have made no attempt to provide the most representative <strong>and</strong> comprehensive summaries of<br />
each str<strong>and</strong> of literature; this is not our goal. Rather it is to give a sufficient summary to enable<br />
subsequent identificati<strong>on</strong> of how the study can be taken forward to develop analytical tools for<br />
assessing market dynamics in competiti<strong>on</strong> policy investigati<strong>on</strong>s.<br />
This secti<strong>on</strong> provides a brief summary of some key elements of the different str<strong>and</strong>s of literature<br />
overviewed in much of this work.<br />
Industrial organizati<strong>on</strong><br />
Studies of dynamic competiti<strong>on</strong> within ec<strong>on</strong>omics have developed from <strong>on</strong>ce<br />
inherently static neoclassical ec<strong>on</strong>omic theories to modern analyses of innovati<strong>on</strong> <strong>and</strong> dynamic<br />
competiti<strong>on</strong>.<br />
N<strong>on</strong>etheless, a str<strong>on</strong>g relati<strong>on</strong>ship with neoclassical ec<strong>on</strong>omics remains in the general<br />
methodological approach adopted in these theories, which are characterized by the analysis of<br />
equilibrium models where fully rati<strong>on</strong>al ec<strong>on</strong>omic agents make optimizing choices.<br />
Generally, traditi<strong>on</strong>al models of innovati<strong>on</strong> focus <strong>on</strong> the study of firms’ incentives to invest<br />
resources in Research <strong>and</strong> <strong>Development</strong> (R&D) activities. Game-theoretical models developed<br />
in the industrial organizati<strong>on</strong> literature have investigated firms’ R&D decisi<strong>on</strong>s in strategic<br />
1<br />
219
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
envir<strong>on</strong>ments. (Endogenous growth models, discussed in the next sub-secti<strong>on</strong>, have developed<br />
the study of market dynamics in models that explain the relati<strong>on</strong>ship between firms’ investments,<br />
innovati<strong>on</strong> <strong>and</strong> ec<strong>on</strong>omic growth.)<br />
Game-theoretical models suggest that there are two main forces that underlie firms’ investment in<br />
R&D: the search for higher profits <strong>and</strong> the threat posed by falling behind potential innovating<br />
rivals. Game theoretical models study these forces in a variety of market situati<strong>on</strong>s <strong>and</strong> address<br />
issues such as the interplay between innovati<strong>on</strong> <strong>and</strong> market structure, the dynamics of<br />
competiti<strong>on</strong> <strong>and</strong> the nature of the relati<strong>on</strong>ship between intensity of competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong>.<br />
These models provide a rich picture of what the possible strategies <strong>and</strong> industry equilibria in<br />
dynamic markets can be. However, general predicti<strong>on</strong>s, that can be c<strong>on</strong>sidered appropriate<br />
across all situati<strong>on</strong>s <strong>and</strong> industries, are scarce. On <strong>on</strong>e h<strong>and</strong>, this is a result that seems to stress<br />
the lack of predictive power of these models, i.e. (almost) everything can be rati<strong>on</strong>alized; <strong>on</strong> the<br />
other h<strong>and</strong>, however, the variety of results seems well to fit with the variety of observed<br />
behaviours <strong>and</strong> “equilibria”. There is no general model that can uncritically be applied to any<br />
case: the underst<strong>and</strong>ing of the specific characteristics of the single situati<strong>on</strong> needs to underpin<br />
any appropriate choice of a modelling framework.<br />
Despite the absence of general results, these models are certainly useful tools to underst<strong>and</strong><br />
firms’ incentives to invest in R&D activities in strategic envir<strong>on</strong>ments <strong>and</strong> to suggest what main<br />
factors may be central in shaping the nature of dynamic competiti<strong>on</strong>. For instance, these models<br />
suggest that:<br />
• In order to underst<strong>and</strong> R&D investments in strategic envir<strong>on</strong>ments it is necessary to<br />
underst<strong>and</strong> how innovati<strong>on</strong> may affect profits both of successful <strong>and</strong> n<strong>on</strong>-successful<br />
innovators. The first perspective captures the idea that firms want to innovate to increase<br />
their profits; the sec<strong>on</strong>d captures the idea that firms want to innovate to maintain<br />
competitiveness.<br />
•<br />
•<br />
The relati<strong>on</strong>ship between c<strong>on</strong>centrati<strong>on</strong> of an industry <strong>and</strong> its rate of technological<br />
innovati<strong>on</strong> is certainly complex <strong>and</strong> in general not a causal <strong>on</strong>e: both should be thought of<br />
as the outcomes of the operati<strong>on</strong> of market forces <strong>and</strong> exogenous factors such as the<br />
nature of dem<strong>and</strong>, technological opportunity <strong>and</strong> the c<strong>on</strong>diti<strong>on</strong>s governing appropriability.<br />
Dynamic competiti<strong>on</strong> may be characterized by persistent dominance of the incumbent<br />
leader or by acti<strong>on</strong>-reacti<strong>on</strong> whereby incumbents are overtaken by a rival whose<br />
incumbency is itself then short-lived. The nature of market dynamics depends <strong>on</strong> a<br />
number of factors, such as the type of innovati<strong>on</strong>, i.e. drastic or n<strong>on</strong>-drastic, the<br />
uncertainties involved in R&D activities, the nature of patent protecti<strong>on</strong> <strong>and</strong> of knowledge<br />
spill-overs, the intensity of product market competiti<strong>on</strong>, etc.<br />
•<br />
When the relati<strong>on</strong>ship between competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> is investigated, it is necessary<br />
to be clear what the noti<strong>on</strong> of intensity of competiti<strong>on</strong> describes <strong>and</strong> how this relates (or<br />
does not relate) to market structure. Indeed a market where competiti<strong>on</strong> is tougher may<br />
be more c<strong>on</strong>centrated simply because inefficient firms cannot survive. There may be a<br />
2<br />
220
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
trade-off between the intensity of static competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong>. In general, the<br />
relati<strong>on</strong>ship between intensity of competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> need not be m<strong>on</strong>ot<strong>on</strong>ic at all.<br />
Endogenous growth models have recently developed al<strong>on</strong>g the earlier game theoretical literature <strong>on</strong><br />
innovati<strong>on</strong> in the c<strong>on</strong>text of studies that seek to explain the relati<strong>on</strong>ship between innovati<strong>on</strong> <strong>and</strong><br />
ec<strong>on</strong>omic growth. These models suggest that innovati<strong>on</strong>, resulting from intenti<strong>on</strong>al R&D<br />
investments by profit-maximizing firms or simply by unintenti<strong>on</strong>al learning-by-doing, is a<br />
fundamental driver of ec<strong>on</strong>omic growth in the l<strong>on</strong>g run.<br />
Early Schumpeterian endogenous growth models stressed the importance of ex-post rents for<br />
innovati<strong>on</strong>: competiti<strong>on</strong> would have a detrimental effect <strong>on</strong> innovati<strong>on</strong> by decreasing the rents that<br />
an innovator would be able to appropriate. More recent models have emphasized another<br />
mechanism by which competiti<strong>on</strong> affects innovati<strong>on</strong>: tougher competiti<strong>on</strong> may increase the<br />
incentives of firms to innovate in order to escape from fierce competiti<strong>on</strong>. These recent studies<br />
suggest that the relati<strong>on</strong>ship between competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> may not well be m<strong>on</strong>ot<strong>on</strong>ic <strong>and</strong><br />
that instead, <strong>on</strong>e should expect an inverse-U shaped relati<strong>on</strong>ship: when competiti<strong>on</strong> is low, an<br />
increase in competiti<strong>on</strong> would foster innovati<strong>on</strong>; the reverse would happen when competiti<strong>on</strong> is<br />
fierce.<br />
The result that competiti<strong>on</strong> may be c<strong>on</strong>ducive to innovati<strong>on</strong> is also obtained in studies where the<br />
traditi<strong>on</strong>al behavioral assumpti<strong>on</strong> of profit-maximizing firms is relaxed. When principal-agent<br />
c<strong>on</strong>siderati<strong>on</strong>s are introduced to explain managers’ behaviors, another mechanism by which<br />
competiti<strong>on</strong> may favor innovati<strong>on</strong> is suggested: the speed of innovati<strong>on</strong> may be retarded by the<br />
slack of managers who tend to avoid private costs associated with innovati<strong>on</strong>. When competiti<strong>on</strong><br />
intensifies, the higher threat of bankruptcy may force managers to speed up the process at which<br />
new ideas are adopted. Hence, competiti<strong>on</strong> may be c<strong>on</strong>ducive to faster rates of innovati<strong>on</strong>.<br />
New ec<strong>on</strong>omic geography<br />
New ec<strong>on</strong>omic geography is a branch of ec<strong>on</strong>omics which is mainly c<strong>on</strong>cerned with spatial<br />
aspects of ec<strong>on</strong>omics. In particular, it seeks to explain why <strong>and</strong> how given ec<strong>on</strong>omic activities<br />
c<strong>on</strong>centrate geographically, either within individual countries (agglomerati<strong>on</strong>s) or between<br />
different countries (industrial clustering). Furthermore, it c<strong>on</strong>siders the inter-relati<strong>on</strong>ships between<br />
geographical c<strong>on</strong>centrati<strong>on</strong> of industry, internati<strong>on</strong>al trade <strong>and</strong> ec<strong>on</strong>omic development.<br />
At a broad level, several themes emerge from the literature.<br />
•<br />
•<br />
Geographical c<strong>on</strong>centrati<strong>on</strong> of ec<strong>on</strong>omic activity will occur where transport costs are low<br />
enough <strong>and</strong> the labor force is mobile.<br />
Key characteristics of a particular industry have also an important role to play. The higher<br />
the scope for vertical integrati<strong>on</strong>, the higher the share of intermediate goods in output <strong>and</strong><br />
the lower the labor intensity are, the higher will be the drive for c<strong>on</strong>centrati<strong>on</strong>.<br />
3<br />
221
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Industrial c<strong>on</strong>centrati<strong>on</strong> implies the possibility of geographical proximity of firms that lays<br />
the foundati<strong>on</strong> for technological cooperati<strong>on</strong> <strong>and</strong> thus lead to increased innovative<br />
activity.<br />
•<br />
A high degree of c<strong>on</strong>centrati<strong>on</strong> leads to very large markets — <strong>and</strong> the larger the size of<br />
the market, the higher the potential reward of successful innovati<strong>on</strong>s.<br />
The evoluti<strong>on</strong>ary ec<strong>on</strong>omics of innovati<strong>on</strong><br />
The evoluti<strong>on</strong>ary approach to the study of innovati<strong>on</strong> has been developed <strong>on</strong> very different<br />
methodological basis than those underlying traditi<strong>on</strong>al ec<strong>on</strong>omic models of innovati<strong>on</strong>. In<br />
particular, we observe the rejecti<strong>on</strong> of the modelling assumpti<strong>on</strong>s of rati<strong>on</strong>ality <strong>and</strong> equilibrium<br />
that are fundamental to the traditi<strong>on</strong>al approach.<br />
Evoluti<strong>on</strong>ary ec<strong>on</strong>omics looks from the outset at dynamic processes. In particular, it is associated<br />
with the use of analogies from evoluti<strong>on</strong>ary biology to explain ec<strong>on</strong>omic growth <strong>and</strong> the process<br />
of competiti<strong>on</strong>. Thus the cornerst<strong>on</strong>es of an evoluti<strong>on</strong>ary analysis of competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong><br />
are variety, selecti<strong>on</strong> <strong>and</strong> imitati<strong>on</strong>.<br />
At a basic level, using Darwinian analogies, we can begin to appreciate the role of the market in<br />
selecting the more fit firms (efficient <strong>and</strong> profitable), products <strong>and</strong> techniques at the expense of<br />
less fit firms, products or techniques. In additi<strong>on</strong> to this effect, we would expect to see imitati<strong>on</strong> of<br />
winning ideas by those whose survival is otherwise threatened (although this is limited by the tacit<br />
nature of knowledge).<br />
Inherent in this model of competiti<strong>on</strong> is the associati<strong>on</strong> between competiti<strong>on</strong> <strong>and</strong> experimentati<strong>on</strong><br />
<strong>and</strong> variety. A variety of experiments allows, through the process of selecti<strong>on</strong>, for greater<br />
ec<strong>on</strong>omic progress than would be available through uniform optimizati<strong>on</strong>.<br />
Thus ec<strong>on</strong>omic development <strong>and</strong> innovati<strong>on</strong> can be seen as a combined effect of selecti<strong>on</strong> (via<br />
competiti<strong>on</strong>) from a variety of competing routines <strong>and</strong> practices as well as the more endogenous<br />
process of agents seeking improved routines <strong>and</strong> practices. While the latter is certainly<br />
incorporated, if treated somewhat differently, under the mainstream neoclassical approach, the<br />
emphasis <strong>on</strong> selecti<strong>on</strong> from variety seems an important additi<strong>on</strong> to this approach.<br />
This suggests that models found in other branches of ec<strong>on</strong>omics might miss something important<br />
when they analyze dynamics with reference to homogenous profit-maximizing firms: namely the<br />
benefits of selecti<strong>on</strong> from heterogeneity in capabilities <strong>and</strong> innovative experimentati<strong>on</strong>.<br />
Systems of innovati<strong>on</strong><br />
At an aggregate level, the evoluti<strong>on</strong>ary approach to the study of ec<strong>on</strong>omic growth draws attenti<strong>on</strong><br />
to the importance of instituti<strong>on</strong>s in the process of ec<strong>on</strong>omic growth. The key findings of the<br />
literature are as follows:<br />
• Innovati<strong>on</strong> <strong>and</strong> its diffusi<strong>on</strong> take place within systems of interc<strong>on</strong>nected organizati<strong>on</strong>s <strong>and</strong><br />
instituti<strong>on</strong>s. Important c<strong>on</strong>stituent elements of such systems are organizati<strong>on</strong>s such as<br />
4<br />
222
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
firms, governments <strong>and</strong> universities. Instituti<strong>on</strong>s, with which these organizati<strong>on</strong>s interact,<br />
reflect laws <strong>and</strong> statutes (e.g. the instituti<strong>on</strong> of patent protecti<strong>on</strong>) as well as more abstract<br />
elements (e.g. cultural aspects of the ec<strong>on</strong>omy such as the spirit of entrepreneurial<br />
activity).<br />
•<br />
•<br />
Innovati<strong>on</strong> within the system will depend not <strong>on</strong>ly <strong>on</strong> single instituti<strong>on</strong>s but also <strong>on</strong> the<br />
nature <strong>and</strong> intensity of interacti<strong>on</strong>s between the various elements of the system.<br />
When c<strong>on</strong>sidering the interacti<strong>on</strong> between competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong>, we must remain<br />
aware that the effects of such interacti<strong>on</strong> will depend <strong>on</strong> the evolving instituti<strong>on</strong>al<br />
background against which agents in the ec<strong>on</strong>omy operate.<br />
The overview of the cliometric literature has shown that cliometrics is a methodological framework<br />
for the study of ec<strong>on</strong>omic history. The existing body of cliometrics draws attenti<strong>on</strong> to the following<br />
factors relevant to “systems of innovati<strong>on</strong>”:<br />
• Ec<strong>on</strong>omic growth is a result of a complex interacti<strong>on</strong> between instituti<strong>on</strong>s <strong>and</strong> markets.<br />
Examples of instituti<strong>on</strong>s that are frequently important for innovati<strong>on</strong> are the protecti<strong>on</strong> of<br />
property rights (for example, against piracy) <strong>and</strong> patent policy.<br />
Innovati<strong>on</strong> plays a critical role in ec<strong>on</strong>omic progress <strong>and</strong> development. This involves not<br />
<strong>on</strong>ly technological innovati<strong>on</strong>s but also changes in instituti<strong>on</strong>s. Thus, instituti<strong>on</strong>s<br />
themselves evolve according to changing circumstances. These two last points<br />
complement the c<strong>on</strong>cepts of systems of innovati<strong>on</strong> found in the macroec<strong>on</strong>omic field of<br />
the evoluti<strong>on</strong>ary approach.<br />
5<br />
223
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
INTRODUCTION<br />
1<br />
1.1<br />
1.2<br />
1.3<br />
1.4<br />
2<br />
2.1<br />
2.2<br />
2.3<br />
3<br />
3.1<br />
3.2<br />
3.3<br />
4<br />
4.1<br />
4.2<br />
4.3<br />
4.4<br />
4.5<br />
4.6<br />
ECONOMICS OF INNOVATION<br />
Introducti<strong>on</strong><br />
Industrial Organizati<strong>on</strong><br />
Endogenous Growth Models<br />
New Ec<strong>on</strong>omic Geography<br />
THE EVOLUTIONARY ECONOMICS OF INNOVATION<br />
Introducti<strong>on</strong><br />
Microec<strong>on</strong>omics of Innovati<strong>on</strong><br />
Systems of Innovati<strong>on</strong> <strong>and</strong> the Role of Instituti<strong>on</strong>s<br />
STRATEGIES OF INNOVATIVE FIRMS<br />
Innovati<strong>on</strong> <strong>and</strong> Competitive Advantage<br />
Timing of Innovati<strong>on</strong> <strong>and</strong> Life Cycles<br />
Issues Relating to Network Industries<br />
BIBLIOGRAPHY<br />
Industrial Organizati<strong>on</strong><br />
Endogenous Growth<br />
New Ec<strong>on</strong>omic Geography<br />
Evoluti<strong>on</strong>ary Ec<strong>on</strong>omics<br />
Cliometrics<br />
Strategic Management <strong>and</strong> Networks<br />
224
7<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
INTRODUCTION<br />
This is an overview, with references to the seminal literature <strong>on</strong> the spectrum of dynamic<br />
approaches in micro <strong>and</strong> meso ec<strong>on</strong>omics covering at least the last two decades – the<br />
overview should avoid a pure focus <strong>on</strong> mainstream approaches <strong>and</strong> draw benefits from<br />
overlaps with neighboring disciplines such as systems theory, cliometrics <strong>and</strong><br />
instituti<strong>on</strong>alism.<br />
This is a brief overview <strong>and</strong> classificati<strong>on</strong> based <strong>on</strong> the relevant literature of the most<br />
important ec<strong>on</strong>omic strategies applied by innovative companies to enter new markets or<br />
introduce new products.<br />
Secti<strong>on</strong> 1 presents the survey of mainstream theories <strong>and</strong> models of innovati<strong>on</strong> <strong>and</strong> Secti<strong>on</strong> 2<br />
discusses evoluti<strong>on</strong>ary ec<strong>on</strong>omics, systems theory, cliometrics <strong>and</strong> instituti<strong>on</strong>alism.<br />
Secti<strong>on</strong> 3 focuses instead <strong>on</strong> firms’ strategies related to innovative activities.<br />
225
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
1<br />
1.1<br />
ECONOMICS OF INNOVATION<br />
Introducti<strong>on</strong><br />
The neo-classical approach to innovati<strong>on</strong><br />
In neo-classical theories, producti<strong>on</strong> opportunities are described by specifying a producti<strong>on</strong><br />
functi<strong>on</strong>, which describes the technological relati<strong>on</strong>ship between inputs <strong>and</strong> output. Technological<br />
progress is captured by the ability of a firm, or the ec<strong>on</strong>omy, to produce, for each combinati<strong>on</strong> of<br />
inputs used, more output. However, such progress was for a l<strong>on</strong>g time essentially c<strong>on</strong>sidered<br />
exogenous, as if it were knowledge that falls, like manna, from heaven.<br />
Nevertheless, neo-classical ec<strong>on</strong>omists were c<strong>on</strong>scious that the c<strong>on</strong>siderati<strong>on</strong> of innovati<strong>on</strong> as an<br />
ec<strong>on</strong>omic phenomen<strong>on</strong> posed some challenges to established theories of perfect competiti<strong>on</strong><br />
<strong>and</strong> that a trade-off may exist between static <strong>and</strong> dynamic efficiency.<br />
Innovati<strong>on</strong> is essentially the producti<strong>on</strong> <strong>and</strong> commercial applicati<strong>on</strong> of new knowledge. Neoclassical<br />
ec<strong>on</strong>omists, following the pi<strong>on</strong>eering work of Arrow (1962), have usually treated<br />
knowledge as informati<strong>on</strong>, that is, as a pure n<strong>on</strong>-rivalrous <strong>and</strong> n<strong>on</strong>-excludable good. The<br />
c<strong>on</strong>cepts of n<strong>on</strong>-rivalry <strong>and</strong> n<strong>on</strong>-excludability are crucial in underst<strong>and</strong>ing the neo-classical<br />
ec<strong>on</strong>omic analysis of innovati<strong>on</strong> <strong>and</strong> can be explained as follows:<br />
(a) N<strong>on</strong>-rivalry implies that the same piece of knowledge can be c<strong>on</strong>sumed by different<br />
individuals, the marginal cost involved being essentially zero. Knowledge is<br />
c<strong>on</strong>sidered to be essentially disembodied (as opposed, for instance, to knowledge<br />
embodied in organizati<strong>on</strong>al routines — a central feature of the evoluti<strong>on</strong>ary approach<br />
overviewed in secti<strong>on</strong> 2) <strong>and</strong> available in a comm<strong>on</strong> stock or pool. In other words<br />
knowledge is essentially assimilated to informati<strong>on</strong>.<br />
(b) N<strong>on</strong>-excludability is the inability of the innovator to prevent others from accessing the<br />
knowledge produced. N<strong>on</strong>-excludability arises largely because it may not be feasible,<br />
or it is extremely costly, to m<strong>on</strong>itor <strong>and</strong> charge for the use <strong>and</strong> enjoyment of a piece of<br />
knowledge. In traditi<strong>on</strong>al models of innovati<strong>on</strong>, it is often related to the noti<strong>on</strong> of<br />
knowledge spillovers, which refers to the positive externality by which the knowledge<br />
generated by an agent enhances the technological capabilities of other agents or their<br />
ability to discover new knowledge. N<strong>on</strong>-excludability is not entirely an intrinsic quality<br />
of knowledge <strong>and</strong> depends largely <strong>on</strong> the legal regime in place <strong>and</strong>, in particular, <strong>on</strong><br />
the system of property rights (e.g. patents). Indeed, a patent might allow the innovator<br />
to appropriate the value of an (otherwise n<strong>on</strong>-excludable) innovati<strong>on</strong> by granting the<br />
exclusive right of its commercial use for a certain period of time.<br />
The neo-classical analysis of the ec<strong>on</strong>omic forces underlying the producti<strong>on</strong> of new technological<br />
knowledge follows from the general theoretical framework described above <strong>and</strong> is developed <strong>on</strong><br />
the assumpti<strong>on</strong> that ec<strong>on</strong>omic agents are rati<strong>on</strong>al <strong>and</strong> act to maximize their utility or profit.<br />
When n<strong>on</strong>-rivalry is coupled with n<strong>on</strong>-excludability, <strong>and</strong> both are c<strong>on</strong>sidered to be essential<br />
features of technological knowledge, innovati<strong>on</strong> acquires the characteristics of a pure public good<br />
<strong>and</strong>, accordingly, profit maximizing agents are c<strong>on</strong>sidered not to have incentives to invest in its<br />
producti<strong>on</strong>.<br />
8<br />
226
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
In fact, if the process of imitati<strong>on</strong> of an innovati<strong>on</strong> were immediate, a firm would not have the<br />
incentive to invest resources in R&D activities since ex-post competiti<strong>on</strong> would not allow the<br />
innovator to earn profits to pay back its investment. However, although a single firm would not<br />
have incentives to discover new knowledge, innovati<strong>on</strong> may well be beneficial for society so that<br />
a role for public interventi<strong>on</strong> in the form of direct R&D investment <strong>and</strong>/or appropriate public policy<br />
is usually called for.<br />
These c<strong>on</strong>siderati<strong>on</strong>s suggest that a trade-off may exist between static <strong>and</strong> dynamic efficiency<br />
<strong>and</strong> provide the ec<strong>on</strong>omic rati<strong>on</strong>ale for some important public policies such as public funding of<br />
research <strong>and</strong> the definiti<strong>on</strong> of a patent system that protects the innovator from imitati<strong>on</strong>. The<br />
basic trade-off of a patent system is that between the static costs that derive from the innovator’s<br />
potential m<strong>on</strong>opoly power <strong>and</strong> the l<strong>on</strong>g-term benefits associated with the discovery of the new<br />
knowledge.<br />
Core models of dynamic competiti<strong>on</strong><br />
In the last two decades, the analysis of innovati<strong>on</strong> <strong>and</strong> technological change has attracted much<br />
attenti<strong>on</strong> within mainstream ec<strong>on</strong>omics, due to the development of two closely related disciplines<br />
— game-theoretical industrial organizati<strong>on</strong> <strong>and</strong> endogenous growth theories. As a c<strong>on</strong>sequence,<br />
many of the themes that were <strong>on</strong>ce a feature of the less c<strong>on</strong>venti<strong>on</strong>al theories of technological<br />
advance have now found their way into ec<strong>on</strong>omics.<br />
Studies of dynamic competiti<strong>on</strong> within mainstream ec<strong>on</strong>omics have developed <strong>on</strong>ce inherently<br />
static neoclassical ec<strong>on</strong>omic theories into modern analyses of innovati<strong>on</strong> <strong>and</strong> dynamic<br />
competiti<strong>on</strong>. N<strong>on</strong>etheless, a str<strong>on</strong>g relati<strong>on</strong>ship with neoclassical ec<strong>on</strong>omics remains in the<br />
general methodological approach adopted in these theories, which are characterized by the<br />
analysis of equilibrium models where fully rati<strong>on</strong>al ec<strong>on</strong>omic agents make optimizing choices.<br />
Core models of dynamic competiti<strong>on</strong> are highly stylized <strong>and</strong> this is also reflected in the<br />
definiti<strong>on</strong> of innovati<strong>on</strong> adopted. In these models, innovati<strong>on</strong> is usually defined as the reducti<strong>on</strong> of<br />
producti<strong>on</strong> costs or as the commercializati<strong>on</strong> of new <strong>and</strong>/or better products. Apparently this<br />
definiti<strong>on</strong> identifies innovati<strong>on</strong> with technological change <strong>and</strong> downsizes the importance of other<br />
forms of innovati<strong>on</strong>, such as managerial <strong>and</strong> organizati<strong>on</strong>al changes.<br />
However, notwithst<strong>and</strong>ing the emphasis placed by these models <strong>on</strong> innovati<strong>on</strong> as the commercial<br />
use of new or improved products <strong>and</strong> processes, this stylized definiti<strong>on</strong> of innovati<strong>on</strong> can also be<br />
c<strong>on</strong>sidered to encompass n<strong>on</strong>-technological changes inasmuch these ultimately matter because<br />
result in a reducti<strong>on</strong> of producti<strong>on</strong> costs or new <strong>and</strong> better products.<br />
What mainstream models suggest is that the particular form of innovati<strong>on</strong> (e.g. new technological<br />
knowledge rather than new organizati<strong>on</strong>al knowledge) is not central to the underst<strong>and</strong>ing of firms’<br />
incentives to exert effort to improve their ability to produce efficiently <strong>and</strong> to market new <strong>and</strong> or<br />
better products, which is the main questi<strong>on</strong> addressed in these theories.<br />
9<br />
227
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Similar to the definiti<strong>on</strong> of innovati<strong>on</strong> , also the treatment of the process by which new knowledge is<br />
generated is very stylized in these models. Knowledge is usually c<strong>on</strong>sidered as deriving from<br />
learning-by-doing or from intenti<strong>on</strong>al R&D investments, optimally chosen by profit-maximizing<br />
agents. The R&D process is captured by the definiti<strong>on</strong> of a deterministic or stochastic knowledge<br />
producti<strong>on</strong> functi<strong>on</strong> which related the input used in R&D — notably capital, labor, available<br />
knowledge – to the new knowledge produced <strong>and</strong>/or the timing of innovati<strong>on</strong>.<br />
Game-theoretical models study firms’ incentives to invest resources in R&D activities in strategic<br />
settings <strong>and</strong> address issues such as the interplay between innovati<strong>on</strong> <strong>and</strong> market structure, the<br />
dynamics of competiti<strong>on</strong> <strong>and</strong> the nature of the relati<strong>on</strong>ship between intensity of competiti<strong>on</strong> <strong>and</strong><br />
innovati<strong>on</strong>.<br />
In general, these models are structured <strong>on</strong> the fundamental distincti<strong>on</strong> between static <strong>and</strong><br />
dynamic competiti<strong>on</strong>. Static or product-market competiti<strong>on</strong> refers to the strategic interacti<strong>on</strong> that<br />
arises between firms, taking as given the capabilities of these firms in terms of the variety, quality<br />
<strong>and</strong> producti<strong>on</strong> costs of the products in their portfolios. Dynamic competiti<strong>on</strong> refers instead to the<br />
strategic interacti<strong>on</strong> am<strong>on</strong>g rival firms to change these capabilities.<br />
This distincti<strong>on</strong> is usually captured by the structure of game-theoretical models: firms are<br />
c<strong>on</strong>sidered to play a game where they first invest in innovative activities <strong>and</strong> then compete at the<br />
product-market level. The distincti<strong>on</strong> is a fictitious <strong>on</strong>e to a certain extent, since in reality both<br />
dimensi<strong>on</strong>s of competiti<strong>on</strong> are c<strong>on</strong>tinuously overlapping, but it is a c<strong>on</strong>vincing <strong>and</strong> powerful<br />
c<strong>on</strong>ceptual categorizati<strong>on</strong> since it allows distinguishing activities that occur over different timehoriz<strong>on</strong>s,<br />
i.e. short-term price competiti<strong>on</strong> versus. l<strong>on</strong>g-term dynamic competiti<strong>on</strong>. When firms<br />
compete in the short-term they have to take as given a number of c<strong>on</strong>straints in terms of their <strong>and</strong><br />
their rivals’ “capabilities” that derive from previous innovative activities; when competiti<strong>on</strong> is<br />
c<strong>on</strong>sidered over a l<strong>on</strong>ger-time horiz<strong>on</strong>, however, these “capabilities” become endogenous to the<br />
competitive process as well.<br />
Despite the absence of general results, game-theoretical models are certainly useful to<br />
underst<strong>and</strong> what fundamental strategic forces <strong>and</strong> incentives drive firms’ investments in R&D <strong>and</strong><br />
what factors may lead to different types of dynamic competiti<strong>on</strong> (i.e. persistence of m<strong>on</strong>opoly or<br />
acti<strong>on</strong>-reacti<strong>on</strong>).<br />
Endogenous (or new) growth models, pi<strong>on</strong>eered by Romer, Lucas, Aghi<strong>on</strong> <strong>and</strong> Howitt have tried<br />
to explain the ec<strong>on</strong>omic forces driving the technological progress underlying ec<strong>on</strong>omic growth. In<br />
these models ec<strong>on</strong>omic growth, in general, derives from the growth of the stock of knowledge<br />
available to society, which in turn is the outcome of ec<strong>on</strong>omic choices of profit-maximizing agents.<br />
Earlier models developed in the 1960s were instead mainly focused <strong>on</strong> capital accumulati<strong>on</strong> <strong>and</strong><br />
did not treat technological change as an ec<strong>on</strong>omic phenomen<strong>on</strong> amenable to ec<strong>on</strong>omic analysis.<br />
Although these models are based <strong>on</strong> the equilibrium methodology typical of the neo-classical<br />
approach they are inherently dynamic. The evoluti<strong>on</strong> of the stock of knowledge is explained as<br />
either being the product of unintenti<strong>on</strong>al learning-by-doing activities or the outcome of R&D<br />
investments undertaken by profit-maximizing agents <strong>and</strong> rewarded by the m<strong>on</strong>opolistic rents that<br />
the innovator is able to appropriate. Externalities involved in the producti<strong>on</strong> of new knowledge<br />
10<br />
228
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
usually play an important role in these models, <strong>and</strong> in particular knowledge spillovers by which<br />
the knowledge generated by an agent enhances the technological capabilities of other agents or<br />
their ability to discover new knowledge.<br />
This secti<strong>on</strong> c<strong>on</strong>siders first game-theoretical microec<strong>on</strong>omic models of innovati<strong>on</strong> developed in<br />
the industrial organizati<strong>on</strong> literature. It then c<strong>on</strong>siders endogenous growth theories <strong>and</strong> the new<br />
ec<strong>on</strong>omic geography theories.<br />
1.2 Industrial Organizati<strong>on</strong><br />
The game-theoretical literature <strong>on</strong> innovati<strong>on</strong>, which was developed mainly in the 1980s <strong>and</strong><br />
1990s, addresses the strategic interacti<strong>on</strong> that characterizes R&D investment decisi<strong>on</strong>s in<br />
oligopolistic industries. These models provide the fundamental building blocks of more recent<br />
endogenous growth models, which cannot be properly assessed without reference to this earlier<br />
literature.<br />
More recently, game-theoretical models have focused <strong>on</strong> the analysis of competiti<strong>on</strong> <strong>and</strong><br />
innovati<strong>on</strong> in markets characterized by network effects, <strong>on</strong> the analysis of co-operati<strong>on</strong> in<br />
research <strong>and</strong> development activities <strong>and</strong> <strong>on</strong> the study of the relati<strong>on</strong>ship between intensity of<br />
competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong>.<br />
The starting point of this secti<strong>on</strong> is a categorizati<strong>on</strong> of game-theoretical models that should act as<br />
a guiding map to this literature. Underst<strong>and</strong>ing the nature of the models, <strong>and</strong> not <strong>on</strong>ly their<br />
results, is unavoidable to assess the adequacy of a particular model as theoretical reference in a<br />
practical case. The results of game theoretical models are generally very sensitive to the specific<br />
assumpti<strong>on</strong>s adopted <strong>and</strong> <strong>on</strong>ly by having clear the link between the former <strong>and</strong> the latter can<br />
these studies offer useful guidance for practical analyses of competiti<strong>on</strong> <strong>and</strong> market dynamics.<br />
The following secti<strong>on</strong>s focus <strong>on</strong> some of the issues that are addressed in the game-theoretical<br />
literature <strong>and</strong> that are most relevant to the analysis of competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> for antitrust<br />
purposes:<br />
•<br />
•<br />
•<br />
•<br />
the relati<strong>on</strong>ship between market structure <strong>and</strong> innovati<strong>on</strong>;<br />
the issue of persistence of m<strong>on</strong>opoly in dynamic markets;<br />
the relati<strong>on</strong>ship between toughness of product market competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong>; <strong>and</strong><br />
R&D co-operati<strong>on</strong>.<br />
An overview of the game-theoretical literature <strong>on</strong> R&D<br />
Game-theoretical models of innovati<strong>on</strong> explain the strategic interacti<strong>on</strong> that underlies firms’<br />
investments in research <strong>and</strong> development activities. The focus <strong>on</strong> strategic behavior<br />
distinguishes this literature from earlier decisi<strong>on</strong> theoretic models, e.g. Kamien <strong>and</strong> Schwartz<br />
(1980).<br />
11<br />
229
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Game-theoretical models of innovati<strong>on</strong> can be categorized as bel<strong>on</strong>ging to <strong>on</strong>e of three broad<br />
paradigms:<br />
• N<strong>on</strong>-tournament models. These models capture the strategic interacti<strong>on</strong> that can arise in<br />
those cases where firms pursue different paths of technological advance so that a<br />
discovery made al<strong>on</strong>g <strong>on</strong>e path can be used irrespectively <strong>on</strong> whether a competitor<br />
innovates al<strong>on</strong>g another technological path. These models have been investigated by<br />
Dasgupta <strong>and</strong> Stiglitz (1980a), Spence (1984) <strong>and</strong> Judd (1985).<br />
•<br />
•<br />
Deterministic aucti<strong>on</strong> models. Aucti<strong>on</strong> models of innovati<strong>on</strong> describe R&D competiti<strong>on</strong> in<br />
terms of rival firms bidding for a given process or product innovati<strong>on</strong>. Important<br />
c<strong>on</strong>tributi<strong>on</strong>s include Gilbert <strong>and</strong> Newbery (1982), Katz <strong>and</strong> Shapiro (1987) <strong>and</strong> Vickers<br />
(1986).<br />
Stochastic tournament models (or “patent races”). These models capture the strategic<br />
interacti<strong>on</strong> that arises when firms compete to be the first to introduce an innovati<strong>on</strong> <strong>and</strong><br />
the issue of timing is crucial to firms’ strategic interacti<strong>on</strong>. Papers in this category include<br />
Loury (1979), Lee <strong>and</strong> Wilde (1980), Dasgupta <strong>and</strong> Stiglitz (1980b), Grossman <strong>and</strong><br />
Shapiro (1987), Harris <strong>and</strong> Vickers (1987), Reinganum (1981,1982).<br />
Each of these approaches describes R&D investments in a particular setting, which may be an<br />
appropriate depicti<strong>on</strong> of reality in some circumstances <strong>and</strong> a less appropriate theoretical<br />
reference in relati<strong>on</strong> to some other industries or to particular stages of a technology life cycle.<br />
Despite the differences, however, game-theoretical models of innovati<strong>on</strong> illustrate that there are<br />
two central ec<strong>on</strong>omic incentives at the basis of firms’ investments in R&D activities, whose<br />
relative importance varies according to the specific envir<strong>on</strong>ment c<strong>on</strong>sidered: the profit incentive<br />
<strong>and</strong> the competitive threat (or replacement effect).<br />
The profit incentive, or st<strong>and</strong>-al<strong>on</strong>e incentive, is related to the noti<strong>on</strong> that allocating resources<br />
to innovative research <strong>and</strong> development would, if successful, increase a firm’s profits. This<br />
incentive can be usually thought of as the incentive to invest in R&D of a firm that takes the<br />
innovati<strong>on</strong> decisi<strong>on</strong> in isolati<strong>on</strong>. In fact, it is equal to the difference between the profits that the<br />
firm would get if it innovates <strong>and</strong> those that it would get if it does not innovate, all else equal.<br />
The competitive threat, or pre-empti<strong>on</strong> incentive, arises when strategic interacti<strong>on</strong> in innovati<strong>on</strong><br />
activities is c<strong>on</strong>sidered: a firm not <strong>on</strong>ly has to take into account the benefits c<strong>on</strong>nected to<br />
innovati<strong>on</strong> but also the possible loss of competitiveness where it does not innovate <strong>and</strong> a<br />
competitor does. This incentive reflects the threat posed by the existence of an active rival <strong>and</strong><br />
can be thought of as the difference between the profits if the firm innovates <strong>and</strong> the profits if it is<br />
left to a rival firm to innovate.<br />
12<br />
230
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
In general settings, both the profit incentive <strong>and</strong> the competitive threat may play a role in shaping<br />
the nature of dynamic competiti<strong>on</strong>. This would depend crucially <strong>on</strong> the specific characteristics of<br />
the market envir<strong>on</strong>ment c<strong>on</strong>sidered <strong>and</strong>, in the literature, <strong>on</strong> the specific assumpti<strong>on</strong>s adopted in<br />
each model.<br />
N<strong>on</strong>-tournament models<br />
N<strong>on</strong>-tournament models of innovati<strong>on</strong> investigate R&D investments in a setting where there can<br />
be multiple discoverers <strong>and</strong> firms do not compete to be the first to innovate. In these models<br />
symmetric competitors cannot prevent each other from getting equivalent improvements from<br />
spending equivalent amounts of resources in R&D.<br />
For this reas<strong>on</strong>, such models may explain R&D investments in those envir<strong>on</strong>ments where firms<br />
pursue different research paths so that an advance made <strong>on</strong> a single path can be appropriated<br />
irrespective of rivals’ advances <strong>on</strong> other research trajectories.<br />
A typical n<strong>on</strong>-tournament model, such as the <strong>on</strong>e by Dasgupta <strong>and</strong> Stiglitz (1980a), is structured<br />
in 3 stages: an entry stage, an R&D stage <strong>and</strong> a market competiti<strong>on</strong> stage. Firms first decide<br />
whether or not to enter an industry; active firms would then invest resources in R&D activities to<br />
decrease their marginal cost of producti<strong>on</strong> or increase the quality of their products <strong>and</strong> would<br />
finally compete in the marketplace.<br />
R&D activities are described in terms of a deterministic producti<strong>on</strong> functi<strong>on</strong> that specifies the size<br />
of innovati<strong>on</strong> according to the amount of cost reducti<strong>on</strong> or quality improvement per unit of R&D<br />
investment. Each firm can reduce its costs or improve the qualities of its products independently<br />
of parallel innovati<strong>on</strong> by rival firms, although it is usually assumed that a firm can indirectly benefit<br />
from a rival’s R&D activity for the existence of knowledge spillovers.<br />
Knowledge spillovers capture the noti<strong>on</strong> that innovative knowledge produced by R&D activities<br />
may, to some extent, be n<strong>on</strong>-excludable so that firms can, at least partially, appropriate the results<br />
of R&D activities of a rival. Cohen <strong>and</strong> Levinthal (1989) suggest that such learning may not be<br />
costless <strong>and</strong> that firms may have to invest in own R&D programs in order to access external<br />
knowledge <strong>and</strong> benefit from spillovers from innovating rivals.<br />
N<strong>on</strong>-tournament models have especially been employed to assess the extent of innovati<strong>on</strong><br />
achieved by the operati<strong>on</strong> of market forces <strong>and</strong> how it compares to the socially optimal <strong>on</strong>e, to<br />
discuss technological policy, <strong>and</strong> to investigate the relati<strong>on</strong>ship between market structure (i.e.<br />
c<strong>on</strong>centrati<strong>on</strong>) <strong>and</strong> innovati<strong>on</strong>.<br />
However, n<strong>on</strong>-tournament models assume that there are a large number of research paths <strong>and</strong><br />
the relati<strong>on</strong>ship between R&D investments <strong>and</strong> rewards from innovati<strong>on</strong> is essentially c<strong>on</strong>tinuous.<br />
For this reas<strong>on</strong> they are not well suited to describe market envir<strong>on</strong>ments where innovati<strong>on</strong> takes<br />
the form of a race <strong>and</strong> competiti<strong>on</strong> is for the market rather than in the market.<br />
Although this may be an appropriate descripti<strong>on</strong> for technological change in some markets, in<br />
many others innovati<strong>on</strong> is a more disruptive process that leads to winners <strong>and</strong> losers in very<br />
different competitive positi<strong>on</strong>s. In these circumstances, the rewards to R&D investments are<br />
13<br />
231
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
disc<strong>on</strong>tinuous, performance is rewarded <strong>on</strong> the basis of the rank within the set of realized<br />
performances, <strong>and</strong> competiti<strong>on</strong> takes the form of rival firms racing to be first to innovate<br />
(Dasgupta, 1986). An example of this type of race could be that characterizing competiti<strong>on</strong><br />
between Intel <strong>and</strong> AMD to introduce better PC processors.<br />
Aucti<strong>on</strong> models<br />
Aucti<strong>on</strong> models of innovati<strong>on</strong> describe R&D competiti<strong>on</strong> in terms of rival firms bidding for a given<br />
process or product innovati<strong>on</strong>. The innovati<strong>on</strong> is allocated to the firm which makes the highest bid<br />
<strong>and</strong> the issue is to investigate the identity of the firm which would have the larger willingness to<br />
pay for the innovati<strong>on</strong>.<br />
In general, the maximum bid that a firm would make is equal to the difference between the profits<br />
it would get if it were successful in the aucti<strong>on</strong> <strong>and</strong> those it would get were it not. For this reas<strong>on</strong>,<br />
aucti<strong>on</strong> models of innovati<strong>on</strong> emphasize the competitive threat as the incentive underlying<br />
competiti<strong>on</strong> in innovati<strong>on</strong>.<br />
Aucti<strong>on</strong> models have been widely used to investigate dynamic competiti<strong>on</strong> <strong>and</strong> whether market<br />
dynamics are characterized by persistent dominance by a technological leader or by “creative<br />
destructi<strong>on</strong>” whereby the incumbent is overtaken by some rival whose incumbency is itself shortlived.<br />
In a single aucti<strong>on</strong> model, the issue of dynamic competiti<strong>on</strong> is investigated by c<strong>on</strong>sidering a set of<br />
firms with different technological levels, i.e. marginal costs in the case of process innovati<strong>on</strong>, <strong>and</strong><br />
studying their incentives to purchase an innovati<strong>on</strong>. Market persistence is the result of such<br />
competiti<strong>on</strong> if the winner of the aucti<strong>on</strong> is the technological leader, otherwise dynamic competiti<strong>on</strong><br />
is characterized by acti<strong>on</strong>-reacti<strong>on</strong>.<br />
Assessing market dynamics in a single-innovati<strong>on</strong> aucti<strong>on</strong> model, however, may not help to<br />
underst<strong>and</strong> dynamic competiti<strong>on</strong> because in these models firms do not take into account how the<br />
outcome of the current bid may affect the outcome of future aucti<strong>on</strong>s. These inter-temporal links,<br />
however, may be an important element to explain firms’ incentives to invest in R&D in some<br />
c<strong>on</strong>texts where a sequence of innovative opportunities is in prospect. For this reas<strong>on</strong>, the<br />
questi<strong>on</strong> of dynamic competiti<strong>on</strong> may be better posed in the c<strong>on</strong>text of sequences of innovati<strong>on</strong><br />
rather than of a single innovati<strong>on</strong>.<br />
Aucti<strong>on</strong> models are intrinsically deterministic <strong>and</strong> this leads to the somewhat disappointing result<br />
that losers do not commit resources in R&D <strong>and</strong> that the process of technological advance is not<br />
really explained.<br />
14<br />
232
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Stochastic tournament models<br />
Tournament models of innovati<strong>on</strong> capture the particular interacti<strong>on</strong> that arises in those markets<br />
where dynamic competiti<strong>on</strong> takes the form of a race to be the first to make a discovery. As<br />
opposed to deterministic aucti<strong>on</strong> models, tournament models c<strong>on</strong>sider the case where there is<br />
technological uncertainty. This can be captured, for instance, by c<strong>on</strong>sidering a stochastic<br />
relati<strong>on</strong>ship between R&D effort <strong>and</strong> success.<br />
As such, these models emphasize timing in the c<strong>on</strong>text of R&D competiti<strong>on</strong> <strong>and</strong> are well suited to<br />
describe those envir<strong>on</strong>ments where competiti<strong>on</strong> is for the market rather than in the market. This<br />
nature of market dynamics can arise because technological change proceeds al<strong>on</strong>g a small<br />
number of research trajectories <strong>and</strong> there is competiti<strong>on</strong> to be the first to innovate because<br />
success by a firm would somehow prevent appropriability by subsequent innovators.<br />
The advantage of being the first to innovate is usually explained <strong>on</strong> the basis of the ability of the<br />
winner to appropriate the value of the innovati<strong>on</strong> by means of patent protecti<strong>on</strong>. More generally,<br />
however, appropriability may depend <strong>on</strong> factors other than patent protecti<strong>on</strong>, such as lead-time,<br />
access to complementary resources <strong>and</strong> secrecy. The relative importance of various<br />
appropriability factors is likely to be different across industries. (Levin et al., 1987)<br />
In aucti<strong>on</strong> models it is the competitive threat that characterizes firms’ incentives to invest in R&D<br />
activities. When r<strong>and</strong>omness is introduced, however, both the profit incentive <strong>and</strong> the competitive<br />
threat are relevant in firms’ strategic interacti<strong>on</strong>.<br />
In general, these forces need not be symmetric <strong>and</strong> have to be c<strong>on</strong>sidered for each firm involved:<br />
<strong>on</strong>e firm may face the greater competitive threat <strong>and</strong> another may have the greater profit<br />
incentive. Clearly, if a firm faces both the greater profit incentive <strong>and</strong> competitive threat it would<br />
undertake more R&D <strong>and</strong> hence be more likely to win the race. However, where there are<br />
asymmetric incentives, the nature of the outcome is not as straightforward.<br />
The literature usually focuses <strong>on</strong> the case where the competitive threat exceeds the profit<br />
incentive for all firms involved in the race. However, it could be the case that the profit incentive<br />
for both firms exceeds the competitive threat <strong>and</strong> that competiti<strong>on</strong> in R&D takes the form of a<br />
waiting game where firms tend to behave as free-riders <strong>and</strong> invest little resources in R&D. This<br />
type of situati<strong>on</strong> has been c<strong>on</strong>sidered, for instance, by Katz <strong>and</strong> Shapiro (1987) who explicitly<br />
account for the possibility of imitati<strong>on</strong> <strong>and</strong> licensing of innovati<strong>on</strong>.<br />
An important distincti<strong>on</strong> am<strong>on</strong>g tournament models is related to the assumpti<strong>on</strong> <strong>on</strong> the nature of<br />
R&D expenditures <strong>and</strong> whether these are c<strong>on</strong>tractual (e.g. Dasgupta <strong>and</strong> Stiglitz, 1980, Loury,<br />
1979) or n<strong>on</strong>-c<strong>on</strong>tractual (e.g. Lee <strong>and</strong> Wilde, 1980; <strong>and</strong> Reinganum, 1981,1982). In the first<br />
case R&D outlays take the form of a lump-sum cost incurred at the outset; in the sec<strong>on</strong>d case<br />
R&D can be c<strong>on</strong>sidered as a flow — a cost that it is incurred until the successful innovati<strong>on</strong> is<br />
achieved.<br />
Beath, Katsoulacos <strong>and</strong> Ulph (1994) discuss the implicati<strong>on</strong>s of these different assumpti<strong>on</strong>s.<br />
When R&D is of the c<strong>on</strong>tractual kind, equilibrium investments can be explained <strong>on</strong> a relatively<br />
15<br />
233
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
simple marginal benefit/marginal cost c<strong>on</strong>diti<strong>on</strong>. On the other h<strong>and</strong>, when R&D is of the n<strong>on</strong>c<strong>on</strong>tractual<br />
kind, it is the competitive threat that is the main determinant of a firm’s R&D.<br />
As with aucti<strong>on</strong> models, tournament models have also been developed to c<strong>on</strong>sider a sequence of<br />
tournaments in which firms complete a number of R&D stages before competing in the market.<br />
Examples of models that fall in this category are Fudenberg et al. (1983), Harris <strong>and</strong> Vickers<br />
(1987) <strong>and</strong> Grossman <strong>and</strong> Shapiro (1987).<br />
In these models innovati<strong>on</strong> is obtained <strong>on</strong>ly as the result of the completi<strong>on</strong> of a number of stages.<br />
A crucial issue that arises in such a setting is the nature of the innovative process <strong>and</strong> in particular<br />
whether each firm has to discover each technological step itself before it can move <strong>on</strong> to discover<br />
the next or whether instead the follower firm can compete directly with the leader for the new<br />
state-of-the-art technology.<br />
The latter case may arise if knowledge spillovers are relevant so that a laggard firm has access to<br />
(but may not have right to use) the same technological knowledge of the technological leader. On<br />
the other h<strong>and</strong>, knowledge may be excludable to some extent or in-house R&D is very important<br />
to assimilate external knowledge so that the technological fr<strong>on</strong>tier has to be achieved<br />
independently by each firm.<br />
The Grossman <strong>and</strong> Shapiro, Park <strong>and</strong> Harris <strong>and</strong> Vickers step-by-step models assume that each<br />
firm has to discover each technological step itself before it can move <strong>on</strong> to discover the next.<br />
Reinganum (1985) <strong>and</strong> Beath et al. (1987) are examples of leapfrog models where a follower is<br />
allowed at any time to compete directly with the leader for the new best-practice technology.<br />
In a step-by-step model firms can compete neck-to-neck or have different technological levels. In<br />
a leapfrog model firms are always asymmetric in terms of their technological ability. This<br />
distincti<strong>on</strong> turns out to be crucial in the analysis of the relati<strong>on</strong>ship between competiti<strong>on</strong> <strong>and</strong><br />
innovati<strong>on</strong>, as recent c<strong>on</strong>tributi<strong>on</strong>s in the endogenous growth literature surveyed below have<br />
shown.<br />
Endogenous growth theories have further developed these n<strong>on</strong>-tournament models by<br />
c<strong>on</strong>sidering sequences of tournaments where firms compete to innovate <strong>and</strong> compete in the<br />
product market before starting the following race to innovate.<br />
Market structure <strong>and</strong> innovati<strong>on</strong><br />
A l<strong>on</strong>g-st<strong>and</strong>ing questi<strong>on</strong> in the ec<strong>on</strong>omics of technological change has been the nature of the<br />
relati<strong>on</strong>ship between market structure <strong>and</strong> innovati<strong>on</strong>.<br />
The relati<strong>on</strong>ship between c<strong>on</strong>centrati<strong>on</strong> of an industry <strong>and</strong> its rate of technological innovati<strong>on</strong> is<br />
complex. Market structure may have an impact <strong>on</strong> the rate of innovati<strong>on</strong>, but innovati<strong>on</strong> is also<br />
an important factor that shapes market structure. In fact, it is necessary to recognize that the<br />
16<br />
234
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
relati<strong>on</strong>ship between c<strong>on</strong>centrati<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> is not a causal <strong>on</strong>e: both are the endogenous<br />
outcomes of the operati<strong>on</strong> of market forces <strong>and</strong> exogenous factors such as the nature of<br />
dem<strong>and</strong>, technological opportunity, the c<strong>on</strong>diti<strong>on</strong>s governing appropriability, <strong>and</strong> pure chance.<br />
The classical point of departure in the ec<strong>on</strong>omic analysis of the relati<strong>on</strong>ship between static market<br />
structure (i.e. c<strong>on</strong>centrati<strong>on</strong>) <strong>and</strong> innovati<strong>on</strong> is the work by Arrow (1962). Arrow c<strong>on</strong>siders the<br />
case where a cost-reducing innovati<strong>on</strong> is exogenously available <strong>and</strong> investigates firms’<br />
willingness to pay for the innovati<strong>on</strong> under different market structures. For a drastic innovati<strong>on</strong>,<br />
Arrow’s analysis shows that a firm that is already a m<strong>on</strong>opolist would have lower incentives to<br />
innovate than a firm that is currently in a perfectly competitive envir<strong>on</strong>ment, essentially because it<br />
would have the lower profit incentive.<br />
On the other h<strong>and</strong>, innovati<strong>on</strong> is certainly an important factor that affects market structure:<br />
innovati<strong>on</strong> is a means by which a firm tries to escape the c<strong>on</strong>straints imposed by competiti<strong>on</strong>.<br />
Studies in the Schumpeterian traditi<strong>on</strong> have emphasized the importance of ex-post market power<br />
for firms’ incentives to innovate. Some degree of market power is necessary for a firm to cover its<br />
R&D outlays: dynamic <strong>and</strong> static efficiency are somehow c<strong>on</strong>flicting. This is a theme that has<br />
been well developed in the recent literature <strong>on</strong> endogenous growth.<br />
Dasgupta <strong>and</strong> Stiglitz (1980a) discuss the relati<strong>on</strong>ship between c<strong>on</strong>centrati<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> in a<br />
n<strong>on</strong>-tournament model. If there are exogenous entry barriers, an increase in the number of firms<br />
causes each firm to spend less <strong>on</strong> R&D in equilibrium, however total R&D expenditure increases<br />
with the number of firms. When entry is c<strong>on</strong>sidered to be endogenous, <strong>on</strong>e would observe more<br />
innovati<strong>on</strong> in those industries that are characterized by a higher degree of m<strong>on</strong>opoly power,<br />
although no causality should be imputed to this relati<strong>on</strong>ship.<br />
Sutt<strong>on</strong> (1998) adopts an innovative approach to the study of technology <strong>and</strong> market structure <strong>and</strong><br />
achieves a very general, albeit somehow loose, result <strong>on</strong> the relati<strong>on</strong>ship between c<strong>on</strong>centrati<strong>on</strong><br />
<strong>and</strong> R&D. He is able to characterize, under very general assumpti<strong>on</strong>s, a lower bound to market<br />
c<strong>on</strong>centrati<strong>on</strong> (in large markets) that holds in those industries where the effectiveness of R&D in<br />
raising c<strong>on</strong>sumers’ willingness to pay is high <strong>and</strong> the product groups within the industry are<br />
sufficiently close substitutes. By c<strong>on</strong>trast, in those industries where R&D is not effective or where<br />
product groups are not close substitutes, very fragmented market structures can also be<br />
equilibrium c<strong>on</strong>figurati<strong>on</strong>s no matter how large the market size is.<br />
The models discussed above c<strong>on</strong>sider the relati<strong>on</strong>ship between product market structure (i.e.<br />
c<strong>on</strong>centrati<strong>on</strong>) <strong>and</strong> innovati<strong>on</strong>. In the c<strong>on</strong>text of models where firms race to innovate another<br />
interesting questi<strong>on</strong> c<strong>on</strong>cerns the relati<strong>on</strong>ship between the number of firms that are part of the<br />
race <strong>and</strong> the pace of technological advance.<br />
17<br />
235
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Loury (1979) c<strong>on</strong>siders a tournament model of innovati<strong>on</strong> where R&D expenditures are<br />
committed upfr<strong>on</strong>t (that is, the probability of success depends <strong>on</strong> the scale of the R&D lab) <strong>and</strong><br />
shows that increasing the number of firms reduces the expected date of inventi<strong>on</strong>. Lee <strong>and</strong> Wilde<br />
(1980) c<strong>on</strong>sider the case where the probability of innovati<strong>on</strong> is related to the research intensity<br />
<strong>and</strong> show that as the number of firms in the industry increases, the equilibrium rate of investment<br />
per firm increases <strong>and</strong> success by any <strong>on</strong>e firm is hastened by an increase in the number of<br />
firms. However, the first success date is hastened by an increase in the number of firms in both<br />
models.<br />
Reinganum (1982) c<strong>on</strong>siders a setting where firms can vary the research intensity as in Lee <strong>and</strong><br />
Wilde (1980) but does not assume that the rate of expenditure is c<strong>on</strong>stant over time. Instead,<br />
firms may adjust R&D intensity in resp<strong>on</strong>se to elapsed time <strong>and</strong> rival progress. Reinganum<br />
shows that, in this setting, when imitati<strong>on</strong> is not possible, an increase in the number of firms<br />
increases the equilibrium rate of investment for each firm <strong>and</strong> decreases the expected time of<br />
innovati<strong>on</strong>. When there is no full patent protecti<strong>on</strong>, the relati<strong>on</strong>ship is ambiguous <strong>and</strong> depends <strong>on</strong><br />
the relative payoffs to the innovator <strong>and</strong> the imitators.<br />
The persistence of dominance<br />
In many industries characterized by fundamental l<strong>on</strong>g-term market dynamics competiti<strong>on</strong> may<br />
take the form of competiti<strong>on</strong> for the market rather than competiti<strong>on</strong> in the market.<br />
In these markets the issue is not whether more or less c<strong>on</strong>centrati<strong>on</strong> is associated with faster<br />
technological progress but whether market dynamics would be characterised by persistent<br />
dominance of the incumbent leader or by acti<strong>on</strong>-reacti<strong>on</strong> whereby incumbents are overtaken by<br />
some rivals whose incumbency is itself short-lived.<br />
The dynamic evoluti<strong>on</strong> of market structure depends <strong>on</strong> both abilities <strong>and</strong> incentives of the<br />
incumbent <strong>and</strong> the rivals to innovate. Game theoretical models are well suited to analyse the<br />
incentives underlying R&D investments <strong>and</strong> the resulting evoluti<strong>on</strong> of market structure. If we<br />
focus <strong>on</strong> ec<strong>on</strong>omic incentives, <strong>and</strong> set aside differences in R&D abilities, persistence of m<strong>on</strong>opoly<br />
or acti<strong>on</strong>-reacti<strong>on</strong> can be related to the different profit incentive <strong>and</strong> competitive threat faced by<br />
the leader <strong>and</strong> the follower(s).<br />
Gilbert <strong>and</strong> Newbery (1982) use the aucti<strong>on</strong> model to examine potential competiti<strong>on</strong> at the R&D<br />
stage <strong>and</strong> reverse Arrow’s (1962) result: potential competiti<strong>on</strong> matters <strong>and</strong> a m<strong>on</strong>opolist may<br />
have more incentives to innovate than a potential entrant because the incumbent m<strong>on</strong>opolist may<br />
face a larger competitive threat.<br />
The profit of a successful incumbent who innovates is that of a m<strong>on</strong>opolist firm, whereas if it were<br />
the entrant to innovate each firm would get the profit of a duopolist. Hence, the competitive threat<br />
of the incumbent can be measured as the difference between the profit of a m<strong>on</strong>opolist firm <strong>and</strong><br />
the profit of a duopolist firm. The incumbent’s competitive threat, instead, is simply equal to the<br />
profit of a duopolist. This implies that the incumbent would have more incentives to innovate (i.e.<br />
the larger competitive threat) if, as it is normally the case, the profit of a m<strong>on</strong>opolist is greater than<br />
the combined profits of two duopolists.<br />
18<br />
236
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Katz <strong>and</strong> Shapiro (1987) study a general aucti<strong>on</strong> model <strong>and</strong> show that for minor innovati<strong>on</strong>s, the<br />
industry leader will typically be the innovator, whether or not imitati<strong>on</strong> <strong>and</strong> licensing are feasible.<br />
For markets where patent protecti<strong>on</strong> is str<strong>on</strong>g, they predict that the major innovati<strong>on</strong>s will be<br />
made by industry leaders. But if imitati<strong>on</strong> is easy, industry followers or entrants will make the<br />
major discoveries.<br />
Reinganum (1983) in a tournament model shows that when the first innovator captures a<br />
sufficiently high share of the post innovati<strong>on</strong> market, then the incumbent firm invests less <strong>on</strong> a<br />
given project that does the potential entrant. This is because the incumbent has less incentive<br />
than the challenger to shorten the period of its incumbency.<br />
Beath, Katsoulacos <strong>and</strong> Ulph (1995) show that in a tournament model under Cournot<br />
competiti<strong>on</strong>, a large innovati<strong>on</strong> or a large initial gap results in persistent dominance while a small<br />
initial gap <strong>and</strong> a small innovati<strong>on</strong> results in acti<strong>on</strong>-reacti<strong>on</strong>. However, if market competiti<strong>on</strong> is<br />
Bertr<strong>and</strong> there is persistent dominance irrespective of the size of the innovati<strong>on</strong> or the initial gap.<br />
This is a result that points towards the possible trade-off between static <strong>and</strong> dynamic efficiency<br />
suggested also by Vickers (1986).<br />
Market dynamics have also been investigated in models that c<strong>on</strong>sider a sequence of innovati<strong>on</strong>s.<br />
Reinganum (1985) c<strong>on</strong>siders a sequence of drastic innovati<strong>on</strong>s <strong>and</strong> shows that market dynamics<br />
are characterised by a process that resembles Schumpeter’s process of creative destructi<strong>on</strong>: the<br />
incumbent invests less than each challenger in each stage.<br />
Vickers (1986) c<strong>on</strong>siders a sequence of n<strong>on</strong>-drastic process innovati<strong>on</strong> in the c<strong>on</strong>text of the<br />
aucti<strong>on</strong> model. He compares market dynamics under Bertr<strong>and</strong> <strong>and</strong> Cournot competiti<strong>on</strong> <strong>and</strong><br />
finds that when the product market is very competitive then there is increasing dominance, but<br />
when it is not very competitive there is acti<strong>on</strong>-reacti<strong>on</strong>. When innovati<strong>on</strong> is drastic, then market<br />
dynamics are characterised by increasing dominance.<br />
Reinganum (1989) observes that the differences in the results obtained in different models can be<br />
ascribed to the different roles that the profit incentive <strong>and</strong> the competitive threat play in stochastic<br />
tournament <strong>and</strong> deterministic aucti<strong>on</strong> models. In a deterministic model the incentive to pre-empt<br />
(larger for the incumbent) dominates the firms’ decisi<strong>on</strong>. When success is stochastic, however,<br />
the threat from the rival innovating is less acute. In the case of drastic innovati<strong>on</strong>s, the<br />
competitive threat is the same for both firms <strong>and</strong> it is the profit incentive (which is larger for<br />
challengers) that determines R&D investments. The relevance of the profit incentive extends to<br />
the case for some n<strong>on</strong>-drastic innovati<strong>on</strong>.<br />
Competiti<strong>on</strong> intensity <strong>and</strong> innovati<strong>on</strong><br />
Market structure is often associated with the c<strong>on</strong>cept of competitiveness: usually, a high level of<br />
c<strong>on</strong>centrati<strong>on</strong> in an industry is interpreted as weak competiti<strong>on</strong>. This view is based <strong>on</strong> a<br />
symmetric Cournot model, where price-cost margins are higher as the number of firms increases.<br />
However, it is preferable to disentangle the noti<strong>on</strong>s of market structure <strong>and</strong> toughness of price<br />
competiti<strong>on</strong> as, am<strong>on</strong>g others, Sutt<strong>on</strong> (1998) <strong>and</strong> Bo<strong>on</strong>e (2000, 2001) have d<strong>on</strong>e.<br />
19<br />
237
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
In a simple world with homogeneous firms, toughness of price competiti<strong>on</strong> can be c<strong>on</strong>sidered as<br />
being related to the level of price-cost margins given any level of c<strong>on</strong>centrati<strong>on</strong>. Hence, a<br />
differentiated Bertr<strong>and</strong> market would be c<strong>on</strong>sidered to be more competitive than a differentiated<br />
Cournot market because price-cost margins would be lower in the former, for any level of<br />
c<strong>on</strong>centrati<strong>on</strong>. As a result, more competitive markets may allow fewer firms to profitably survive.<br />
Recently, Bo<strong>on</strong>e (2000,2001) has offered an interesting c<strong>on</strong>tributi<strong>on</strong> to the study of the<br />
relati<strong>on</strong>ship between toughness of competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong>. Bo<strong>on</strong>e defines intensity of<br />
competiti<strong>on</strong> <strong>on</strong> the basis of an axiomatic approach, which allows the result to be general to a<br />
large class of specific game-theoretical models. Bo<strong>on</strong>e c<strong>on</strong>siders four axioms:<br />
•<br />
•<br />
•<br />
•<br />
the lower limit of competiti<strong>on</strong> is such that firms are not affected by opp<strong>on</strong>ents’ acti<strong>on</strong>s;<br />
the least efficient firm in the market loses as competiti<strong>on</strong> becomes more intense;<br />
if the leader is far enough ahead, he gains as competiti<strong>on</strong> becomes more intense; <strong>and</strong><br />
if all active firms have similar costs, they all lose as competiti<strong>on</strong> becomes more intense.<br />
Bo<strong>on</strong>e (2001) shows that this definiti<strong>on</strong> encompasses, for instance, a switch from Cournot to<br />
Bertr<strong>and</strong> competiti<strong>on</strong> <strong>and</strong> a reducti<strong>on</strong> in travel costs in typical horiz<strong>on</strong>tal product differentiati<strong>on</strong><br />
models.<br />
When the noti<strong>on</strong> of intensity of competiti<strong>on</strong> is defined in such a way, it is normally inversely related<br />
to equilibrium c<strong>on</strong>centrati<strong>on</strong>. For instance, in Bo<strong>on</strong>e’s setting, if intensity of competiti<strong>on</strong> is low,<br />
then a large number of firms can be active in the market because the less efficient firms can<br />
survive as well. On the other h<strong>and</strong>, fierce competiti<strong>on</strong> is associated with high c<strong>on</strong>centrati<strong>on</strong><br />
because <strong>on</strong>ly the most efficient firms can survive.<br />
On the basis of this definiti<strong>on</strong> of product market competiti<strong>on</strong>, Bo<strong>on</strong>e (2001) investigates the<br />
relati<strong>on</strong>ship between intensity of competiti<strong>on</strong> <strong>and</strong> firms’ incentives to invest in R&D. The central<br />
finding of this study is that the relati<strong>on</strong>ship between market competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> may not<br />
be m<strong>on</strong>ot<strong>on</strong>ic; this is so because varying intensity of competiti<strong>on</strong> is associated with different<br />
identity of the innovator.<br />
Bo<strong>on</strong>e (2001) explains this interesting result as follow. When intensity of competiti<strong>on</strong> is low it is<br />
the follower that would be the next innovator, whereas when intensity of competiti<strong>on</strong> is large, it is<br />
the current technological leader that is likely to innovate. When it is the follower that innovates,<br />
tougher competiti<strong>on</strong> implies lower profits <strong>and</strong> hence lower incentives to innovate. However, when<br />
it is the current leader to innovate, an increase of toughness of price competiti<strong>on</strong> would further<br />
increase the profits related to his technological leadership <strong>and</strong> hence would increase the value of<br />
innovati<strong>on</strong> for the firm. Hence, the relati<strong>on</strong>ship between toughness of price competiti<strong>on</strong> <strong>and</strong><br />
innovati<strong>on</strong> would be U-shaped.<br />
Denicolò (2002) observes that Bo<strong>on</strong>e’s paper generalises the findings of a number of other<br />
studies. Delb<strong>on</strong>o <strong>and</strong> Denicolò (1990) c<strong>on</strong>sidered a homogeneous product market <strong>and</strong> found<br />
that Bertr<strong>and</strong> duopolists have greater incentive to innovate than Cournot duopolists. Bester <strong>and</strong><br />
20<br />
238
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Petrakis (1993) <strong>and</strong> B<strong>on</strong>anno <strong>and</strong> Haworth (1998) showed respectively that horiz<strong>on</strong>tal <strong>and</strong><br />
vertical product differentiati<strong>on</strong> could reverse this result. Qiu (1997) compared the two regimes in<br />
a n<strong>on</strong>-tournament model <strong>and</strong> find that is such a framework the incentive to innovate is greater<br />
with Cournot competiti<strong>on</strong> even in the case of homogenous products.<br />
Other studies that investigate the relati<strong>on</strong>ship between toughness of competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong><br />
fall within the endogenous growth literature ,Aghi<strong>on</strong> <strong>and</strong> Griffith (2005) <strong>and</strong> are surveyed in the<br />
relevant secti<strong>on</strong> of this<br />
review.<br />
R&D Co-operati<strong>on</strong><br />
Increasingly, research joint ventures (RJVs) have become a widespread form of industrial cooperati<strong>on</strong>.<br />
In parallel, the study of the effects of R&D co-operati<strong>on</strong> has also emerged as an<br />
important research topic when c<strong>on</strong>sidering industry policy.<br />
Co-operative R&D ventures can take many forms, ranging from R&D joint ventures, to crosslicensing<br />
agreements <strong>and</strong> various informal types of technology trading or informati<strong>on</strong>-sharing<br />
agreements.<br />
Research Joint Ventures themselves can cover a variety of arrangements. One type of RJV is<br />
the traditi<strong>on</strong>al joint venture, in which two or more parties create a separate entity in which they all<br />
have equity interests to c<strong>on</strong>duct well-defined R&D projects for their benefits. Another type is the<br />
research c<strong>on</strong>sortium. A third form of RJV is the venture capital investment by firms in a st<strong>and</strong>al<strong>on</strong>e<br />
start-up company.<br />
Ec<strong>on</strong>omists have investigated the extent to which RJVs might allow firms to internalize spillovers,<br />
co-ordinate their research activities <strong>and</strong> achieve higher R&D efficiency. Co-operative R&D is<br />
thought to be socially beneficial for several reas<strong>on</strong>s:<br />
•<br />
•<br />
•<br />
RJVs can alleviate the under-provisi<strong>on</strong> of R&D effort that results from technological<br />
spillovers <strong>and</strong> other sub-optimal R&D decisi<strong>on</strong>s.<br />
Co-operati<strong>on</strong> can lead to greater disseminati<strong>on</strong> of R&D results. It can improve R&D<br />
efficiency through good research design <strong>and</strong> informati<strong>on</strong> sharing, avoiding needless<br />
duplicati<strong>on</strong> of resources.<br />
Co-operative R&D enables the firms to share risks, to exploit synergies, pool different<br />
complementary assets, <strong>and</strong> exploit increasing returns to scale in R&D. It can enable firms<br />
to overcome a cost-of-development barrier impenetrable to any <strong>on</strong>e of them al<strong>on</strong>e.<br />
Adding to these effects, an important issue that is not very developed in the industrial organizati<strong>on</strong><br />
literature but is emphasized in the evoluti<strong>on</strong>ary ec<strong>on</strong>omics <strong>and</strong> the management literature <strong>on</strong><br />
21<br />
239
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
inter-firm co-operati<strong>on</strong> is that cooperative ventures may be an important means by which firms<br />
exchange specialised know-how <strong>and</strong> tacit knowledge.<br />
Countering these social benefits is the fear that firms participating in co-operative R&D might use<br />
the venture to engage in anti-competitive practices or that they might free ride <strong>on</strong> each other.<br />
RJVs <strong>and</strong> R&D investments<br />
Since the early c<strong>on</strong>tributi<strong>on</strong>s to the literature <strong>on</strong> RJVs, the effects of R&D co-operati<strong>on</strong> <strong>on</strong> the<br />
amount of resources invested in innovative activities have been extensively investigated.<br />
This attenti<strong>on</strong> is immediately related to the study of the potential of RJVs to correct the suboptimal<br />
R&D decisi<strong>on</strong>s that may characterise independent decisi<strong>on</strong>-making.<br />
The results of these studies accord support to the presumpti<strong>on</strong> that RJVs, by internalising<br />
externalities which affect the efficiency of n<strong>on</strong>-cooperative decisi<strong>on</strong>-making, can help to correct<br />
the n<strong>on</strong>-optimality of independent investments in R&D. In explaining what impact RJVs may<br />
have <strong>on</strong> R&D investments, knowledge spillovers are widely acknowledged as a crucial factor, <strong>and</strong><br />
their c<strong>on</strong>siderati<strong>on</strong> plays an important role in most ec<strong>on</strong>omic models of RJVs.<br />
In their seminal paper, d’Asprem<strong>on</strong>t <strong>and</strong> Jacquemin (1988) pi<strong>on</strong>eered an influential approach to<br />
study the effects of co-operati<strong>on</strong> <strong>on</strong> R&D investments. They c<strong>on</strong>sider a two-stage model in which<br />
duopolists first c<strong>on</strong>duct partially inappropriable research leading to a reducti<strong>on</strong> in unit cost, <strong>and</strong><br />
then compete à la Cournot in the product market. The focus of their analysis is <strong>on</strong> the<br />
comparis<strong>on</strong> of the magnitude of cost-reducing technical advance achieved when firms c<strong>on</strong>duct<br />
R&D competitively versus co-operatively, in the presence of spillover effects.<br />
The difference between the independent decisi<strong>on</strong>-making equilibrium <strong>and</strong> the co-operative <strong>on</strong>e<br />
results from the balance of the spillover internalisati<strong>on</strong> effect <strong>and</strong> the reduced strategic incentive<br />
effect. Joint decisi<strong>on</strong> <strong>on</strong> the levels of R&D expenditures internalises two externalities disregarded<br />
under independent decisi<strong>on</strong>-making:<br />
• Research knowledge is partially inappropriable, <strong>and</strong> it can benefit other firms without the<br />
innovator being able to fully appropriate the value of the knowledge created. This<br />
externality causes sub-optimality of independent R&D investments. Its internalisati<strong>on</strong><br />
under joint-decisi<strong>on</strong> making (spillover internalisati<strong>on</strong> effect) will tend to increase cooperative<br />
R&D investment relative to the n<strong>on</strong>-cooperative level.<br />
•<br />
The innovative activity of a firm comprises also a negative pecuniary externality <strong>on</strong> the<br />
profits of the rival through a clear strategic market interacti<strong>on</strong>. Joint decisi<strong>on</strong>-making<br />
internalises this externality (reduced strategic incentive effect) <strong>and</strong> tends to decrease the<br />
co-operative level of R&D investment relative to the n<strong>on</strong>-cooperative equilibrium.<br />
The balance of these opposite forces is crucial to underst<strong>and</strong> the effects of co-operati<strong>on</strong> <strong>on</strong> the<br />
level of resources invested in research <strong>and</strong>, in turn, <strong>on</strong> the appropriateness of co-operati<strong>on</strong> to<br />
reduce the gap between the private <strong>and</strong> social incentives for doing R&D.<br />
22<br />
240
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
In case of “high” spillover rates, the spillover internalizati<strong>on</strong> effect would dominate the reduced<br />
strategic incentive effect, the net effect being an increase in the R&D expenditures. By c<strong>on</strong>trast, if<br />
spillovers are “low”, R&D co-operati<strong>on</strong> would cause R&D investments to decrease relative to the<br />
n<strong>on</strong>-cooperative equilibrium.<br />
The c<strong>on</strong>tributi<strong>on</strong>s of Kamien, Muller <strong>and</strong> Zang (1992), Suzumura (1992) Simps<strong>on</strong> <strong>and</strong> V<strong>on</strong>ortas<br />
(1994), have proposed extensi<strong>on</strong>s of the basic modelling framework to cover an arbitrary number<br />
of firm, general dem<strong>and</strong> <strong>and</strong> cost assumpti<strong>on</strong>s.<br />
RJVs <strong>and</strong> R&D efficiency<br />
As we have discussed, c<strong>on</strong>siderable attenti<strong>on</strong> has been devoted to the study of the effects of<br />
RJVs <strong>on</strong> the amount of resources allocated to R&D, <strong>and</strong> of the advantages of technological cooperati<strong>on</strong><br />
in terms of correcting inefficient R&D efforts that independent decisi<strong>on</strong>-making involves.<br />
Another comm<strong>on</strong>ly acknowledged advantage of R&D co-operati<strong>on</strong> is identified in the increased<br />
efficiency of R&D activities: RJVs not <strong>on</strong>ly can affect the absolute levels of R&D spending, but<br />
also the amount of R&D spending per unit of cost reducti<strong>on</strong>/quality improvement achieved.<br />
Indeed, to the extent that co-operative R&D is more widely disseminated than individually<br />
c<strong>on</strong>ducted R&D, ex ante co-operati<strong>on</strong> increases the efficiency of R&D efforts because a single<br />
investment benefits a greater number of firms.<br />
The efficiency gain can have three types of positive effects. First, sharing lowers the cost of<br />
investment for each firm, which may induce them to c<strong>on</strong>duct more R&D. Sec<strong>on</strong>d, for a given<br />
level of R&D investment it might increase the effective amount of R&D. Third, co-operative R&D<br />
can eliminate the wasteful duplicati<strong>on</strong> that would occur if several firms separately undertook the<br />
same projects. Even if several firms c<strong>on</strong>tinue to c<strong>on</strong>duct separate R&D under an ex ante<br />
agreement, they still can improve the efficiency of their efforts by co-ordinating them.<br />
Some studies have formalized these intuiti<strong>on</strong>s, by investigating the effects of co-operati<strong>on</strong> <strong>on</strong> the<br />
level of informati<strong>on</strong>-sharing <strong>and</strong> the co-ordinati<strong>on</strong> of research activities.<br />
Katz (1986) presented an early, albeit isolated, study in which firms choose the level of spillovers<br />
in the RJV, by determining the amount of informati<strong>on</strong>-sharing achieved. However, the<br />
endogeneity of spillovers is restricted to the co-operative equilibrium <strong>on</strong>ly, <strong>and</strong> this limits the<br />
significance of the comparis<strong>on</strong> of the (exogenous) n<strong>on</strong>-cooperative <strong>and</strong> (endogenous) cooperative<br />
informati<strong>on</strong>-sharing levels.<br />
To fully exploit the endogenizati<strong>on</strong> of knowledge spillovers, <strong>and</strong> explain the differences of<br />
informati<strong>on</strong>-sharing levels in the co-operative <strong>and</strong> n<strong>on</strong>-cooperative equilibrium it is necessary to<br />
23<br />
241
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
model both as the result of firms’ choices. Katsoulacos <strong>and</strong> Ulph (1998) have presented a<br />
stochastic model of innovati<strong>on</strong> with this feature.<br />
The authors adopt a structured approach to model knowledge spillovers by distinguishing two<br />
relevant dimensi<strong>on</strong>s: the adaptability of the research to the other firm <strong>and</strong> the amount of<br />
informati<strong>on</strong>-sharing that actually takes place.<br />
They show that the n<strong>on</strong>-cooperative equilibrium generally produces minimal spillovers; in this<br />
case, RJVs might lead to a higher informati<strong>on</strong>-sharing. However, if firms operate in different but<br />
complementary industries, it is possible to achieve maximum informati<strong>on</strong>-sharing also in the n<strong>on</strong>cooperative<br />
equilibrium.<br />
Poyago-Theotoky (1999) analyses endogenous spillovers in the c<strong>on</strong>text of a simple n<strong>on</strong>tournament<br />
model of R&D where firms are engaged in cost-reducing innovati<strong>on</strong>. It is shown that<br />
when spillovers of informati<strong>on</strong> are treated as endogenous firms never disclose any of their<br />
informati<strong>on</strong> when choosing their R&D n<strong>on</strong>-cooperatively. Under co-operative R&D, firms will<br />
always choose to fully share their informati<strong>on</strong>, i.e. a research joint venture will operate with a<br />
maximum spillover value.<br />
The relati<strong>on</strong>ship between the research paths followed by firms can be manifold, spanning a<br />
c<strong>on</strong>tinuum from perfect complementarity to perfect substitutability. Co-operati<strong>on</strong> in R&D can take<br />
advantage of joint research design, by choosing the number of labs to operate for optimally<br />
exploiting such substitutability or complementarity.<br />
The issue of organisati<strong>on</strong> design in an RJV has been further c<strong>on</strong>sidered by Katsoulacos <strong>and</strong> Ulph<br />
(1998) who relate this choice to the nature of research paths followed <strong>and</strong> the endogenous level<br />
of informati<strong>on</strong>-sharing chosen by firms. An important result of this study is that cost<br />
c<strong>on</strong>siderati<strong>on</strong>s, dominant when research outputs are close substitutes <strong>and</strong> firms are willing to<br />
share a lot of informati<strong>on</strong>, are <strong>on</strong>ly <strong>on</strong>e of the factors behind the RJVs decisi<strong>on</strong> as to whether to<br />
operate <strong>on</strong>e or two labs.<br />
Indeed, the RJV may close a lab for anti-competitive reas<strong>on</strong>s, to prevent having to face a very<br />
competitive situati<strong>on</strong> when both firms discover. Or, when there are very str<strong>on</strong>g complementarities<br />
between the research outputs of the two firms, the RJV might prefer to keep both labs open in<br />
order to exploit such complementarities.<br />
1.3 Endogenous Growth Models<br />
Technological change is a fundamental driver of ec<strong>on</strong>omic growth. However, <strong>on</strong>ly recently have<br />
the ec<strong>on</strong>omic forces that underlie technological progress in a dynamic ec<strong>on</strong>omy been<br />
investigated. Advances in this field have been associated with the development of endogenous<br />
growth theories, <strong>on</strong>e of the most fertile grounds for ec<strong>on</strong>omic research in the last 15 years. The<br />
relevant research agenda has been set <strong>on</strong> the study of ec<strong>on</strong>omic growth as the result of<br />
knowledge producti<strong>on</strong> — we shall focus <strong>on</strong> technological knowledge in this survey — which is<br />
explained as the outcome of ec<strong>on</strong>omic decisi<strong>on</strong>s.<br />
24<br />
242
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Before the development of endogenous growth theories ec<strong>on</strong>omic models of growth were<br />
essentially models of capital accumulati<strong>on</strong> developed <strong>on</strong> the basis of Solow’s (1956) seminal<br />
c<strong>on</strong>tributi<strong>on</strong>. These models suggested that the simple accumulati<strong>on</strong> of physical capital cannot<br />
sustain l<strong>on</strong>g-run ec<strong>on</strong>omic growth as l<strong>on</strong>g as its marginal productivity is decreasing <strong>and</strong> not<br />
bounded from zero. L<strong>on</strong>g run ec<strong>on</strong>omic growth had to be attributed to the effect of some other<br />
factor, exogenous to the ec<strong>on</strong>omic choices analyzed in these models, such as technological<br />
change.<br />
In fact, even in older exogenous growth theories that studied capital accumulati<strong>on</strong>, technological<br />
change was acknowledged as <strong>on</strong>e of the major factors that explain the ability of an ec<strong>on</strong>omy to<br />
grow in the l<strong>on</strong>g run. Certainly, advances in technology are a desirable element in any<br />
explanati<strong>on</strong> of growth: “a story of growth that neglects technological progress is both ahistorical<br />
<strong>and</strong> implausible”(Grossman <strong>and</strong> Helpman,1994,p.26). Cliometric studies have also examined the<br />
relati<strong>on</strong>ship between knowledge<br />
<strong>and</strong> ec<strong>on</strong>omic growth as shown in the historical background below.<br />
Nevertheless, for a l<strong>on</strong>g time the forces underlying advances were not explained <strong>on</strong> the basis of<br />
ec<strong>on</strong>omic incentives but were essentially related to the exogenous evoluti<strong>on</strong> of scientific<br />
knowledge <strong>and</strong> its technological applicati<strong>on</strong>s.<br />
Endogenous growth theories or new growth theory have certainly c<strong>on</strong>tributed to fill this gap.<br />
These theories are based <strong>on</strong> product differentiati<strong>on</strong> <strong>and</strong> imperfect competiti<strong>on</strong>, ec<strong>on</strong>omies of<br />
scale <strong>and</strong> increasing returns instead of perfect competiti<strong>on</strong> <strong>and</strong> diminishing returns put forward by<br />
neo-classical theory. Freeman (1994) argues that these theories have been c<strong>on</strong>siderably<br />
influenced by the neo-Schumpeterian school of innovati<strong>on</strong>.<br />
New growth theory not <strong>on</strong>ly emphasizes knowledge accumulati<strong>on</strong> <strong>and</strong> diffusi<strong>on</strong>, in the form<br />
mainly of human capital (Lucas, 1988, Rebelo, 1991) or technological innovati<strong>on</strong> (Romer, 1990;<br />
Aghi<strong>on</strong> <strong>and</strong> Howitt 1992), as fundamental drivers of ec<strong>on</strong>omic growth but also the ec<strong>on</strong>omic<br />
decisi<strong>on</strong>s that underlie both.<br />
The model developed by Rebelo (1991) represents the first generati<strong>on</strong> of endogenous growth<br />
models that underscore the importance of human capital accumulati<strong>on</strong>. In this model, there is no<br />
exogenous productivity growth given that the entirety of output growth is explained by human<br />
capital <strong>and</strong> capital stock accumulati<strong>on</strong>.<br />
As such, this review focuses <strong>on</strong> the sec<strong>on</strong>d generati<strong>on</strong> of endogenous growth models that<br />
c<strong>on</strong>sider technological innovati<strong>on</strong> as well as human capital accumulati<strong>on</strong>. In general, these<br />
models relate ec<strong>on</strong>omic growth to innovati<strong>on</strong>s that improve firms’ productive efficiency <strong>and</strong>/or<br />
lead to the producti<strong>on</strong> of new or better intermediate <strong>and</strong> c<strong>on</strong>sumer goods.<br />
8<br />
9<br />
25<br />
243
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Innovati<strong>on</strong>s result from new knowledge produced as the outcome of firms’ learning-by-doing<br />
<strong>and</strong>/or intenti<strong>on</strong>al R&D activities. In general, R&D models of endogenous growth have drawn<br />
heavily from the earlier game theoretical literature <strong>on</strong> innovati<strong>on</strong> <strong>and</strong> extended those analyses to<br />
more dynamic settings.As such, they have developed the analysis of l<strong>on</strong>g-st<strong>and</strong>ing questi<strong>on</strong>s<br />
investigating, for instance, the relati<strong>on</strong>ship between the intensity of competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong><br />
<strong>and</strong> the nature of market dynamics.<br />
Recent ec<strong>on</strong>omic analysis in this field has offered interesting new insights <strong>on</strong> the relati<strong>on</strong>ship<br />
between competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong>. Despite some suggestive c<strong>on</strong>clusi<strong>on</strong>s, however,<br />
endogenous growth models suffer from the same limitati<strong>on</strong>s as their c<strong>on</strong>stituent micro<br />
foundati<strong>on</strong>s: results often depend crucially from the assumpti<strong>on</strong>s adopted <strong>and</strong> the specific<br />
structure of the model studied. Unfortunately, in many cases some of these important<br />
assumpti<strong>on</strong>s relate to features of market envir<strong>on</strong>ments that are hard to observe <strong>and</strong> evaluate.<br />
Ec<strong>on</strong>omic theory provides no general results that can be c<strong>on</strong>sidered to hold in any market situati<strong>on</strong>,<br />
but rather specific c<strong>on</strong>siderati<strong>on</strong>s that suggest which features of a market envir<strong>on</strong>ment are worth<br />
c<strong>on</strong>sidering to assess the nature of competiti<strong>on</strong>, market dynamics ,their relati<strong>on</strong>ship <strong>and</strong> growth.<br />
(Aghi<strong>on</strong> <strong>and</strong> Griffith, 2005).<br />
Learning-by-doing<br />
Knowledge generated by learning-by-doing can be a significant engine of ec<strong>on</strong>omic growth <strong>and</strong><br />
models of endogenous growth with technological progress driven by learning-by-doing have been<br />
studied by Romer (1986), Young (1991,1993), Stokey (1988) <strong>and</strong> Aghi<strong>on</strong> <strong>and</strong> Howitt (1998).<br />
Learning-by-doing, since the seminal work of Arrow (1962), Solow (1997) has been c<strong>on</strong>sidered an<br />
important source of technological knowledge. Firms can produce new knowledge without investing<br />
in instituti<strong>on</strong>alized research <strong>and</strong> development activities but rather, as a by-product of their usual<br />
producti<strong>on</strong> activity. Indeed, experience has been shown to be an important source of<br />
technological improvements in many industries such as artificial fibres, semic<strong>on</strong>ductors <strong>and</strong><br />
memory chips.<br />
26<br />
244
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Generally, in these models learning-by-doing is the unintenti<strong>on</strong>al by-product of producti<strong>on</strong> or<br />
investment activities <strong>and</strong> it is usually c<strong>on</strong>sidered to result in technological knowledge that is <strong>on</strong>ly<br />
partially appropriable. The existence of knowledge spillovers is in fact generally acknowledged: a<br />
firm can, in part at least, benefit of the knowledge generated by the learning-by-doing of other<br />
firms.<br />
Young (1991, 1993) studies a setting where new technologies are discovered through R&D<br />
activities but are initially inferior to the older technologies they seek to replace. Experience,<br />
however, generates incremental improvements over time, which allow the new technology to<br />
supplant older technologies. These incremental improvements are <strong>on</strong>ly achievable to some<br />
extent, in the sense that there is a bound to the technological advance that can be generated by<br />
learning-by-doing. This hybrid model emphasises that inventive activity <strong>and</strong> producti<strong>on</strong><br />
experience are complementary forces in driving technological change.<br />
R&D models<br />
Endogenous growth models have also investigated ec<strong>on</strong>omic growth as driven by the<br />
accumulati<strong>on</strong> <strong>and</strong> diffusi<strong>on</strong> of knowledge generated by Research <strong>and</strong> <strong>Development</strong> activities,<br />
undertaken by profit-maximising firms.<br />
These models investigate the process of discovery of new or better products. Accordingly, a<br />
distincti<strong>on</strong> can be drawn between those models that treat innovati<strong>on</strong> in terms of the development<br />
of new varieties of products <strong>and</strong> those that c<strong>on</strong>sider sequential improvements of the quality of<br />
existing products.<br />
The major difference between these two types of models is that the latter capture an important<br />
characteristic of the innovative process, namely that new inventi<strong>on</strong>s entail an element of<br />
destructi<strong>on</strong> because they make old technologies or products obsolete. For this reas<strong>on</strong>, early<br />
endogenous growth models with sequential improvements in the quality of products are called<br />
Schumpeterian.<br />
Despite these differences however, the two types of models share the same fundamental<br />
structure. In both models, firms invest resources to acquire the exclusive right to manufacture a<br />
new product <strong>and</strong> it is usually assumed that R&D activities generate knowledge spillovers that<br />
benefit other research firms. The producti<strong>on</strong> of new knowledge <strong>and</strong> its diffusi<strong>on</strong> drive ec<strong>on</strong>omic<br />
growth in the l<strong>on</strong>g run.<br />
Variety of products<br />
Romer (1990) provides the seminal model where technological progress is captured in terms of<br />
an increasing variety of products. The increase of the variety of intermediate or final goods is<br />
reflected in higher productivity or utility <strong>and</strong> hence ec<strong>on</strong>omic growth.<br />
27<br />
245
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
New varieties of products are discovered as outcomes of R&D investments <strong>and</strong> remunerated by<br />
the rents that accrue to the successful innovator who can benefit from patent protecti<strong>on</strong> (or, more<br />
generally, of other appropriability factors such as lead time, secrecy, access to complementary<br />
resources).<br />
In these models, it is generally assumed that R&D activities lead both to new knowledge that can<br />
be appropriated by the innovator <strong>and</strong> new general knowledge that cannot be appropriated but<br />
instead c<strong>on</strong>tributes to a stock of publicly available knowledge that facilitates further discoveries.<br />
Not <strong>on</strong>ly is the producti<strong>on</strong> of new knowledge necessary for ec<strong>on</strong>omic growth, but also its<br />
diffusi<strong>on</strong>.<br />
A distinguishing feature of Romer’s (1990) growth model is that it studies horiz<strong>on</strong>tal product<br />
innovati<strong>on</strong>s that involve no obsolescence: new goods are never close substitutes for existing<br />
goods (i.e. each new product finds its own horiz<strong>on</strong>tal niche).<br />
Although this feature of the model may detract from its realism is some cases, it is not necessarily<br />
so. There are some industries where innovati<strong>on</strong> is fundamentally directed at the introducti<strong>on</strong> of<br />
new varieties of products rather than at improving products’ qualities. An example discussed in<br />
the literature is the flowmeter industry where technological advance takes the form of the<br />
development of new types of flowmeters. In this industry there is no escalati<strong>on</strong> of R&D <strong>on</strong> a<br />
single technological trajectory <strong>and</strong> related product group but proliferati<strong>on</strong> of product groups.<br />
Quality of products<br />
Aghi<strong>on</strong> <strong>and</strong> Howitt (1992) examine growth as driven by industrial innovati<strong>on</strong>s that improve the<br />
quality of products.<br />
Vertical innovati<strong>on</strong> models of this kind introduce, relative to models with exp<strong>and</strong>ing varieties of<br />
products, the c<strong>on</strong>siderati<strong>on</strong> of the process of creative destructi<strong>on</strong>: a new innovati<strong>on</strong> makes, to<br />
some extent, old products obsolete <strong>and</strong> negatively affects the ec<strong>on</strong>omic rents of previous innovators.<br />
For this reas<strong>on</strong>, models with qualitative improvements are also called Schumpeterian because they<br />
embody Schumpeter’s idea of creative destructi<strong>on</strong> by emphasizing the process by which new<br />
products (<strong>and</strong> innovators) displace old products (<strong>and</strong> innovators).<br />
Endogenous growth models with vertical innovati<strong>on</strong>s have very elaborated micro-foundati<strong>on</strong>s<br />
describing the nature of market dynamics, which are similar to <strong>and</strong> develop further earlier gametheoretical<br />
models of innovati<strong>on</strong>. As such these models have developed the study of dynamic<br />
competiti<strong>on</strong>.<br />
28<br />
246
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Aghi<strong>on</strong> <strong>and</strong> Howitt (1992) study a model of ec<strong>on</strong>omic growth based <strong>on</strong> creative destructi<strong>on</strong>.<br />
Innovati<strong>on</strong> c<strong>on</strong>sists of a higher quality intermediate good that can be used to produce output<br />
more efficiently than before. Research firms invest resources in R&D activities that stochastically<br />
determine the time of the new innovati<strong>on</strong>; the arrival rate of new innovati<strong>on</strong>s depends <strong>on</strong>ly up<strong>on</strong><br />
the current flow of input to research. Firms’ R&D investments are motivated by the prospect of<br />
m<strong>on</strong>opoly rents that can be captured when a successful innovati<strong>on</strong> is patented. However, unlike<br />
the horiz<strong>on</strong>tal innovati<strong>on</strong>s c<strong>on</strong>sidered in Romer (1990), these rents would last <strong>on</strong>ly until the next<br />
innovati<strong>on</strong> occurs, when the current innovati<strong>on</strong> would become obsolete.<br />
Aghi<strong>on</strong> <strong>and</strong> Howitt c<strong>on</strong>sider the cases of both drastic <strong>and</strong> n<strong>on</strong>-drastic innovati<strong>on</strong>. In the first case<br />
the new innovator is unc<strong>on</strong>strained by potential competiti<strong>on</strong> from the previous patent; in the latter<br />
case, such a c<strong>on</strong>straint is binding.<br />
In the case of drastic innovati<strong>on</strong>s market dynamics take the form of a c<strong>on</strong>tinuous process of<br />
acti<strong>on</strong> <strong>and</strong> reacti<strong>on</strong> whereby at each point in time the market is dominated by a m<strong>on</strong>opolist whose<br />
incumbency is short lived. In theory, a new innovati<strong>on</strong> could be introduced by either the current<br />
m<strong>on</strong>opolist or an outside research firm. The value of innovati<strong>on</strong> is the expected incremental<br />
present value of the flow of m<strong>on</strong>opoly profits generated by the innovati<strong>on</strong> over an interval whose<br />
length depends <strong>on</strong> the occurrence of the next innovati<strong>on</strong>. Because this incremental profit is lower<br />
for an incumbent m<strong>on</strong>opolist, <strong>on</strong>ly outside research firms would innovate <strong>and</strong> the m<strong>on</strong>opolist<br />
chooses to do no research.<br />
The reas<strong>on</strong> for this result lies in the fact that, given the assumpti<strong>on</strong> <strong>and</strong> the structure of the model,<br />
firms’ incentives are driven <strong>on</strong>ly by the profit incentive, which in the specific case is always larger<br />
for an external firm. This is the replacement effect emphasized by Arrow (1962) whereby a<br />
m<strong>on</strong>opolist would have lower incremental profits from innovati<strong>on</strong> than a perfectly competitive firm.<br />
In the case of drastic innovati<strong>on</strong> the competitive threat does not play a role because the flow of<br />
profit is independent of the identity of the innovator.<br />
Innovati<strong>on</strong>s are n<strong>on</strong>-drastic if the previous incumbent could make a positive profit when the<br />
current <strong>on</strong>e is charging the unc<strong>on</strong>strained optimal m<strong>on</strong>opolistic price. In the case of n<strong>on</strong>-drastic<br />
innovati<strong>on</strong> the competitive threat may potentially play a role in shaping equilibrium investments in<br />
R&D activities. However, Aghi<strong>on</strong> <strong>and</strong> Howitt assume that the m<strong>on</strong>opolist chooses to do no<br />
research also in the case of n<strong>on</strong>-drastic innovati<strong>on</strong>s, i.e. assume that the efficiency effect is<br />
smaller than the replacement effect. As a result market dynamics take the form of c<strong>on</strong>tinuous<br />
acti<strong>on</strong>/reacti<strong>on</strong>.<br />
Finally, Aghi<strong>on</strong> <strong>and</strong> Howitt c<strong>on</strong>sider the case in which firms can affect not <strong>on</strong>ly the frequency but<br />
also the size of innovati<strong>on</strong>s. Their finding is that innovati<strong>on</strong>s would be too small if they were<br />
drastic; in the n<strong>on</strong>-drastic case, the tendency to make innovati<strong>on</strong>s too small is in part mitigated by<br />
the incentive for innovators to move away from their competitive fringe.<br />
Grossman <strong>and</strong> Helpman (1991) have built <strong>on</strong> the model by Aghi<strong>on</strong> <strong>and</strong> Howitt to c<strong>on</strong>sider an<br />
ec<strong>on</strong>omy where a c<strong>on</strong>tinuum of, <strong>and</strong> not <strong>on</strong>ly <strong>on</strong>e, intermediate goods. Each product has its own<br />
quality ladder <strong>and</strong> entrepreneurs target individual products <strong>and</strong> race to bring about the next<br />
generati<strong>on</strong>. In each industry, success occurs with a probability per unit of time that is proporti<strong>on</strong>al<br />
29<br />
247
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
to the total R&D resources invested to improving that product. The authors allow free entry in the<br />
race for the next generati<strong>on</strong> of products <strong>and</strong> assume that potential entrants can learn enough<br />
about the state of knowledge to compete to produce the new state-of-the-art quality. These<br />
assumpti<strong>on</strong>s imply that, without a cost advantage, industry leaders do not invest resources to<br />
improve the quality of their own state-of-the-art products. This is because of the replacement<br />
effect whereby the leader would obtain an incremental flow of profits that is less than that of an<br />
external firm.<br />
Aghi<strong>on</strong> <strong>and</strong> Howitt (1998) introduce an element of heterogeneity in innovative activity that<br />
captures the distincti<strong>on</strong> between fundamental <strong>and</strong> sec<strong>on</strong>dary research. They first investigate this<br />
heterogeneity by c<strong>on</strong>sidering fundamental research as deriving from R&D activities <strong>and</strong><br />
sec<strong>on</strong>dary innovati<strong>on</strong> from learning-by-doing. Each innovati<strong>on</strong> resulting from research c<strong>on</strong>sists of<br />
a potential new product, <strong>and</strong> learning-by-doing leads to improvements of the quality of goods that<br />
have been invented. They also c<strong>on</strong>sider the possibility that learning-by-doing c<strong>on</strong>tributed to a<br />
general stock of knowledge that creates new opportunities for research as well. Moreover,<br />
Aghi<strong>on</strong> <strong>and</strong> Howitt (1998) c<strong>on</strong>sider both the case where learning is not appropriated by the firm<br />
but is shared by all firms <strong>and</strong> where the quality enhancement of learning is fully internalized.<br />
Finally, they c<strong>on</strong>sider the case where the distincti<strong>on</strong> between fundamental <strong>and</strong> sec<strong>on</strong>dary<br />
innovati<strong>on</strong>s is captured in terms of research versus development.<br />
A feature of Aghi<strong>on</strong> <strong>and</strong> Howitt’s (1992) influential model, shared by most early models of<br />
endogenous growth with quality ladders <strong>and</strong> leapfrogging, is that more competiti<strong>on</strong> (i.e. higher<br />
elasticity of dem<strong>and</strong> for each firm) is associated with less innovati<strong>on</strong> <strong>and</strong> growth, a very<br />
Schumpeterian result. The reas<strong>on</strong> for this c<strong>on</strong>clusi<strong>on</strong> is that these models capture <strong>on</strong>ly an<br />
increase of ex post product market competiti<strong>on</strong> <strong>and</strong> not of ex ante competiti<strong>on</strong> as well. Some<br />
recent studies modify this feature of earlier models <strong>and</strong> reach a different c<strong>on</strong>clusi<strong>on</strong> <strong>on</strong> the<br />
relati<strong>on</strong>ship between competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong>: tougher competiti<strong>on</strong> may increase the<br />
incentives of firms to innovate in order to escape from a more competitive market state.<br />
In the model developed by Aghi<strong>on</strong>, Harris <strong>and</strong> Vickers (1997) <strong>and</strong> subsequently analysed by<br />
Aghi<strong>on</strong> et al. (2001) <strong>and</strong> Aghi<strong>on</strong> et al. (2002), the study of the relati<strong>on</strong>ship between product<br />
market competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> is carried out in the c<strong>on</strong>text of a model of endogenous growth<br />
with vertical differentiati<strong>on</strong> in a duopolistic industry where the incumbent firm can innovate.<br />
The model c<strong>on</strong>siders a sequence of tournaments with step-by-step innovati<strong>on</strong>s: the technological<br />
follower has first to catch up with the leading-edge technology before being able to become the<br />
new technological leader. This assumpti<strong>on</strong> can be justified <strong>on</strong> the grounds that innovative<br />
knowledge is to some extent tacit so that a rival has to invest in its own R&D to catch up <strong>and</strong><br />
cannot simply rely <strong>on</strong> knowledge spillovers.<br />
Aghi<strong>on</strong>, Harris <strong>and</strong> Vickers (1997) focus <strong>on</strong> the comparis<strong>on</strong> between Cournot <strong>and</strong> Bertr<strong>and</strong><br />
competiti<strong>on</strong> to capture different levels of market competiti<strong>on</strong>. Aghi<strong>on</strong> et al. (2001) <strong>and</strong> Aghi<strong>on</strong> et<br />
al. (2002) use the same model to investigate the relati<strong>on</strong>ship between innovati<strong>on</strong> <strong>and</strong> competiti<strong>on</strong><br />
where the latter is captured in terms of a substitutability parameter at the product market<br />
competiti<strong>on</strong> level.<br />
30<br />
248
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
The substantial difference of these models compared to earlier Schumpeterian models lies in the<br />
fact that, by c<strong>on</strong>sidering the possibility that the incumbent innovates, it is the incremental profits of<br />
each firm that drive R&D investments <strong>and</strong> the impact of product market competiti<strong>on</strong> has to be<br />
evaluated both ex ante <strong>and</strong> ex post. Tougher product market competiti<strong>on</strong> might reduce a firm’s<br />
pre-innovati<strong>on</strong> rents by more than it reduces post-innovati<strong>on</strong> rents, hence it may increase R&D<br />
firms’ investments.<br />
In these models, technological progress does not involve leapfrogging. Hence, a typical industry<br />
can be in either of two states: it can be a leveled industry where firms compete neck-to-neck or<br />
an unleveled industry where <strong>on</strong>e firm is technologically ahead of the follower.<br />
The effect of an increase in product market competiti<strong>on</strong> would have to be evaluated c<strong>on</strong>sidering<br />
the impact in both leveled <strong>and</strong> unleveled industries. The impact is actually twofold: <strong>on</strong> the <strong>on</strong>e<br />
h<strong>and</strong> the degree of product market competiti<strong>on</strong> affects equilibrium R&D investments in each of<br />
the possible state of the industry (leveled <strong>and</strong> unleveled); additi<strong>on</strong>ally, product market<br />
competiti<strong>on</strong> has an impact <strong>on</strong> how often the industry will be in <strong>on</strong>e of either states.<br />
In a leveled industry increased competiti<strong>on</strong> would spur innovati<strong>on</strong> because it would increase the<br />
incentives of firms’ to get ahead of the rival (the new “escape competiti<strong>on</strong> effect”); in an unleveled<br />
industry, increased competiti<strong>on</strong> would reduce the incentives to innovate for the classic<br />
Schumpeterian effect.<br />
The reas<strong>on</strong>ing in Aghi<strong>on</strong> et al. (2002), which leads to the c<strong>on</strong>clusi<strong>on</strong> that the relati<strong>on</strong>ship between<br />
intensity of competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> has an inverted-U shape, is discussed below.<br />
C<strong>on</strong>sider the case of very large innovati<strong>on</strong>s, in which case the leader never innovates <strong>and</strong> the<br />
largest gap between the leader <strong>and</strong> the laggard is <strong>on</strong>e technological step.<br />
When there is not much product market competiti<strong>on</strong>, there is hardly any incentive for firms to<br />
innovate when they compete neck-to-neck, <strong>and</strong> the overall innovati<strong>on</strong> rate will be highest when<br />
the industry is in an asymmetric state. Thus the industry will spend relatively more time in the<br />
levelled state where the escape-competiti<strong>on</strong> effect dominates. An increase in product market<br />
competiti<strong>on</strong> would then result in larger incentives for the firms to innovate. Hence, if the degree of<br />
competiti<strong>on</strong> is initially low, an increase will result in a faster innovati<strong>on</strong> rate.<br />
When initial product market competiti<strong>on</strong> is high, there is relatively little incentive for the laggard<br />
firm in an unleveled state to innovate <strong>and</strong> a relatively large incentive for a neck-to-neck firm to<br />
leave the leveled state. The c<strong>on</strong>sequence is that the industry will spend most of the time in the<br />
unleveled state where it is the Schumpeterian effect that dominates. Tougher product market<br />
competiti<strong>on</strong> reduces the Schumpeterian effect so that when the degree of competiti<strong>on</strong> is initially<br />
high, an increase may result in a slower average innovati<strong>on</strong> rate.<br />
These results extend to the general model with no upper bound to the technological gap.<br />
Aghi<strong>on</strong> et al. (2002) also provide empirical evidence that support their claim of an inverted-U<br />
relati<strong>on</strong>ship between intensity of competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong>.<br />
31<br />
249
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Aghi<strong>on</strong> et al. (2001) use the same framework to investigate the relati<strong>on</strong>ship between imitati<strong>on</strong> <strong>and</strong><br />
innovati<strong>on</strong> <strong>and</strong> c<strong>on</strong>clude that this is also characterized by an inverted-U shape. Again this may<br />
be the case because easier imitati<strong>on</strong> may decrease a firm’s pre-innovati<strong>on</strong> rents by more than it<br />
reduces post-innovati<strong>on</strong> rent.<br />
Denicolò <strong>and</strong> Zanchettin (2002) c<strong>on</strong>sider the relati<strong>on</strong>ship between competiti<strong>on</strong> <strong>and</strong> growth in the<br />
case of n<strong>on</strong>-drastic innovati<strong>on</strong>s by comparing innovati<strong>on</strong> <strong>and</strong> growth under Bertr<strong>and</strong> <strong>and</strong> Cournot<br />
competiti<strong>on</strong> in a st<strong>and</strong>ard leapfrogging model. The main finding of this study is that:<br />
“[…] when the size of innovati<strong>on</strong>s is sufficiently large, the equilibrium rate of growth is<br />
unambiguously greater with Bertr<strong>and</strong> than with Cournot competiti<strong>on</strong>. For smaller<br />
innovati<strong>on</strong>, Cournot competiti<strong>on</strong> may (but need not) create greater incentive to innovate.<br />
The intuiti<strong>on</strong> is that more competiti<strong>on</strong> entails lower prices but at the same time leads to greater<br />
productive efficiency (hence greater rates of growth) in that it lowers the market shares of less<br />
productive firms (Aghi<strong>on</strong> <strong>and</strong> Shankerman, 2000). When firms are symmetric <strong>on</strong>ly the first effect is<br />
at work. When technical progress creates asymmetries between firms the productive efficiency effect<br />
becomes important <strong>and</strong> the effect of more competiti<strong>on</strong> <strong>on</strong> innovators’ profits is generally ambiguous.<br />
However, the productive efficiency effect must dominate if innovati<strong>on</strong>s are close to being drastic.<br />
D’Asprem<strong>on</strong>t et al. (2002) investigate the relati<strong>on</strong>ship between competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> in a<br />
tournament model under uncertainty where, unlike Aghi<strong>on</strong> et al. (2001), the number of firms in<br />
each industry is endogenously determined <strong>and</strong> the race entails the possibility of several<br />
simultaneous winners <strong>and</strong> spillover effects. The main c<strong>on</strong>clusi<strong>on</strong> of the paper is that the<br />
relati<strong>on</strong>ship between toughness of market competiti<strong>on</strong> <strong>and</strong> the incentives to innovate is n<strong>on</strong>m<strong>on</strong>ot<strong>on</strong>e.<br />
The key for this result lies in the possible multiplicity of successful innovators in the<br />
same period.<br />
Introducti<strong>on</strong> of Agency C<strong>on</strong>siderati<strong>on</strong>s<br />
The adopti<strong>on</strong> of a step-by-step technological assumpti<strong>on</strong> is not the <strong>on</strong>ly means by which the<br />
relati<strong>on</strong>ship between competiti<strong>on</strong> <strong>and</strong> growth takes a different nature that in the traditi<strong>on</strong>al<br />
Schumpeterian model. Aghi<strong>on</strong> <strong>and</strong> Howitt (1998) c<strong>on</strong>sider also an alternative approach, based<br />
<strong>on</strong> a study published later by Aghi<strong>on</strong>, Dewatrip<strong>on</strong>t <strong>and</strong> Rey (1999) that leads to c<strong>on</strong>clusi<strong>on</strong>s <strong>on</strong><br />
the relati<strong>on</strong>ship between competiti<strong>on</strong> <strong>and</strong> growth that are different from those of the classic<br />
Schumpeterian model.<br />
This approach is based <strong>on</strong> relaxing the traditi<strong>on</strong>al behavioural assumpti<strong>on</strong> that firms are profit<br />
maximizing. N<strong>on</strong>-profit maximizing behavior is a comm<strong>on</strong> assumpti<strong>on</strong> in the literature that<br />
emphasises the existence of agency problems between producers <strong>and</strong> their financiers. Aghi<strong>on</strong> et<br />
al (1997) c<strong>on</strong>sider a principal-agent model of the firms where managers are interested in<br />
maximizing some private benefits of c<strong>on</strong>trol while minimizing private costs that innovati<strong>on</strong> would<br />
entail such as the costs of reorganizing the firm or training costs.<br />
32<br />
250
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Within this new setting, the impact of intensity of competiti<strong>on</strong> <strong>on</strong> growth can be positive because<br />
tougher competiti<strong>on</strong> may lead to managers to speed up the adopti<strong>on</strong> of new technologies to avoid<br />
bankruptcy <strong>and</strong> the loss of c<strong>on</strong>trol rights <strong>and</strong> hence to faster ec<strong>on</strong>omic growth.<br />
The literature <strong>on</strong> the importance of competiti<strong>on</strong> for incentives is extensive <strong>and</strong> it suggests that<br />
there are other important mechanisms though which competiti<strong>on</strong> may have a positive impact <strong>on</strong><br />
incentives is through the improvement of the ability of assessing managers’ efforts <strong>and</strong> hence of<br />
m<strong>on</strong>itoring them. At the same time, similarly to other areas of research, theoretical predicti<strong>on</strong>s of<br />
these studies are not as immediate as these simple c<strong>on</strong>siderati<strong>on</strong>s may suggest.<br />
Nickell (1996) provides empirical evidence <strong>on</strong> the relati<strong>on</strong>ship between intensity of competiti<strong>on</strong><br />
<strong>and</strong> innovati<strong>on</strong> by studying the impact of competiti<strong>on</strong> <strong>on</strong> both the level <strong>and</strong> the growth of total<br />
factor productivity of a large number of UK manufacturing companies. The main findings of this<br />
paper are that<br />
•<br />
•<br />
market power, which is captured in terms of market share, generates lower levels of<br />
productivity; <strong>and</strong> that<br />
competiti<strong>on</strong>, which is captured by the number of competitors or by lower levels or rents,<br />
leads to faster total factor productivity growth.<br />
Blundell, Griffith <strong>and</strong> van Reenen (1999) also study the relati<strong>on</strong>ship between competiti<strong>on</strong> <strong>and</strong><br />
innovati<strong>on</strong>. Their study provides evidence that<br />
“[…] within industries in was the high market share firms who tended to<br />
commercializemore innovati<strong>on</strong> although increased product market competiti<strong>on</strong> in the<br />
industry tended to<br />
stimulate innovative activity” (p. 550)<br />
For a very comprehensive survey of empirical studies <strong>on</strong> competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> see the<br />
recent OECD study by Ahn (2002).<br />
1.4 New Ec<strong>on</strong>omic Geography<br />
New ec<strong>on</strong>omic geography was effectively initiated by Krugman (1991). The main goal of this<br />
str<strong>and</strong> of research is to provide a theoretical explanati<strong>on</strong> of why ec<strong>on</strong>omic activity <strong>and</strong> populati<strong>on</strong><br />
tend to c<strong>on</strong>centrate, that is to say, why cities <strong>and</strong> industrial belts come into existence <strong>and</strong> keep <strong>on</strong><br />
exp<strong>and</strong>ing at the expense of the periphery. This approach is similar to new growth theory in that<br />
the models assume imperfect competiti<strong>on</strong> <strong>and</strong> product differentiati<strong>on</strong>, ec<strong>on</strong>omies of scale <strong>and</strong><br />
increasing returns.<br />
33<br />
251
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
The underlying story is that manufactured goods will be produced in regi<strong>on</strong>s where nearby<br />
dem<strong>and</strong> is higher, as this allows firms to minimize transport costs. Dem<strong>and</strong> will come either from<br />
agriculture or from the manufacturing sector itself (manufacturing workers). Because of a<br />
possible positive feedback between producti<strong>on</strong> of <strong>and</strong> dem<strong>and</strong> for manufactured goods,<br />
manufacturing producti<strong>on</strong> will tend to relocate into these regi<strong>on</strong>s.<br />
This phenomen<strong>on</strong> is formalized by the can<strong>on</strong>ical modelling of core <strong>and</strong> periphery (Krugman<br />
(1991)) that assumes that the ec<strong>on</strong>omy is split into two sectors - namely manufacturing <strong>and</strong><br />
agriculture – <strong>and</strong> into two regi<strong>on</strong>s (the core <strong>and</strong> periphery). In manufacturing, m<strong>on</strong>opolistic<br />
competiti<strong>on</strong> à la Dixit-Stiglitz (1977) prevails: manufacturing produces differentiated products in<br />
the presence of increasing returns. In additi<strong>on</strong>, labor, specific to manufacturing, is mobile<br />
between the two regi<strong>on</strong>s. Wages, both in nominal <strong>and</strong> real terms, can differ between the two<br />
regi<strong>on</strong>s. The driving force behind labor mobility is indeed differences in real wages.<br />
Transportati<strong>on</strong> costs when shipping from <strong>on</strong>e regi<strong>on</strong> to the other occur in the form of Samuels<strong>on</strong>’s<br />
iceberg cost (Samuels<strong>on</strong> (1954)). This means that transport costs are not explicitly c<strong>on</strong>sidered<br />
but rather it is assumed that <strong>on</strong>ly a fracti<strong>on</strong> of goods arrives, with the difference “smelting down”<br />
during transportati<strong>on</strong>.<br />
By c<strong>on</strong>trast, agriculture is characterized by perfect competiti<strong>on</strong> with homogenous goods <strong>and</strong><br />
c<strong>on</strong>stant returns to scale. Wages equal prices <strong>and</strong> labor, specific to agriculture, is immobile<br />
between the two regi<strong>on</strong>s. Furthermore, no transportati<strong>on</strong> costs are assumed to occur when<br />
goods are moved from <strong>on</strong>e regi<strong>on</strong> to the other. Assuming c<strong>on</strong>stant returns to scale <strong>and</strong> the<br />
absence of transport costs secures wage equalizati<strong>on</strong> in agriculture. It should be also noted that<br />
agriculture is evenly divided between the two regi<strong>on</strong>s.<br />
For c<strong>on</strong>centrati<strong>on</strong> in manufacturing to take place in <strong>on</strong>e of the regi<strong>on</strong>s, several c<strong>on</strong>diti<strong>on</strong>s have to<br />
be fulfilled. First, a large share of populati<strong>on</strong> should be allocated to the manufacturing sector so<br />
as it generates large dem<strong>and</strong> (that triggers the virtuous circle). Sec<strong>on</strong>d, ec<strong>on</strong>omies of scale<br />
should be high enough <strong>and</strong> transportati<strong>on</strong> costs low.<br />
Three cases can be c<strong>on</strong>sidered.<br />
1 First, in the event that transport costs are large, there is an equal divisi<strong>on</strong> of<br />
manufacturing between the two regi<strong>on</strong>s.<br />
2<br />
3<br />
If transportati<strong>on</strong> costs are low, manufacturing still can remain equally divided between<br />
regi<strong>on</strong>s. However, this situati<strong>on</strong> is not a stable <strong>on</strong>e (unstable equilibrium) for the following<br />
reas<strong>on</strong>. If manufacturing were <strong>on</strong>ly slightly more important in <strong>on</strong>e of the regi<strong>on</strong>s, this<br />
would lead to an increasing share of manufacturing in this sector accompanied by a<br />
dwindling of manufacturing in the other regi<strong>on</strong>. At the end of this process, manufacturing<br />
ends up being c<strong>on</strong>centrated in <strong>on</strong>e of the regi<strong>on</strong>s, <strong>and</strong> this can be referred to as the coreperiphery<br />
pattern.<br />
The third case, characterized by intermediate transportati<strong>on</strong> costs, can lead to either<br />
complete specializati<strong>on</strong> or equal divisi<strong>on</strong> between regi<strong>on</strong>s.<br />
34<br />
252
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Extensi<strong>on</strong>s<br />
The three-regi<strong>on</strong> model<br />
A possible extensi<strong>on</strong> of the basic model is to simultaneously c<strong>on</strong>sider three regi<strong>on</strong>s. Keeping the<br />
assumpti<strong>on</strong>s of the two-regi<strong>on</strong> model would yield basically the same results: No c<strong>on</strong>centrati<strong>on</strong> of<br />
manufacturing activity in the presence of high transport costs, extreme c<strong>on</strong>centrati<strong>on</strong> in <strong>on</strong>e of the<br />
three regi<strong>on</strong>s if transport costs are low, <strong>and</strong> the combinati<strong>on</strong> of the two previous situati<strong>on</strong>s in the<br />
case of intermediate transport costs.<br />
Transport costs in agriculture<br />
The sec<strong>on</strong>d extensi<strong>on</strong> of the basic two-regi<strong>on</strong> model c<strong>on</strong>sists in releasing the assumpti<strong>on</strong> related<br />
to the absence of transport costs in agriculture. Empirical evidence clearly indicates that transport<br />
costs exist not <strong>on</strong>ly for differentiated goods produced in the manufacturing sector, but also in the<br />
other sectors, <strong>and</strong> that these costs are likely to be at least as high as the cost for manufacturing<br />
products. For this reas<strong>on</strong>, iceberg-type transport costs are allowed for in agriculture. As a<br />
corollary, wages <strong>and</strong> c<strong>on</strong>sequently prices of agricultural products do not equalise between the two<br />
sectors.<br />
The introducti<strong>on</strong> of transportati<strong>on</strong> costs for agricultural products when shipped from <strong>on</strong>e regi<strong>on</strong> to<br />
the other results in higher agricultural prices in the destinati<strong>on</strong> regi<strong>on</strong> (with highly c<strong>on</strong>centrated<br />
manufacturing at the expenses of the other regi<strong>on</strong>). Hence, higher prices for agricultural<br />
products, translated into higher costs of living would actually discourage labour mobility between<br />
the two regi<strong>on</strong>s, <strong>and</strong> thus hinder the c<strong>on</strong>centrati<strong>on</strong> of manufacturing activity.<br />
It seems very unrealistic to assume that the same homogenous agricultural good is produced in<br />
the two regi<strong>on</strong>s. The next step is therefore to c<strong>on</strong>sider differentiated agricultural products. The<br />
results are similar to the previous <strong>on</strong>es. In general, it is possible to reach the c<strong>on</strong>clusi<strong>on</strong> that<br />
while high transport costs in agriculture tend to prohibit c<strong>on</strong>centrati<strong>on</strong> in manufacturing, a<br />
decrease in costs leads to c<strong>on</strong>centrati<strong>on</strong>.<br />
Urban agglomerati<strong>on</strong>s<br />
A sec<strong>on</strong>d str<strong>and</strong> of the new ec<strong>on</strong>omic geography is c<strong>on</strong>cerned with urban agglomerati<strong>on</strong>s <strong>and</strong><br />
attempts to answer questi<strong>on</strong>s as follows:<br />
•<br />
•<br />
•<br />
•<br />
How do cities come into existence <strong>and</strong> why do they exist in the l<strong>on</strong>ger run?<br />
How do new cities form <strong>and</strong> which are the mechanisms underlying the growth of cities?<br />
What is the reas<strong>on</strong> that cities turn out to be linked <strong>on</strong>e to another through a hierarchy?<br />
What is the role of natural advantages, such as harbors, in the development of cities?<br />
35<br />
253
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Internati<strong>on</strong>al trade<br />
New trade theory, as developed by Krugman (1980), has much in comm<strong>on</strong> with new ec<strong>on</strong>omic<br />
geography. The former aims at investigating internati<strong>on</strong>al specializati<strong>on</strong> phenomena in the case<br />
of two or more countries, which can be viewed as the equivalent of agglomerati<strong>on</strong> <strong>and</strong><br />
c<strong>on</strong>centrati<strong>on</strong> within an individual country. It should be noted that the logic <strong>and</strong> the mechanisms<br />
at work are very similar <strong>and</strong> what makes the difference are different restricti<strong>on</strong>s to the basic model<br />
as discussed below.<br />
Fujita et al. (2001) develops a model with intermediate goods. In this two-country framework,<br />
each individual country is composed of two sectors, namely manufacturing <strong>and</strong> agriculture with<br />
labor being assumed to be perfectly mobile between agriculture <strong>and</strong> the manufacturing sector.<br />
By c<strong>on</strong>trast, labor is assumed to be immobile between the two countries. It is worth drawing<br />
attenti<strong>on</strong> to the fact that this model corresp<strong>on</strong>ds exactly to the can<strong>on</strong>ical model with the difference<br />
that two countries instead of two regi<strong>on</strong>s are c<strong>on</strong>sidered. There is no labor mobility between the<br />
two countries (regi<strong>on</strong>s) <strong>and</strong> intermediate goods are added to the model.<br />
The introducti<strong>on</strong> of intermediary goods is normally associated with input-output matrices. Put<br />
differently, firms are assumed to be linked with each other through input <strong>and</strong> output linkages. The<br />
model avoids the inclusi<strong>on</strong> of new sectors by assuming that the range of goods produced in<br />
manufacturing serves not <strong>on</strong>ly for final c<strong>on</strong>sumpti<strong>on</strong> but also as input for manufacturing itself.<br />
As a result, manufacturing uses labor <strong>and</strong> intermediary goods as inputs. In accordance with the<br />
model, the greater the gamut of intermediary goods, the lower the costs of producti<strong>on</strong>. In additi<strong>on</strong><br />
to that, if those goods are available locally, prices do not include transport costs. As far as sales<br />
are c<strong>on</strong>cerned, what firms produce is sold either to c<strong>on</strong>sumer as final goods or to manufacturing<br />
firms as intermediate goods.<br />
On the <strong>on</strong>e h<strong>and</strong>, high trade costs leave manufacturing equally divided between the two countries<br />
because firms have to supply c<strong>on</strong>sumers. On the other h<strong>and</strong>, low trade costs translate into the<br />
full c<strong>on</strong>centrati<strong>on</strong> of manufacturing activity in <strong>on</strong>e of the countries <strong>on</strong> the grounds of backward<br />
<strong>and</strong> forward linkages. In the presence of intermediate trade costs, c<strong>on</strong>centrati<strong>on</strong> is possible but<br />
not inevitable<br />
This model also manages to give an explanati<strong>on</strong> for differing nominal <strong>and</strong> real wage levels in<br />
different countries. According to the model, the real wage in the country where manufacturing<br />
c<strong>on</strong>centrates rises relative to those in the other country. One reas<strong>on</strong> for this is that dem<strong>and</strong> for<br />
labor increases <strong>and</strong> so does the wage level as a c<strong>on</strong>sequence of c<strong>on</strong>centrati<strong>on</strong> in the first<br />
country. Another reas<strong>on</strong> is that the country without manufacturing producti<strong>on</strong> has to pay the bill of<br />
transport costs of the imported manufactured goods, which in turn decreases the real wage.<br />
36<br />
254
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Ec<strong>on</strong>omic development: a sequential process<br />
In this stage of the analysis, an exogenous growth process is added to the model <strong>and</strong> it is<br />
assumed that dem<strong>and</strong> for manufactured goods rises faster than potential supply. A model<br />
c<strong>on</strong>sisting of two countries is c<strong>on</strong>sidered. The model seeks to explain under which c<strong>on</strong>diti<strong>on</strong>s<br />
industrial c<strong>on</strong>centrati<strong>on</strong> starts spilling out from <strong>on</strong>e country to the other.<br />
The point of departure is to c<strong>on</strong>sider <strong>on</strong>e country with highly c<strong>on</strong>centrated manufacturing. It is<br />
shown that c<strong>on</strong>centrati<strong>on</strong> will c<strong>on</strong>tinue until the gains of backward <strong>and</strong> forward linkages in the<br />
industry exceed the potential loss that could be associated with the fact that wages are<br />
c<strong>on</strong>siderable higher in the industrialized country. If the losses are higher than the gains (that is<br />
the wage gap exceeds a given threshold) firms start moving into the other country. These firms<br />
then create their inter-firm linkages in the other country, which, through a positive feedback<br />
process will attract other firms <strong>and</strong> thus accelerate the c<strong>on</strong>centrati<strong>on</strong> process.<br />
However, there are some interesting questi<strong>on</strong>s to answer. For instance, which industry is the first<br />
to move, <strong>and</strong> how does the industrial structure of countries at different stages of industrializati<strong>on</strong><br />
change? This is studied using a five-country model in which each country is composed of seven<br />
different industries. According to the model, the most labor intensive industry will be the first to<br />
move into the other country, followed progressively by less-labor intensive industries. However,<br />
the latter will move faster.<br />
The authors suggest that the inter-industry structure of the ec<strong>on</strong>omies changes depending <strong>on</strong><br />
several factors, including:<br />
•<br />
•<br />
the higher the share of final c<strong>on</strong>sumpti<strong>on</strong> goods in sales, the higher the probability that<br />
firms move (forward linkage); <strong>and</strong><br />
the lower the intermediate input requirement, the quicker firm moves (backward linkage).<br />
Industrial clustering<br />
Fujita et al. (2001) c<strong>on</strong>struct a two-country, two-industry model with a single factor of producti<strong>on</strong>.<br />
Agriculture is dropped in the analysis <strong>and</strong> two manufacturing industries are used instead, both<br />
m<strong>on</strong>opolistically competitive. The questi<strong>on</strong> is: under which circumstances do c<strong>on</strong>centrati<strong>on</strong> or<br />
dispersi<strong>on</strong> occur?<br />
The model provides us with two c<strong>on</strong>clusi<strong>on</strong>s. First, in the event that <strong>on</strong>e industry is str<strong>on</strong>gly<br />
c<strong>on</strong>nected to the other industry through input-output linkages, c<strong>on</strong>centrati<strong>on</strong> will not happen. In<br />
c<strong>on</strong>trast, in the case of str<strong>on</strong>g intra-industry links, the model predicts a high c<strong>on</strong>centrati<strong>on</strong>, that is,<br />
each country specializes in <strong>on</strong>e of the industrial activities. When c<strong>on</strong>sidering trade costs, the<br />
results are similar to the previous <strong>on</strong>es: the two industries remain present in both countries if<br />
trade costs are high, but a dwindling in trade costs allows for <strong>and</strong> subsequently leads to<br />
specializati<strong>on</strong> of the countries.<br />
37<br />
255
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
The authors suggest that with deeper ec<strong>on</strong>omic integrati<strong>on</strong> bringing about lower trade costs,<br />
European ec<strong>on</strong>omies are expected to undergo a c<strong>on</strong>centrati<strong>on</strong> process <strong>and</strong> will c<strong>on</strong>verge<br />
towards the ec<strong>on</strong>omic structure of the US.<br />
The two-country, two-industry framework is then replaced by a model including more than two<br />
industrial sectors. Although the model cannot predict why <strong>on</strong>e industry c<strong>on</strong>centrates in <strong>on</strong>e<br />
country <strong>and</strong> not another, it can provide a theoretical underpinning in favor of the argument that in<br />
case <strong>on</strong>e country loses a particular industry as a result of some sort of exogenous shock, there<br />
exists no mechanism that could ensure that, after the shock, the industry in questi<strong>on</strong> regains<br />
ground in the country under c<strong>on</strong>siderati<strong>on</strong>.<br />
Empirical works related to industrial clustering show that the US experienced a high c<strong>on</strong>centrati<strong>on</strong><br />
of ec<strong>on</strong>omic activities <strong>and</strong> a high degree of industrial clustering chiefly during the period prior to<br />
WW-I. (see e.g. Kim (1995)). Patterns of c<strong>on</strong>centrati<strong>on</strong> <strong>and</strong> industrial clustering remained<br />
relatively unchanged afterwards. Studies focusing <strong>on</strong> Europe indicate a str<strong>on</strong>g c<strong>on</strong>centrati<strong>on</strong> <strong>and</strong><br />
clustering which has occurred during the last 20 years or so. Even so, the level of c<strong>on</strong>centrati<strong>on</strong><br />
<strong>and</strong> sectoral clustering in Europe is not as high as in the US (e.g. Midelfart-Knarvik et al. (2000)).<br />
In a recent study with a str<strong>on</strong>g cliometric influence (see secti<strong>on</strong> 2), Davis <strong>and</strong> Weinstein (2002)<br />
set out to investigate agglomerati<strong>on</strong> patterns in Japan based <strong>on</strong> a unique data set stretching over<br />
8000 years. Data drawn from archaeological findings <strong>and</strong> populati<strong>on</strong> census indicate that<br />
dispersi<strong>on</strong> in populati<strong>on</strong> has been historically high in Japan. However, variati<strong>on</strong> in populati<strong>on</strong><br />
density c<strong>on</strong>siderably widened from the outset of the industrial revoluti<strong>on</strong>. The authors argue that<br />
although increasing returns can hardly explain the origins of geographical agglomerati<strong>on</strong>, they<br />
play a prominent role in explaining temporal dynamics, i.e. the accelerati<strong>on</strong> of geographical<br />
c<strong>on</strong>centrati<strong>on</strong> throughout the last 200 years. New light is also shed <strong>on</strong> the temporary nature of<br />
even very large shocks. For instance, the massive bombardment of Japanese cities at the end of<br />
WW-II does not appear to have permanently altered the spatial structure of the Japanese industrial<br />
structure.<br />
38<br />
256
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
2 THE EVOLUTIONARY ECONOMICS OF INNOVATION<br />
This secti<strong>on</strong> discusses the main findings from the evoluti<strong>on</strong>ary approach to the study of<br />
innovati<strong>on</strong> <strong>and</strong> technical change. It then discusses the relati<strong>on</strong>ship between innovati<strong>on</strong> <strong>and</strong><br />
instituti<strong>on</strong>s (known as “systems of innovati<strong>on</strong>”).<br />
2.1 Introducti<strong>on</strong><br />
The evoluti<strong>on</strong>ary approach to the study of innovati<strong>on</strong> <strong>and</strong> technological change arose largely<br />
because the mainstream theory was c<strong>on</strong>sidered inadequate to tackle the inherent disequilibrium<br />
dynamics that characterise technological innovati<strong>on</strong>. Nels<strong>on</strong> <strong>and</strong> Winter (1984) point out that<br />
both the ec<strong>on</strong>omy as a whole <strong>and</strong> the ec<strong>on</strong>omic actors taken individually can be c<strong>on</strong>sidered to<br />
evolve c<strong>on</strong>tinuously, similar to the evoluti<strong>on</strong>ary process described in biology. Hence the name of<br />
evoluti<strong>on</strong>ary ec<strong>on</strong>omics.<br />
Dissatisfacti<strong>on</strong> with the traditi<strong>on</strong>al approach is almost pervasive <strong>and</strong> embraces the definiti<strong>on</strong> of<br />
innovati<strong>on</strong> <strong>and</strong> knowledge, the descripti<strong>on</strong> of the R&D process <strong>and</strong> more generally of the<br />
innovati<strong>on</strong> process, the modelling approach based <strong>on</strong> the noti<strong>on</strong> of full rati<strong>on</strong>ality <strong>and</strong> equilibrium,<br />
<strong>and</strong> the lack of c<strong>on</strong>siderati<strong>on</strong> of instituti<strong>on</strong>al factors.<br />
Though these themes trace back to earlier researchers — in particular, the Schumpeterian<br />
perspective, which places emphasis <strong>on</strong> dynamics <strong>and</strong> competiti<strong>on</strong> through innovati<strong>on</strong> — dynamic<br />
theorizing in evoluti<strong>on</strong>ary ec<strong>on</strong>omics is often seen to start with the models of Nels<strong>on</strong> <strong>and</strong> Winter<br />
(1982).<br />
Although evoluti<strong>on</strong>ary ec<strong>on</strong>omics is perhaps not a coherent theory with a comm<strong>on</strong> methodology,<br />
it may be categorised as a group of various approaches that share the same criticisms of the neoclassical<br />
approach. For example:<br />
• First, evoluti<strong>on</strong>ary ec<strong>on</strong>omics does not treat ec<strong>on</strong>omic agents as rati<strong>on</strong>al, in the sense<br />
they are in neo-classical ec<strong>on</strong>omics. They do not maximize returns through calculating<br />
the optimal strategy from a range of probabilistic outcomes.<br />
•<br />
Sec<strong>on</strong>d, evoluti<strong>on</strong>ary ec<strong>on</strong>omics does not seek to analyse ec<strong>on</strong>omic phenomena from<br />
equilibrium outcomes. In particular, dynamic change within the ec<strong>on</strong>omy is c<strong>on</strong>sidered<br />
best analyzed in terms of the processes that govern it, which are inherently disequilibrium<br />
dynamics.<br />
Evoluti<strong>on</strong>ary models draw their theoretical foundati<strong>on</strong>s <strong>on</strong> different sources <strong>and</strong> particularly <strong>on</strong><br />
biological models, evoluti<strong>on</strong>ary game theory <strong>and</strong> instituti<strong>on</strong>alism in order to capture essentially<br />
dynamic <strong>and</strong> irreversible processes.<br />
The literature tends to place emphasis <strong>on</strong> innovati<strong>on</strong> <strong>and</strong> selecti<strong>on</strong> from variety as the force of<br />
ec<strong>on</strong>omic development. Of particular importance is the need to see innovati<strong>on</strong> as more than<br />
simple technological development (<strong>and</strong> much more than the producti<strong>on</strong> of knowledge as a public<br />
good) <strong>and</strong> to analyze ec<strong>on</strong>omic development through technological <strong>and</strong> organizati<strong>on</strong>al<br />
improvements (which lead to the introducti<strong>on</strong> of new products, inputs <strong>and</strong> techniques).<br />
39<br />
257
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
In evoluti<strong>on</strong>ary theories innovati<strong>on</strong> is in fact defined more broadly than in mainstream models <strong>and</strong><br />
as a multidimensi<strong>on</strong>al phenomen<strong>on</strong>. Innovati<strong>on</strong> is not <strong>on</strong>ly a reducti<strong>on</strong> in producti<strong>on</strong> costs or the<br />
commercializati<strong>on</strong> of new products but also a more efficient reorganizati<strong>on</strong> of a firm’s activities,<br />
the adopti<strong>on</strong> of more sophisticated managerial techniques, access to new markets, etc.<br />
This richer definiti<strong>on</strong> emphasizes that innovati<strong>on</strong> extends outside the realm of technology in two<br />
ways: many innovati<strong>on</strong>s are not technological but involve other aspects of a firm’s activities such<br />
as its organizati<strong>on</strong>al structure; moreover, even technological innovati<strong>on</strong>s often affect n<strong>on</strong>technological<br />
activities of a firm: technological <strong>and</strong> organizati<strong>on</strong>al changes are often importantly<br />
intertwined.<br />
Another important difference with mainstream models of innovati<strong>on</strong> is that evoluti<strong>on</strong>ary models<br />
emphasize the tacit <strong>and</strong> specific nature of knowledge, as opposed to the idea of knowledge as<br />
informati<strong>on</strong> that somehow underlies mainstream models of innovati<strong>on</strong>.<br />
In terms of the process that leads to the discovery of new products, evoluti<strong>on</strong>ary models describe<br />
the importance of learning, <strong>and</strong> of organizati<strong>on</strong>al learning often embodied in organizati<strong>on</strong>al<br />
routines as a source of knowledge. The process through which innovati<strong>on</strong>s emerge is by no<br />
means linear, but it is characterized by complex feedback mechanisms <strong>and</strong> interactive relati<strong>on</strong>s<br />
involving science, technology, learning, producti<strong>on</strong>, instituti<strong>on</strong>s, organizati<strong>on</strong>, policy makers <strong>and</strong><br />
dem<strong>and</strong> (Edquist 1999).<br />
Evoluti<strong>on</strong>ary ec<strong>on</strong>omics can be looked at both from a microec<strong>on</strong>omic <strong>and</strong> macroec<strong>on</strong>omic<br />
perspective. The microec<strong>on</strong>omic aspects of evoluti<strong>on</strong>ary ec<strong>on</strong>omics are mainly c<strong>on</strong>cerned with<br />
how innovati<strong>on</strong> arises at the firm <strong>and</strong> market level. In c<strong>on</strong>trast, the macroec<strong>on</strong>omic str<strong>and</strong> of<br />
evoluti<strong>on</strong>ary ec<strong>on</strong>omics places innovati<strong>on</strong> into the social <strong>and</strong> ec<strong>on</strong>omic c<strong>on</strong>text <strong>and</strong> examines the<br />
dynamic process through which the innovati<strong>on</strong> systems generate innovati<strong>on</strong>s through a complex<br />
<strong>and</strong> reciprocal relati<strong>on</strong>ship network c<strong>on</strong>necting different comp<strong>on</strong>ents of the system.<br />
2.2 Microec<strong>on</strong>omics of Innovati<strong>on</strong><br />
Knowledge <strong>and</strong> rati<strong>on</strong>ality<br />
At the heart of evoluti<strong>on</strong>ary ec<strong>on</strong>omics is a definiti<strong>on</strong> of knowledge <strong>and</strong> of its characteristics that is<br />
substantially different for the neo-classical assimilati<strong>on</strong> of innovati<strong>on</strong> to informati<strong>on</strong>. In each<br />
technology there are elements of tacit <strong>and</strong> specific knowledge that cannot be formalised <strong>and</strong><br />
cannot, therefore, be diffused in the form of informati<strong>on</strong>. These firm-specific, local <strong>and</strong> cumulative<br />
features of knowledge are, to various extents, essential characteristics of innovati<strong>on</strong>.<br />
The term tacit knowledge is used to encapsulate the idea of knowledge that is found to be useful<br />
(e.g. in an organizati<strong>on</strong>al routine or a skill) without being directly accessible to c<strong>on</strong>sciousness<br />
(Pavitt, 2002) or articulable (Winter, 1987). Thus individuals can be c<strong>on</strong>sidered not to underst<strong>and</strong><br />
all that they know — tacit knowledge cannot be fully explained, written down or otherwise<br />
diffused; yet it remains valuable. By c<strong>on</strong>trast the term informati<strong>on</strong> can be taken to mean ‘explicit’<br />
knowledge, which can be codified <strong>and</strong> transmitted.<br />
40<br />
258
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Informati<strong>on</strong>, as explicit knowledge, can be diffused around the ec<strong>on</strong>omy (<strong>and</strong> fits most closely<br />
with the neo-classical c<strong>on</strong>cept of knowledge as a public good). Tacit knowledge cannot be<br />
passed from agent to agent in this way, but it is transferred <strong>and</strong> transmitted implicitly. Firms, <strong>and</strong><br />
other ec<strong>on</strong>omic organizati<strong>on</strong>s, can be seen as vehicles through which tacit knowledge is<br />
replicated through routines <strong>and</strong> techniques (Foster, 2001).<br />
Furthermore, as menti<strong>on</strong>ed in secti<strong>on</strong> 1.2.5, drawing mutual benefits from tacit knowledge that<br />
cannot be explicitly transferred between firms may bring benefits to inter-firm co-operati<strong>on</strong> in<br />
innovative activity that complement those benefits identified in the industrial organizati<strong>on</strong><br />
literature.<br />
Support is found for this c<strong>on</strong>cept of tacit knowledge from psychological experiments. Nels<strong>on</strong> <strong>and</strong><br />
Winter (2002) cite studies that examine the difference between processes acquired through<br />
rati<strong>on</strong>ality (underst<strong>and</strong>ing the underlying nature of the problem <strong>and</strong> calculating strategies for<br />
success) <strong>and</strong> learning-by-doing (learning successful strategies through trial <strong>and</strong> error without a<br />
complete theory <strong>on</strong> why such strategies are successful). They note that “a close logical<br />
c<strong>on</strong>necti<strong>on</strong> between a learned task <strong>and</strong> a newly presented task does not necessarily indicate a<br />
potential for easy transfer”. Thus there is an important difference between knowledge from<br />
underst<strong>and</strong>ing <strong>and</strong> knowledge from practice (tacit knowledge).<br />
Such results indicate that the more knowledge that is embedded in firms, through working<br />
practices, research programs <strong>and</strong> <strong>on</strong>going developments, is tacit, the less transferable this<br />
knowledge is to outsiders. To some extent, if a firm cannot rati<strong>on</strong>alize why a particular routine has<br />
proved beneficial, but c<strong>on</strong>tinues with it <strong>on</strong> the basis of past performance, then it will be less able<br />
to adapt to a slight change in c<strong>on</strong>diti<strong>on</strong>s <strong>and</strong> less able to apply successful strategies to slightly<br />
different problems. Thus, it is perhaps unrealistic to expect firms to re-optimize their strategies<br />
immediately when circumstances change — by its very nature a process of trial <strong>and</strong> error (or<br />
imitati<strong>on</strong>) takes time.<br />
In a similar manner, the idea of tacit knowledge helps explain the large internati<strong>on</strong>al differences in<br />
living st<strong>and</strong>ards <strong>and</strong> ec<strong>on</strong>omic development that seem inc<strong>on</strong>sistent with knowledge being a<br />
public good transferred <strong>and</strong> adopted at minimal cost. At a more micro level, it may be expected<br />
that different organizati<strong>on</strong>s may obtain different levels of output from ostensibly the same inputs.<br />
Evoluti<strong>on</strong>ary ec<strong>on</strong>omists not <strong>on</strong>ly disagree with the neoclassical traditi<strong>on</strong> <strong>on</strong> the definiti<strong>on</strong> of<br />
technological knowledge but also <strong>on</strong> fundamental assumpti<strong>on</strong>s <strong>on</strong> the behavior of ec<strong>on</strong>omic<br />
agents.<br />
Evoluti<strong>on</strong>ary ec<strong>on</strong>omics does not treat ec<strong>on</strong>omic agents as rati<strong>on</strong>al, in the sense they are in neoclassical<br />
ec<strong>on</strong>omics. This is not solely because ec<strong>on</strong>omic agents do not have full informati<strong>on</strong>, but<br />
more importantly because they have limited cognitive powers; they do not profit-maximize (given<br />
their informati<strong>on</strong> set) but rather aim at sufficiently high levels of success to enable survival in a<br />
dynamic envir<strong>on</strong>ment.<br />
41<br />
259
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Related to this, Dosi (1988) emphasizes that innovative search is characterized by str<strong>on</strong>g<br />
uncertainty whereby the list of possible events is unknown <strong>and</strong> <strong>on</strong>e does not know either the<br />
c<strong>on</strong>sequence of particular acti<strong>on</strong>s for any given event.<br />
Instead of being rati<strong>on</strong>al <strong>and</strong> thriving for maximizing profit, ec<strong>on</strong>omic actors are assumed to<br />
develop routine behaviors, which they follow as l<strong>on</strong>g as they turn out to work well. In case the<br />
routine behaviour fails to functi<strong>on</strong>, they will be revised <strong>and</strong> changed so that new routines may be<br />
developed. As a result of the evoluti<strong>on</strong>ary process, viable routines prevail whereas the less<br />
successful <strong>on</strong>es are eliminated. It must be menti<strong>on</strong>ed, however, that these changes are not<br />
taking place c<strong>on</strong>tinuously but are rather unevenly distributed over time <strong>and</strong> tend to occur<br />
sporadically.<br />
Innovati<strong>on</strong> <strong>and</strong> competiti<strong>on</strong><br />
Variety is an important part of evoluti<strong>on</strong>ary theories of innovati<strong>on</strong> <strong>and</strong> competiti<strong>on</strong> <strong>and</strong> can be<br />
traced back to the assumpti<strong>on</strong>s <strong>on</strong> limits to ec<strong>on</strong>omic agents’ rati<strong>on</strong>ality.<br />
A classic issue raised by the c<strong>on</strong>cept of Schumpeterian competiti<strong>on</strong> as a process of creative<br />
destructi<strong>on</strong> is that this, <strong>on</strong> its own, would imply that ec<strong>on</strong>omic growth leads to suppressi<strong>on</strong> of<br />
variety <strong>and</strong> ultimately to m<strong>on</strong>opoly (Nels<strong>on</strong> <strong>and</strong> Winter, 2002, p 35). Thus, to counter this<br />
predicti<strong>on</strong>, there is need to c<strong>on</strong>sider how new variety originates.<br />
Metcalfe (1998, p98) draws the link between the rejecti<strong>on</strong> of the neo-classical assumpti<strong>on</strong> of<br />
rati<strong>on</strong>al maximizing agents, <strong>and</strong> the generati<strong>on</strong> (<strong>and</strong> re-generati<strong>on</strong>) of variety:<br />
“…the positive side of bounded rati<strong>on</strong>ality is that it frees the imaginati<strong>on</strong> from the limits of<br />
calculati<strong>on</strong>. When problems become too complex to be well-defined, let al<strong>on</strong>e solved<br />
analytically, <strong>on</strong>e is inevitably dependent <strong>and</strong> judgment, c<strong>on</strong>jecture <strong>and</strong> the guiding h<strong>and</strong><br />
of experience.”<br />
The point here is that while rati<strong>on</strong>al agents, profit-maximizing under identical resource <strong>and</strong><br />
informati<strong>on</strong> c<strong>on</strong>straints should, by definiti<strong>on</strong>, reach the same strategy, more realistic agents,<br />
guided by judgment <strong>and</strong> experience, will reach a range of preferred strategies. This process<br />
can be expected to lead to significant variety in innovative soluti<strong>on</strong>s, within which there are<br />
elements of path dependence as agents bring past experience to bear <strong>on</strong> current problems.<br />
From these factors leading to the generati<strong>on</strong> of heterogeneity, competiti<strong>on</strong> proceeds to bring<br />
ec<strong>on</strong>omic development. As discussed below, this is because competiti<strong>on</strong> leads to improvement<br />
through selecti<strong>on</strong>, adaptati<strong>on</strong> <strong>and</strong> imitati<strong>on</strong>.<br />
Foster (2001) emphasizes how the traditi<strong>on</strong>al (mainstream) discussi<strong>on</strong> of competiti<strong>on</strong> in<br />
ec<strong>on</strong>omics has much to do with static outcomes <strong>and</strong> little to do with dynamics. By modelling<br />
competiti<strong>on</strong> in terms of equilibrium, competiti<strong>on</strong> is seen as timeless, rather than a process of<br />
development. A key feature of evoluti<strong>on</strong>ary ec<strong>on</strong>omics is its emphasis <strong>on</strong> competiti<strong>on</strong> as a<br />
process rather than an outcome.<br />
42<br />
260
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Adopting a more dynamic approach, evoluti<strong>on</strong>ary ec<strong>on</strong>omics is in part associated with the use of<br />
analogies to evoluti<strong>on</strong>ary biology to explain ec<strong>on</strong>omic growth <strong>and</strong> the process of competiti<strong>on</strong>. At<br />
a very basic level, the noti<strong>on</strong> of competiti<strong>on</strong> through natural selecti<strong>on</strong> associated with Darwin<br />
seems similar to the process by which ec<strong>on</strong>omic competiti<strong>on</strong> selects more fit (efficient <strong>and</strong><br />
profitable) firms at the expense of less fit firms.<br />
In this c<strong>on</strong>text, Mokyr (1999a) defines the Darwinian model “as a system of self-reproducing units<br />
(techniques) that changes over time.” Such a system appears to be best characterized by the<br />
following three properties:<br />
• The relati<strong>on</strong>ship through which the underlying structure determines the manifested<br />
entities. From an ec<strong>on</strong>omic viewpoint, an underlying structure is the “useful knowledge”<br />
that provides the framework for the “feasible” techniques (manifested entity) i.e. what<br />
society potentially is able to do.<br />
•<br />
•<br />
A Darwinian system is likely to undergo dramatic changes over time. Techniques<br />
reproduce themselves between two time periods through either repetiti<strong>on</strong> or learning <strong>and</strong><br />
imitati<strong>on</strong>. This process then entails c<strong>on</strong>stant changes part of which cannot be foreseen<br />
<strong>and</strong> which can turn out to destabilize the current system. An important questi<strong>on</strong> that this<br />
raises is the pace of innovati<strong>on</strong>. Is the innovati<strong>on</strong> process a gradual <strong>on</strong>e or does it occur<br />
in a stepwise manner? This is indeed a crucial issue (for example, in cliometrics as<br />
discussed below, the debate <strong>on</strong> whether the first industrial revoluti<strong>on</strong> as it occurred in Britain<br />
created disc<strong>on</strong>tinuity or whether it can be regarded as a smooth transiti<strong>on</strong>).<br />
Excessive variety in a Darwinian system implies that the actual number of techniques<br />
exceeds the number sustainable in the system. This results in a selecti<strong>on</strong> process.<br />
Mokyr (1999a) argues that selecti<strong>on</strong> can operate through three channels. The first <strong>on</strong>e is<br />
the st<strong>and</strong>ard neo-classical mechanism where techniques are chosen in the event they<br />
maximize an objective functi<strong>on</strong> incorporating supply <strong>and</strong> dem<strong>and</strong> side c<strong>on</strong>siderati<strong>on</strong> as<br />
well as externalities. Sec<strong>on</strong>d, there is inertia in the system that eliminates useless<br />
techniques. Finally, <strong>and</strong> more importantly, selecti<strong>on</strong> not <strong>on</strong>ly happens in the market but<br />
also at the social level.<br />
Inherent in this model of competiti<strong>on</strong> is the associati<strong>on</strong> between competiti<strong>on</strong> <strong>and</strong> experimentati<strong>on</strong>.<br />
A variety of experiments allows for ec<strong>on</strong>omic progress bey<strong>on</strong>d the scope that would be<br />
c<strong>on</strong>sidered achievable through comprehensive calculati<strong>on</strong> <strong>and</strong> strategy (i.e. if each agent had to<br />
specify all permutati<strong>on</strong>s of outcomes, the likelihood of each, <strong>and</strong> the interrelati<strong>on</strong>ships involved).<br />
Nels<strong>on</strong> <strong>and</strong> Winter (2002, p 28) identify that the ec<strong>on</strong>omic development ‘puzzle’ is how the vast<br />
technological progress <strong>and</strong> efficient forms of organizati<strong>on</strong> seen in the modern age could have<br />
arisen given the cognitive limits of humans <strong>and</strong> organizati<strong>on</strong>s. To this end, they propose that<br />
competiti<strong>on</strong> is the means through which the deficiency of human rati<strong>on</strong>ality <strong>and</strong> ability are<br />
mitigated to achieve favorable ec<strong>on</strong>omic outcomes:<br />
“Neo-classical ec<strong>on</strong>omics discovers social value in human selfishness, but not virtue<br />
that<br />
is robust against the human limitati<strong>on</strong> of incompetence — <strong>and</strong> the possible role of the<br />
market process in achieving that robustness is not featured.”<br />
43<br />
261
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Adopting this positi<strong>on</strong>, limited rati<strong>on</strong>ality, rather than being an impediment to well-functi<strong>on</strong>ing<br />
markets, is actually seen as part of the process of competiti<strong>on</strong>, <strong>and</strong> through it, innovati<strong>on</strong>. If<br />
agents possessed far greater cognitive powers <strong>and</strong> rati<strong>on</strong>ality, the importance of competiti<strong>on</strong> in<br />
delivering ec<strong>on</strong>omic development would be reduced.<br />
Arguably there is an interesting corresp<strong>on</strong>dence between this c<strong>on</strong>cept of competiti<strong>on</strong> <strong>and</strong> the role<br />
of competitive markets in c<strong>on</strong>veying informati<strong>on</strong>, most famously associated with Hayek (1945):<br />
“The marvel is that in a case like that of a scarcity of <strong>on</strong>e raw material, without an order<br />
being issued, without more than perhaps a h<strong>and</strong>ful of people knowing the cause, tens of<br />
thous<strong>and</strong>s of people whose identity could not be ascertained by m<strong>on</strong>ths of investigati<strong>on</strong>,<br />
are made to use the material or its products more sparingly; that is, they move in the right<br />
directi<strong>on</strong>.”<br />
In this example, the focus is <strong>on</strong> competiti<strong>on</strong> as a force for distilling informati<strong>on</strong> in an efficient<br />
manner.<br />
Thus the process of competiti<strong>on</strong> can be seen to bring benefits exactly because agents in the<br />
ec<strong>on</strong>omy do not have full informati<strong>on</strong> — for example, through “price signals” competiti<strong>on</strong> tackles<br />
the deficiencies <strong>and</strong> asymmetries of informati<strong>on</strong> across the ec<strong>on</strong>omy. Similarly, as discussed<br />
above, the process of competiti<strong>on</strong> can be seen to bring benefits exactly because agents in the<br />
ec<strong>on</strong>omy are not fully rati<strong>on</strong>al — for example, selecti<strong>on</strong> from experimentati<strong>on</strong> tackles the inability<br />
of ec<strong>on</strong>omic agents to make optimal use of all the resources <strong>and</strong> informati<strong>on</strong> they have at their<br />
disposal at a point in time.<br />
Yet the Darwinian reference to competitive selecti<strong>on</strong> does not quite mean the same in<br />
evoluti<strong>on</strong>ary ec<strong>on</strong>omics as in biology. Foster (2001, p120) argues that competiti<strong>on</strong> in ec<strong>on</strong>omics<br />
should allow for “inheritance of behavioral characteristics acquired from experience in particular<br />
envir<strong>on</strong>ments.” In other words, competiti<strong>on</strong> is not just about selecti<strong>on</strong> between firms (raising<br />
industry st<strong>and</strong>ards through eliminati<strong>on</strong> of weaker firms) but also the benefits of learning through<br />
creative endeavors, experience <strong>and</strong> also imitati<strong>on</strong> that allows individual firms to make progress.<br />
(In fact Foster sees this process as more reflective of the evoluti<strong>on</strong>ary theories of Lamarck than<br />
Darwin — a significant difference being that Lamarck c<strong>on</strong>sidered that organisms evolve because<br />
they are willing or need to evolve, rather than <strong>on</strong> the chance-orientated basis of Darwin).<br />
In particular, any competitive advantage <strong>on</strong>e firm has over another depends <strong>on</strong> the extent to<br />
which the factors underlying that advantage are imitable. While species may not be able to<br />
imitate other species (imitati<strong>on</strong> is fundamentally different from the process through species may<br />
adapt in a similar way to similar circumstances), firms are often able to imitate their competitors at<br />
low cost. Therefore an evoluti<strong>on</strong>ary approach to ec<strong>on</strong>omic should not be so Darwinian as to miss<br />
the importance of imitati<strong>on</strong> in ec<strong>on</strong>omic development. Furthermore, the role of imitati<strong>on</strong> in<br />
ec<strong>on</strong>omic development will be governed by the ability of firms to maintain their competitive<br />
advantage through patents, <strong>and</strong> secrets, as well as relative speeds of innovati<strong>on</strong> where timing<br />
matters to success. (How firms seek competitive advantage is discussed further in secti<strong>on</strong> 3<br />
within the c<strong>on</strong>text of the management <strong>and</strong> strategy literature.)<br />
44<br />
262
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Thus ec<strong>on</strong>omic development <strong>and</strong> innovati<strong>on</strong> can be seen as a combined effect of selecti<strong>on</strong> (via<br />
competiti<strong>on</strong>) from a variety of competing routines <strong>and</strong> practices as well as the more endogenous<br />
process of agents seeking improved routines <strong>and</strong> practices. The role of ‘dynamic’ competiti<strong>on</strong> in<br />
this regard is quite different from its “static” force in c<strong>on</strong>straining prices.<br />
This emphasis <strong>on</strong> competiti<strong>on</strong> as a selecti<strong>on</strong> process is supported by a quite separate str<strong>and</strong> of<br />
empirical literature that seeks to decompose industry-level productivity growth to examine the<br />
‘microstructure’ of this growth. Such studies use panel data of firms (or plants within multi-plant<br />
firms) to examine the extent to which ec<strong>on</strong>omic development at the industry level is driven by<br />
productivity improvements within firms (‘internal restructuring’) versus expansi<strong>on</strong> (<strong>and</strong> entry) of<br />
high-productivity firms <strong>and</strong> c<strong>on</strong>tracti<strong>on</strong> (<strong>and</strong> exit) of low-productivity firms (‘external<br />
restructuring’).<br />
For example, Disney et al. (2000) analyze a dataset which covers the period from 1980 to 1992<br />
<strong>and</strong> c<strong>on</strong>tains data for 140,000 UK manufacturing establishments a year. They examine the<br />
relati<strong>on</strong>ship between industry-level <strong>and</strong> firm-level dynamics. Of particular significance they find<br />
that external restructuring explains approximately 50 per cent of changes in labor productivity,<br />
<strong>and</strong> 90 per cent of changes in total factor productivity, over the period. These results indicate that<br />
in the industries studied, competiti<strong>on</strong> in terms of selecti<strong>on</strong> from variety was a very important driver<br />
of productivity growth.<br />
Note that such studies focus more <strong>on</strong> technological <strong>and</strong> organizati<strong>on</strong> improvements than drastic<br />
innovati<strong>on</strong>s (major new lines of products) that would not be fully captured in the productivity<br />
studies. But it is clear to see how evidence of internal versus external restructuring, in the<br />
c<strong>on</strong>text of industry-level ec<strong>on</strong>omic development fits in with the evoluti<strong>on</strong>ary c<strong>on</strong>cepts of<br />
competiti<strong>on</strong> as a selecti<strong>on</strong> process.<br />
Industry structure <strong>and</strong> innovati<strong>on</strong><br />
Schumpeter is often associated with the hypothesis that large firms with market power are more<br />
innovative than small <strong>on</strong>es, although Fagerberg (2002) argues that Schumpeter was more<br />
interested in the difference between new firms <strong>and</strong> old firms. Evoluti<strong>on</strong>ary ec<strong>on</strong>omics offers an<br />
interesting perspective <strong>on</strong> the relati<strong>on</strong>ship between firm size <strong>and</strong> innovati<strong>on</strong>.<br />
Fagerberg (2002) cites models by Nels<strong>on</strong> <strong>and</strong> Winter from the early 1980s that seek to explain<br />
why larger firms may be more innovative than smaller firms. On the <strong>on</strong>e h<strong>and</strong>, if firms use<br />
retained profits to finance R&D, large successful firms will have advantages over smaller firms.<br />
On the other h<strong>and</strong>, a large firm is able to derive greater benefit from finding a better “routine”<br />
because it can put it to use across a larger number of units of producti<strong>on</strong>. Such analysis would<br />
suggest that large firms are at a competitive advantage, although Nels<strong>on</strong> <strong>and</strong> Winter are reported<br />
to see this effect counteracted by large firms (with more market power) having a higher profit<br />
target (price/cost rati<strong>on</strong>) <strong>and</strong> this may provide some restraint in these dominating smaller firms.<br />
These models also suggest that if large firms are aggressive in pursuing their advantage over<br />
small firms, <strong>and</strong> there is in turn a tendency towards c<strong>on</strong>centrati<strong>on</strong>, large “imitating” firms may do<br />
well at the expense of smaller innovators. Thus where technological progress is endogenous,<br />
productivity may be hindered compared to a situati<strong>on</strong> of a more varied industry structure.<br />
45<br />
263
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
An alternative way of c<strong>on</strong>sidering whether large firms are more innovative than small firms is to<br />
focus more <strong>on</strong> heterogeneity in ability. To the extent that competiti<strong>on</strong> leads to the growth of<br />
innovative firms, we might simply expect causality to run the other way: it is not that large firms<br />
are inherently better innovators than smaller firms, just that the better innovators will be<br />
successful <strong>and</strong> hence grow in size. Theories based al<strong>on</strong>g these lines would therefore be treating<br />
innovati<strong>on</strong> as exogenous to firm size, with causality running from the former to the latter.<br />
This expositi<strong>on</strong> between exogenous <strong>and</strong> endogenous innovati<strong>on</strong> is valid more generally than firm<br />
size. Nels<strong>on</strong> <strong>and</strong> Winter (2002) draw a distincti<strong>on</strong> between industries that are characterized by<br />
“science-based” innovati<strong>on</strong> <strong>and</strong> industries that are characterized by “cumulative” innovati<strong>on</strong>.<br />
Science-based industries are c<strong>on</strong>sidered to be those where the thrust of innovati<strong>on</strong> comes from<br />
R&D activities outside that industry. For example, firms in such an industry might benefit<br />
principally from external scientific developments <strong>and</strong> innovative activities by their suppliers. By<br />
c<strong>on</strong>trast, in cumulative industries, innovati<strong>on</strong> stems from R&D activity within each of the firms in<br />
that industry.<br />
An important difference may be expected between science-based industries <strong>and</strong> cumulative<br />
industries. In the former, where firms reap the benefits of innovative activities from outside their<br />
own industry, new entrants <strong>and</strong> previous laggards may be expected to have similar scope for<br />
development than industry leaders. Just because a firm has performed poorly in the past, or is a<br />
newcomer to an industry, should not totally undermine its ability to take advantage of external<br />
innovati<strong>on</strong>s if these are exogenous.<br />
However, if the innovative activity within an industry comes predominantly from the efforts of firms<br />
within that industry, we may expect a significant degree of path dependency. Successful<br />
innovators in <strong>on</strong>e period will gain advantage over other firms that allows them greater potential for<br />
innovati<strong>on</strong> the next period. The cumulative effect of this is that less innovative firms fall further<br />
<strong>and</strong> further behind, <strong>and</strong> incumbents have some degree of advantage over new entrants.<br />
High-level policy implicati<strong>on</strong>s<br />
In discussing policy implicati<strong>on</strong>s, Fagerberg (2002) draws distincti<strong>on</strong> between evoluti<strong>on</strong>ary<br />
approaches <strong>and</strong> neo-classical approaches. Neo-classical ec<strong>on</strong>omics places emphasis <strong>on</strong> the<br />
failure of markets to deliver “public goods”, <strong>and</strong> makes the case for government support to<br />
provide the public good of knowledge. By c<strong>on</strong>trast, evoluti<strong>on</strong>ary ec<strong>on</strong>omics does not treat<br />
knowledge as a public good in this sense; for example tacit knowledge exists in routines <strong>and</strong><br />
practices rather than the c<strong>on</strong>sciousness <strong>and</strong> cannot be transferred from <strong>on</strong>e party to another in<br />
the way that a formula or design can. Adopting a perspective more in line with the evoluti<strong>on</strong>ary<br />
approach would therefore provide les support for government funding of knowledge as a public<br />
good.<br />
Furthermore, evoluti<strong>on</strong>ary ec<strong>on</strong>omics argues that variety is an important part of the scope for<br />
innovati<strong>on</strong> <strong>and</strong> development, while the role of heterogeneity per se is underplayed in neoclassical<br />
ec<strong>on</strong>omics. Policy proposals stem from this stance. For example:<br />
46<br />
264
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
!Rather than subsidizing R&D in well-established firms in traditi<strong>on</strong>al sectors, <strong>on</strong>e<br />
might put the resources into new types of activities or actors, not necessarily with the<br />
expectati<strong>on</strong> that these would do extremely well, but because the entire system<br />
(including thetraditi<strong>on</strong>al sectors) might benefit from such increased diversity”<br />
(Fagerberg, p 41).<br />
2.3 Systems of Innovati<strong>on</strong> <strong>and</strong> the Role of Instituti<strong>on</strong>s<br />
The emphasis <strong>on</strong> the noti<strong>on</strong> of “systems” of innovati<strong>on</strong> suggests that innovati<strong>on</strong> is not a<br />
phenomen<strong>on</strong> which results from the activities of isolated innovative firms, but the result of<br />
complex important interacti<strong>on</strong>s between firms <strong>and</strong> other organizati<strong>on</strong>s <strong>and</strong> instituti<strong>on</strong>s. Each<br />
innovative activity by a certain organizati<strong>on</strong> involves an element of reliance <strong>on</strong> external sources<br />
<strong>and</strong> innovati<strong>on</strong> in the ec<strong>on</strong>omy results from the complementary c<strong>on</strong>tributi<strong>on</strong>s of different<br />
organizati<strong>on</strong>s in a particular instituti<strong>on</strong>al envir<strong>on</strong>ment.<br />
This str<strong>and</strong> of the literature draws together many comm<strong>on</strong> themes between two related<br />
approaches — evoluti<strong>on</strong>ary ec<strong>on</strong>omics <strong>and</strong> cliometrics. What is cliometrics?<br />
Cliometrics was initiated by Alfred C<strong>on</strong>rad <strong>and</strong> John Meyer (C<strong>on</strong>rad <strong>and</strong> Meyer<br />
1957,1958) <strong>and</strong> has been popularized by Robert Fogel <strong>and</strong> Douglass North.<br />
Cliometrics is also often labelled as new ec<strong>on</strong>omic history or quantitative ec<strong>on</strong>omic<br />
history. It attempts to provide new insights into ec<strong>on</strong>omic history, mainly by<br />
c<strong>on</strong>sidering issues related to ec<strong>on</strong>omic development <strong>and</strong> growth.<br />
A more precise definiti<strong>on</strong> for cliometrics is given by McCloskey (1978): “a<br />
cliometrician is an ec<strong>on</strong>omist applying ec<strong>on</strong>omic theory (usually simple) to historical<br />
facts (not always quantitative) in the interest of history (not ec<strong>on</strong>omics)”. Despite<br />
substantial initial criticism, cliometrics has become a well-established branch of<br />
ec<strong>on</strong>omics <strong>and</strong> in 1993, the NobelPrize in Ec<strong>on</strong>omics was awarded to Robert Fogel<br />
<strong>and</strong> Douglass North.<br />
According to Claude Diebolt , cliometrics can be broken down into three major<br />
categories. First, descriptive cliometrics, which seeks to disentangle the ec<strong>on</strong>omic<br />
<strong>and</strong> social influence that might have exerted an influence <strong>on</strong> historical events <strong>and</strong><br />
processes.<br />
Sec<strong>on</strong>d, positive cliometrics, which is c<strong>on</strong>cerned with organizing data based <strong>on</strong><br />
which then different theoretical c<strong>on</strong>jectures can be tested. Third, normative<br />
cliometrics, which cannot be tested empirically <strong>and</strong> which deal with ethical questi<strong>on</strong>s<br />
such as efficiency,distributive <strong>and</strong> social justice <strong>and</strong> with how to ameliorate social<br />
welfare.<br />
Cliometrics is largely based <strong>on</strong> rec<strong>on</strong>structed l<strong>on</strong>g (secular) time series or large panel<br />
data related to historic events. The theoretical background underlying cliometrics was<br />
initially neo-classical theory, but it has been evolving c<strong>on</strong>tinuously with<br />
developments in<br />
general ec<strong>on</strong>omics over the past 40 years or so. Recent works in this field are<br />
essentially related to evoluti<strong>on</strong>ary ec<strong>on</strong>omics as a theoretical background.<br />
47<br />
265
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Defining ‘Innovati<strong>on</strong> Systems’<br />
Different definiti<strong>on</strong>s of the c<strong>on</strong>cept of innovati<strong>on</strong> systems have been offered in the literature, <strong>and</strong>,<br />
despite the inevitable differences, all point towards the importance of the noti<strong>on</strong> that inventi<strong>on</strong>,<br />
innovati<strong>on</strong> <strong>and</strong> its diffusi<strong>on</strong> result from interacti<strong>on</strong>s between different complementary<br />
organisati<strong>on</strong>s <strong>and</strong> instituti<strong>on</strong>s.<br />
Freeman (1987) defines a system of innovati<strong>on</strong> as:<br />
“[…] a network of instituti<strong>on</strong>s in the public <strong>and</strong> private sectors whose activities <strong>and</strong><br />
interacti<strong>on</strong>s initiate, import, modify <strong>and</strong> diffuse new technology”.<br />
Lundvall (1992) defines a system of innovati<strong>on</strong> as a system that includes all parts <strong>and</strong> aspects of<br />
the ec<strong>on</strong>omic structure <strong>and</strong> the instituti<strong>on</strong>al set-up affecting learning as well searching <strong>and</strong><br />
exploring. Alternatively, Gregers<strong>on</strong> <strong>and</strong> Johns<strong>on</strong> (1998, pp. 5) provides the following descripti<strong>on</strong><br />
for system of innovati<strong>on</strong>:<br />
“…overall innovati<strong>on</strong> performance of an ec<strong>on</strong>omy depends not <strong>on</strong>ly <strong>on</strong> how specific<br />
organizati<strong>on</strong>s like firms <strong>and</strong> research institutes perform, but also <strong>on</strong> how they interact<br />
with each other <strong>and</strong> with the government sector in knowledge producti<strong>on</strong> <strong>and</strong><br />
distributi<strong>on</strong>. Innovating firms operate within a comm<strong>on</strong> instituti<strong>on</strong>al set-up <strong>and</strong> they<br />
jointly depend <strong>on</strong>, c<strong>on</strong>tribute to <strong>and</strong> utilize a comm<strong>on</strong> knowledge infrastructure. It can<br />
be thought of as a system which creates <strong>and</strong> distributes knowledge, utilizes this<br />
knowledge by introducing itinto the ec<strong>on</strong>omy in the form of innovati<strong>on</strong>s, diffuses it<br />
<strong>and</strong> transforms it into something valuable, for example, internati<strong>on</strong>al competitiveness<br />
<strong>and</strong> ec<strong>on</strong>omic growth.”<br />
In general a system of innovati<strong>on</strong> is defined in terms of its comp<strong>on</strong>ents <strong>and</strong> the relati<strong>on</strong>s through<br />
which the comp<strong>on</strong>ents are linked with other another.<br />
The elements of the system of innovati<strong>on</strong> are organizati<strong>on</strong>s <strong>and</strong> instituti<strong>on</strong>s. Instituti<strong>on</strong>s exert a<br />
c<strong>on</strong>siderable influence <strong>on</strong> organizati<strong>on</strong>s, <strong>and</strong> simultaneously, organizati<strong>on</strong>s also impact <strong>on</strong> the<br />
instituti<strong>on</strong>al envir<strong>on</strong>ment. Instituti<strong>on</strong>s may give rise to new organizati<strong>on</strong>s as well as being the<br />
origin of new instituti<strong>on</strong>s.<br />
Edquist (2001) argues that organizati<strong>on</strong>s are formal structures that are created with a well defined<br />
goal in mind <strong>on</strong> the <strong>on</strong>e h<strong>and</strong>, <strong>and</strong> instituti<strong>on</strong>s such comm<strong>on</strong> habits, routines, established<br />
practices, rules or laws <strong>on</strong> the other. Organizati<strong>on</strong>s within the system can be broken down into<br />
two categories - primary <strong>and</strong> sec<strong>on</strong>dary. The activities listed above are associated with primary<br />
organizati<strong>on</strong>s whereas sec<strong>on</strong>dary organizati<strong>on</strong>s impact <strong>on</strong> how primary organizati<strong>on</strong>s behave<br />
when performing fundamental activities.<br />
Schoser (1999) argues that the instituti<strong>on</strong>al setting can be broken down into formal <strong>and</strong> informal<br />
instituti<strong>on</strong>s <strong>and</strong> that the factors identified by Gregers<strong>on</strong> <strong>and</strong> Johns<strong>on</strong> (1998) – instituti<strong>on</strong>al set-up,<br />
knowledge infrastructure, specializati<strong>on</strong> patterns, public <strong>and</strong> private dem<strong>and</strong> structure <strong>and</strong><br />
government policy – may impact <strong>on</strong> innovati<strong>on</strong> either directly or indirectly.<br />
Typical organizati<strong>on</strong>s in an innovati<strong>on</strong> system are private firms, universities, government<br />
laboratories <strong>and</strong> industrial research associati<strong>on</strong>s, am<strong>on</strong>g which a divisi<strong>on</strong> of labor is realized,<br />
48<br />
266
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
mainly coordinated by n<strong>on</strong>-market means. These various organizati<strong>on</strong>s undertake different<br />
activities within the system, depending <strong>on</strong> their particular objectives, research mechanisms <strong>and</strong><br />
incentive structures.<br />
The divisi<strong>on</strong> of labor within a system of innovati<strong>on</strong> explains why the systemic view holds that not<br />
<strong>on</strong>ly the comp<strong>on</strong>ents of a system of innovati<strong>on</strong> are important but crucially also their links <strong>and</strong><br />
interacti<strong>on</strong>s. Metcalfe (1995) observes that c<strong>on</strong>nectivity within a system is achieved by a variety<br />
of mechanisms as for instance:<br />
•<br />
•<br />
•<br />
mobility of employers in the labor market;<br />
grants <strong>and</strong> c<strong>on</strong>tracts for research; <strong>and</strong><br />
informal networks, e.g. the links between user firms <strong>and</strong> their suppliers<br />
How ‘Innovati<strong>on</strong> Systems’ Work<br />
There are various activities undertaken within a system. Liu <strong>and</strong> White (2001), identify five<br />
activities: research; implementati<strong>on</strong>, i.e. manufacturing; end-use; educati<strong>on</strong> <strong>and</strong> linkage. Bergek<br />
<strong>and</strong> Jacobss<strong>on</strong> (2002) put forward the following functi<strong>on</strong>s of a system of innovati<strong>on</strong>: creati<strong>on</strong> of<br />
new knowledge; guidance of the research process; supply of resources; creati<strong>on</strong> of positive<br />
externalities <strong>and</strong> creati<strong>on</strong> of markets.<br />
Pearce (2002) offer a different approach to defining the activities of a system of innovati<strong>on</strong>.<br />
Pearce identifies basic research as the first element <strong>on</strong> which applied research is based. Product<br />
development including market research is then built <strong>on</strong> both basic <strong>and</strong> applied research. The<br />
final stage of the system of innovati<strong>on</strong> is adaptati<strong>on</strong> <strong>and</strong> marketing of the products. The first two<br />
stages depend heavily <strong>on</strong> research instituti<strong>on</strong>s, government support as well as firms. By c<strong>on</strong>trast,<br />
the last two stages mainly c<strong>on</strong>cern private firms.<br />
N<strong>on</strong>etheless, Nels<strong>on</strong> (1996) notes that the noti<strong>on</strong> of systems of innovati<strong>on</strong> does not imply that the<br />
system was c<strong>on</strong>sciously designed or that interacti<strong>on</strong>s between the comp<strong>on</strong>ents of the system<br />
work are smoothly <strong>and</strong> that the system is coherent as a whole.<br />
Systems of innovati<strong>on</strong> are often c<strong>on</strong>sidered to be nati<strong>on</strong>al, although it is debated whether nati<strong>on</strong>al<br />
boundaries can be deemed relevant.<br />
On the <strong>on</strong>e h<strong>and</strong> it is argued that the element of nati<strong>on</strong>ality may derive from different factors such<br />
as the nati<strong>on</strong>al focus of technological <strong>and</strong> other policies, laws <strong>and</strong> regulati<strong>on</strong>s that have an impact<br />
<strong>on</strong> the innovative envir<strong>on</strong>ment (Metcalfe, 1995). Nels<strong>on</strong> (1996) argues that because the<br />
educati<strong>on</strong> <strong>and</strong> research system, public infrastructure, laws <strong>and</strong> financial instituti<strong>on</strong>s keep some of<br />
their nati<strong>on</strong>al characteristics, differences across different nati<strong>on</strong> innovati<strong>on</strong> systems are likely to<br />
persist even in the l<strong>on</strong>ger run.<br />
On the other h<strong>and</strong>, the boundaries of systems of innovati<strong>on</strong> in different countries are getting<br />
blurred with the rise of internati<strong>on</strong>al firms present in several countries <strong>and</strong> increasingly intensifying<br />
cross-country inter-firm c<strong>on</strong>necti<strong>on</strong>s to share knowledge <strong>and</strong> innovate.<br />
49<br />
267
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
It may also be argued that systems of innovati<strong>on</strong> may be local <strong>and</strong> regi<strong>on</strong>al rather than nati<strong>on</strong>al<br />
<strong>and</strong> that <strong>on</strong>e may think of different systems of innovati<strong>on</strong>s at various levels of aggregati<strong>on</strong>.<br />
It is widely recognized in the literature that there is a dilemma in nati<strong>on</strong>al systems of innovati<strong>on</strong>.<br />
This dilemma is whether to create incentives for innovati<strong>on</strong> or whether to foster diffusi<strong>on</strong> of<br />
innovati<strong>on</strong>. At the heart of this problem is the patent system that encourages innovati<strong>on</strong> by<br />
allowing for reaping the benefits <strong>and</strong> which at the same time hinders diffusi<strong>on</strong>.<br />
The role patents play in innovati<strong>on</strong> is c<strong>on</strong>troversial. For instance, Levin et al (1987) report a fairly<br />
limited role of patents when it comes to protecting new technology. A brief discussi<strong>on</strong> of some of<br />
the findings of the literature review regarding patents is presented in the box below.<br />
Mokyr (1992, 1999a) also suggests that some instituti<strong>on</strong>al arrangements may hinder innovati<strong>on</strong>.<br />
He points out that self-organizing systems become the cornerst<strong>on</strong>e of evoluti<strong>on</strong>ary ec<strong>on</strong>omics<br />
<strong>and</strong> notes that:<br />
“…the most interesting property of these systems is that they resist change. Resistance<br />
to change is essential for any system if it is to functi<strong>on</strong> <strong>and</strong> not degenerate into chaos.”<br />
However, he also emphases that resistance in society might have not <strong>on</strong>ly a stabilizing effect but<br />
can also act as a break <strong>on</strong> innovati<strong>on</strong>. Indeed, high resistance can slow down or even eliminate<br />
the introducti<strong>on</strong> of innovati<strong>on</strong>s or even deter innovati<strong>on</strong> activity.<br />
Oppositi<strong>on</strong> against new innovati<strong>on</strong>s might come from both owners <strong>and</strong> employees. Mokyr<br />
(1999a) enumerates the following sources of resistance:<br />
•<br />
•<br />
•<br />
•<br />
ec<strong>on</strong>omically motivated resistance;<br />
ideologically motivated resistance;<br />
systemic resistance; <strong>and</strong><br />
frequency resistance.<br />
However, the success of an innovati<strong>on</strong> hinges <strong>on</strong> two factors. On the <strong>on</strong>e h<strong>and</strong>, the greater the<br />
value of the equipment <strong>and</strong> the more specific the skill that are threatened by a new innovati<strong>on</strong>,<br />
the higher the incentive to resist. On the other h<strong>and</strong>, in the event that the innovati<strong>on</strong> generates<br />
high social gains, the incentive to establish the innovati<strong>on</strong> will be equally high.<br />
Sec<strong>on</strong>d, the introducti<strong>on</strong> of the new technology requires time for people to become familiar with it<br />
<strong>and</strong> to make best use of it. Atkens<strong>on</strong> <strong>and</strong> Kehoe (2002) build a quantitative model of technology<br />
diffusi<strong>on</strong> in order to simulate this slow transiti<strong>on</strong> from a low towards a high-productivity regime.<br />
The model seems to fit actual data rather well for the late-19th century US manufacturing sector.<br />
One important feature of the model is that the new technology can be adopted <strong>on</strong>ly by building<br />
new plants. This assumpti<strong>on</strong> is c<strong>on</strong>sistent with empirical literature <strong>on</strong> the microstructure of<br />
productivity growth that finds that multi-plant firms often achieve productivity gains by closing<br />
50<br />
268
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
down inefficient plants rather than improving efficiency within those plants (e.g. Disney et al,<br />
2000).<br />
For the model to produce the slow transiti<strong>on</strong>, another assumpti<strong>on</strong>s are needed. Manufacturers<br />
using the old technology dispose of a slowly accumulated stock of knowledge. At the outset of<br />
the transiti<strong>on</strong> (from old to new technology), they are hesitant to relinquish this stock of knowledge<br />
in favour of the new technology. The higher the stock of knowledge compared to the <strong>on</strong>e related<br />
to the new technology, the slower they are to go for the new technology.<br />
51<br />
269
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
The role of patents<br />
A number of papers attempt to analyze the history of patents <strong>and</strong> the patent system in<br />
theUS. For instance, Sokoloff (1988) examines 4,500 patents accepted in the US<br />
during the period 1790-1846 <strong>and</strong> sheds light <strong>on</strong> the fact that the number of patents<br />
changed at an uneven pace over time. Apart from an important rise of roughly 300<br />
per cent in patents that can be observed during 1805-10 <strong>and</strong> 1820-36, the number of<br />
patents recorded stayed fairly c<strong>on</strong>stant during the rest of the period.<br />
Using a different set of US patents, Phillips (1992) is c<strong>on</strong>cerned with the period<br />
spanning from 1831 to 1879 <strong>and</strong> c<strong>on</strong>firms the finding of Sokoloff (1988). Accepted<br />
patents rose sharply over the periods 1950-60 <strong>and</strong> 1962-68, but remained relatively<br />
stable in other years. If the number of patents is viewed as a good proxy for<br />
innovative activity, it could be argued that innovati<strong>on</strong> follows a stochastic path.<br />
Sokoloff (1988) suggested that the surge in the number of patents occurred between<br />
1820 <strong>and</strong> 1936 was due to major changes in the patent system that enforced<br />
incentives for innovative activities. In fact, patent theory tells us that the higher the<br />
expected return to an innovati<strong>on</strong>, the higher the incentive to innovate.<br />
Khan (1995) puts forward that the US patent system, which was introduced in 1790<br />
,<strong>and</strong> especially the pro-innovati<strong>on</strong> attitude of judges when accepting patent dem<strong>and</strong>s<br />
fostered substantially innovative activity in 19 century US.<br />
Phillips (1992) emphasizes the dem<strong>and</strong>-driven nature of patent activity. He argues<br />
that not <strong>on</strong>ly accumulati<strong>on</strong> of human capital, but also the increasing dem<strong>and</strong> for<br />
patentable ideas <strong>and</strong> new technology in the c<strong>on</strong>text of increasing industrializati<strong>on</strong>,<br />
which lead to an increase in accepted patents.<br />
However, some others have argued that patents do not truly reflect innovati<strong>on</strong>s.<br />
MacLeod (1986) points out that patentees are likely to react sensitively to the<br />
ec<strong>on</strong>omic <strong>and</strong> financial c<strong>on</strong>juncture. She sets out to show that the increase in<br />
registered patents between 1691-1693 was influenced by the increased availability of<br />
risk capital. It remains an unanswered questi<strong>on</strong>, however, whether innovati<strong>on</strong>s were<br />
also that dependent <strong>on</strong> the ec<strong>on</strong>omic envir<strong>on</strong>ment. MacLeod also notes that the<br />
<strong>on</strong>going war gave a significant impetus to military innovati<strong>on</strong>. Another important<br />
questi<strong>on</strong> here is the degree of development of the patent system. It is argued that<br />
weaknesses in the system probably exerted a deterrent or at least not c<strong>on</strong>ductive for<br />
patenting new innovati<strong>on</strong>.<br />
MacLeod (1992) investigates the 19 century British mechanical engineering industry<br />
<strong>and</strong> argues that the existence of a patent system was <strong>on</strong>e key to innovati<strong>on</strong>. The<br />
system provided certainty for innovators so that they could reap the benefits of their<br />
innovati<strong>on</strong>. Therefore, the uncertainty prevailing before 1830 due to the unreformed<br />
<strong>and</strong> inappropriate patent system encouraged secrecy, which in turn worked against<br />
the diffusi<strong>on</strong> of new technology.<br />
52<br />
270
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Do Different “Systems” Matter?<br />
There are differences of opini<strong>on</strong> as to whether differences between “Innovati<strong>on</strong> Systems “ matter<br />
in terms of promoting innovati<strong>on</strong> <strong>and</strong> technical change.<br />
One factor that can promote technology diffusi<strong>on</strong> is the presence of multinati<strong>on</strong>al companies<br />
worldwide. However, Saviotti (2000) documents substantial differences that persist between<br />
nati<strong>on</strong>al systems of innovati<strong>on</strong> of different countries. He observes that different instituti<strong>on</strong>al <strong>and</strong><br />
organisati<strong>on</strong>al c<strong>on</strong>figurati<strong>on</strong>s result in different specializati<strong>on</strong> patterns across countries (see also<br />
secti<strong>on</strong> 1.4 <strong>on</strong> new ec<strong>on</strong>omic geography.)<br />
Nels<strong>on</strong> (1996) also sets out to analyze the cause of differences in different nati<strong>on</strong>al systems of<br />
innovati<strong>on</strong>. Innovati<strong>on</strong> systems are analyzed in three groups of countries, namely large highincome<br />
countries such as the US, Germany, Japan, small high-income countries such as Sweden<br />
<strong>and</strong> developing, low-income countries, e.g. South Korea <strong>and</strong> Argentina.<br />
Based <strong>on</strong> this categorizati<strong>on</strong> of countries, it is argued that the size <strong>and</strong> the affluence of a<br />
particular country mainly determine its capability to innovate. In additi<strong>on</strong>, ec<strong>on</strong>omic <strong>and</strong> political<br />
circumstances <strong>and</strong> priorities also have a significant role to play in shaping the innovati<strong>on</strong> system.<br />
Further to this, Nels<strong>on</strong> (1996) emphasis the importance of competiti<strong>on</strong>. Str<strong>on</strong>g competiti<strong>on</strong> is<br />
believed to promote innovati<strong>on</strong>s. However, he argues that domestic markets especially in small<br />
countries fail to play this role <strong>and</strong> therefore internati<strong>on</strong>al competiti<strong>on</strong> encountered in world<br />
markets, that is <strong>on</strong> export <strong>and</strong> import markets, is crucial in fostering innovative activity.<br />
A str<strong>on</strong>g <strong>and</strong> competitive export sector is thought to be the engine of ec<strong>on</strong>omic growth in many<br />
ec<strong>on</strong>omies. However, for the export sector to withhold internati<strong>on</strong>al competiti<strong>on</strong>, the nati<strong>on</strong>al<br />
system of innovati<strong>on</strong> should be shaped to increase innovati<strong>on</strong> <strong>and</strong> thus support exports.<br />
N<strong>on</strong>etheless, competitiveness can be viewed <strong>and</strong> defined differently in high-income <strong>and</strong> lowincome<br />
countries. High wage level in developed countries implies that <strong>on</strong>ly new, improved<br />
products <strong>and</strong> the c<strong>on</strong>stant ameliorati<strong>on</strong> of the underlying technological process offer possibilities<br />
to boost competitiveness. By c<strong>on</strong>trast, in less developed countries with lower wage levels,<br />
competitiveness is to be improved by diffusing <strong>and</strong> learning foreign technology.<br />
Security policy c<strong>on</strong>siderati<strong>on</strong>s also seem to forge the shape of the innovati<strong>on</strong> system. Nels<strong>on</strong><br />
(1996) argues that the bulk of government backed research <strong>and</strong> development is c<strong>on</strong>centrating <strong>on</strong><br />
the defense industry in the US <strong>and</strong> Britain. Military inventi<strong>on</strong>s are then applied to civil problems<br />
giving birth to new products.<br />
Furthermore, the current German <strong>and</strong> Japanese systems of innovati<strong>on</strong> deeply root in the past<br />
when the system was designed to back the producti<strong>on</strong> of better weap<strong>on</strong>ry. However, after World<br />
War II, those innovative capacities were turned toward civilian producti<strong>on</strong>, <strong>and</strong> this led to<br />
flourishing exports.<br />
This is a clear sign for the str<strong>on</strong>g c<strong>on</strong>tinuity in nati<strong>on</strong>al innovati<strong>on</strong> systems over time. Nels<strong>on</strong><br />
(1996) points out that the German <strong>and</strong> French innovati<strong>on</strong> systems in 1990 were very similar to<br />
53<br />
271
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
those in 1890. This phenomen<strong>on</strong> seems to hold true for the majority of developed countries<br />
except for the US where a major change occurred after 1945.<br />
Grupp et al. (2001) set out to investigate the system of innovati<strong>on</strong> in Germany throughout the past<br />
150 years <strong>and</strong> come to the c<strong>on</strong>clusi<strong>on</strong> that the innovati<strong>on</strong> system has been very stable over the<br />
period under investigati<strong>on</strong>. According to the authors the industrial research system can be<br />
associated with an increase in innovative activity. When analyzing how R&D expenditures <strong>and</strong><br />
innovati<strong>on</strong> activity have changed from 1850 to 1999, they found that the German industrial<br />
research system has been very stable. In particular, they provide some more in depth<br />
ec<strong>on</strong>ometric analysis, which reveals that during 1850-1913, patent activity can be predicted by<br />
growth in human capital. Furthermore, there is causal link from patent activity to st<strong>and</strong>ard of<br />
livings whereas st<strong>and</strong>ard of living influenced human capital.<br />
These results clearly indicate the presence of a virtuous circle that results in substantial ec<strong>on</strong>omic<br />
growth. Results for the sec<strong>on</strong>d period (1951-99) are less straightforward: the st<strong>and</strong>ard of living<br />
variable appears exogenous <strong>and</strong> turns out to explain patent activity as well as <strong>on</strong> public <strong>and</strong><br />
private expenditures in R&D.<br />
The educati<strong>on</strong> system <strong>and</strong> especially universities have a significant role to play in the formati<strong>on</strong><br />
<strong>and</strong> the successful functi<strong>on</strong>ing of the innovati<strong>on</strong> system. One important questi<strong>on</strong> to be addressed<br />
is how <strong>and</strong> to what extent the research <strong>and</strong> teaching orientati<strong>on</strong> of universities satisfies the need<br />
for technological innovati<strong>on</strong>.<br />
In fact, the design of a good interface between public research organizati<strong>on</strong>s <strong>and</strong> private firms<br />
has been central to innovati<strong>on</strong> policy in most advanced Countries. This interface is c<strong>on</strong>sidered<br />
important both because it may allow research c<strong>on</strong>ducted in public organizati<strong>on</strong>s to be more easily<br />
translated into ec<strong>on</strong>omic growth <strong>and</strong> competitiveness <strong>and</strong> because it may result in fertile<br />
combinati<strong>on</strong> of complementary knowledge-basis <strong>and</strong> skills. Countering these benefits, the<br />
c<strong>on</strong>cern that closer linkages between public <strong>and</strong> private research organizati<strong>on</strong>s, <strong>and</strong> an increased<br />
privatizati<strong>on</strong> of university research, may diminish the variety of reward systems in the ec<strong>on</strong>omy<br />
<strong>and</strong> in the l<strong>on</strong>g-term the potential of public research organizati<strong>on</strong>s to c<strong>on</strong>tribute to the innovati<strong>on</strong><br />
system.<br />
Evidence from Studies using Cliometric techniques<br />
Cliometrics also offers some observati<strong>on</strong>s <strong>on</strong> the role of the relati<strong>on</strong>ship between innovati<strong>on</strong> <strong>and</strong><br />
instituti<strong>on</strong>s.<br />
Neo-classical theory posits that ec<strong>on</strong>omic agents maximise aggregate income in the absence of<br />
transacti<strong>on</strong> costs. However, since Coase (1960), it is clear that transacti<strong>on</strong> costs do matter. For<br />
instance, Wallis <strong>and</strong> North (1986) show that in 1970, 45 per cent of US GNP was linked to the<br />
transacti<strong>on</strong> sector. As a corollary, the instituti<strong>on</strong> setting, which, to a large extent, determines the<br />
magnitude of transacti<strong>on</strong> costs, has a c<strong>on</strong>siderable impact <strong>on</strong> the final outcome.<br />
What matters in the l<strong>on</strong>g- run is the c<strong>on</strong>tinuous interacti<strong>on</strong> between instituti<strong>on</strong>s <strong>and</strong> organizati<strong>on</strong>s<br />
that exerts an influence <strong>on</strong> developments in the instituti<strong>on</strong>al envir<strong>on</strong>ment. Organizati<strong>on</strong>s are born<br />
<strong>and</strong> are operating within the given framework of the instituti<strong>on</strong>s <strong>and</strong> their activities reflect the<br />
54<br />
272
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
opportunities provided by instituti<strong>on</strong>s. If, for example, security is lax, piracy will come into<br />
existence, whereas if, say, property rights are properly enforced, actors may engage in more<br />
innovative activities. (This approach is indeed very close to the systems of innovati<strong>on</strong>)<br />
Ols<strong>on</strong> (1996) puts forward that instituti<strong>on</strong>s are essential for efficiently functi<strong>on</strong>ing markets.<br />
N<strong>on</strong>etheless, it does not necessarily hold true that instituti<strong>on</strong>s provide per se the c<strong>on</strong>diti<strong>on</strong>s<br />
required for efficiency in markets <strong>and</strong> therefore they are subject to changes. If the rules of the<br />
game are perceived by the organizati<strong>on</strong>s as too restrictive, they may attempt to modify the<br />
instituti<strong>on</strong>al setting. The driving force behind changes is, according to North, learning by<br />
individuals. And it is argued that competiti<strong>on</strong> c<strong>on</strong>tributes significantly to the learning process<br />
given that learning is indispensable for survival. That is, the greater the degree of competiti<strong>on</strong>,<br />
the bigger the incentive for learning.<br />
In a series of seminal studies, Douglass North (1990,1991) underscores the importance of social<br />
<strong>and</strong> ec<strong>on</strong>omic instituti<strong>on</strong>s, which are likely to affect innovati<strong>on</strong> <strong>and</strong> hence l<strong>on</strong>g-term ec<strong>on</strong>omic<br />
development. According to North, instituti<strong>on</strong>s form the incentive structure of a society, <strong>and</strong> as a<br />
c<strong>on</strong>sequence, the political <strong>and</strong> ec<strong>on</strong>omic instituti<strong>on</strong>s are the underlying determinants of ec<strong>on</strong>omic<br />
performance.<br />
The instituti<strong>on</strong>al envir<strong>on</strong>ment provides a general framework for social <strong>and</strong> ec<strong>on</strong>omic interacti<strong>on</strong>s.<br />
Instituti<strong>on</strong>s fall into two categories. The first category can be referred to as formal c<strong>on</strong>straints <strong>and</strong><br />
is composed of rules, laws, c<strong>on</strong>stituti<strong>on</strong>s, property rights, whereas the sec<strong>on</strong>d <strong>on</strong>e is labeled as<br />
informal c<strong>on</strong>straints <strong>and</strong> is made up of customs, traditi<strong>on</strong>s, behavior, c<strong>on</strong>venti<strong>on</strong>s, codes of<br />
c<strong>on</strong>ducts etc.<br />
In North (1994), the evoluti<strong>on</strong> of instituti<strong>on</strong>s is analyzed in a historical perspective. It is argued that<br />
there are some primitive instituti<strong>on</strong>al settings that are most unlikely to experience changes<br />
triggered by pressures coming from the inside. In three types of exchanges under c<strong>on</strong>siderati<strong>on</strong>,<br />
namely tribal society, regi<strong>on</strong>al ec<strong>on</strong>omy <strong>and</strong> l<strong>on</strong>g-distance caravan trade, learning, <strong>and</strong> hence the<br />
accumulati<strong>on</strong> of knowledge <strong>and</strong> skill will not lead to changes, simply because innovati<strong>on</strong> is<br />
perceived as something endangering the system’s survival. This is also borne out by Posner<br />
(1980).<br />
North (1968) also explores the impacts that changes in the instituti<strong>on</strong>al envir<strong>on</strong>ment have <strong>on</strong><br />
technological progress. He analyses the case of ocean transportati<strong>on</strong> <strong>and</strong> finds that from 1600 to<br />
1860, the costs of ocean shipping dropped by 50 per cent due to huge increases in total factor<br />
productivity. The questi<strong>on</strong> is whether productivity gains are driven by technological advances or<br />
by other factors.<br />
North argues that in the period 1600-1784 productivity advances were slow, <strong>and</strong> virtually all the<br />
productivity growth came from decreased crew size <strong>and</strong> the better use of shipping time, that is the<br />
reducti<strong>on</strong> of time spent in harbors. This productivity advance may be understood within the<br />
evoluti<strong>on</strong>ary ec<strong>on</strong>omics framework as an improvement in organizati<strong>on</strong>al routines <strong>and</strong> working<br />
practices.<br />
55<br />
273
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Between 1814 <strong>and</strong> 1860, North finds that total factor productivity experienced a c<strong>on</strong>siderable<br />
accelerati<strong>on</strong> compared with the previous period, <strong>and</strong> grew roughly 10 times faster than before.<br />
North argues that the bulk of these productivity gains can be attributed to larger ships <strong>and</strong> to an<br />
increased load factor.<br />
Essentially, an increase in the size of the market <strong>and</strong> a sharp reducti<strong>on</strong> in piracy explain these<br />
changes in the characteristics of ships. On <strong>on</strong>e h<strong>and</strong>, shipping capacity was significantly better<br />
used because of immigrants pouring in America: Ships that used to return to America without<br />
carrying anything but the ballast were transporting men. On the other h<strong>and</strong>, security became<br />
better because of the eliminati<strong>on</strong> of piracy. This led to a reducti<strong>on</strong> in the need for military pers<strong>on</strong>al<br />
<strong>and</strong> equipment <strong>on</strong> board. All in all, North emphasizes that productivity gains were not due to<br />
technological progress but rather the c<strong>on</strong>sequence of organizati<strong>on</strong>al innovati<strong>on</strong>s <strong>and</strong> changes in<br />
the instituti<strong>on</strong>al envir<strong>on</strong>ment.<br />
With regards to the first industrial revoluti<strong>on</strong> in 18th <strong>and</strong> 19th century Britain, L<strong>and</strong>es (1994) argues<br />
that it was not pure accident that the first industrial revoluti<strong>on</strong> occurred in Britain <strong>and</strong> not<br />
elsewhere, e.g. in France. This is because of the structure of the industry <strong>and</strong> the instituti<strong>on</strong>al<br />
setting, which made it more likely that major inventi<strong>on</strong>s would be produced in Britain. However,<br />
Crafts (1995a,b) disagrees with this view <strong>and</strong> suggests that major inventi<strong>on</strong>s follow a stochastic<br />
process <strong>and</strong> thus those inventi<strong>on</strong>s could have occurred with roughly the same probability in<br />
France than in Britain given that the things in comm<strong>on</strong> between the two countries outweighs the<br />
differences between them.<br />
According to Mokyr (1993), two types of inventi<strong>on</strong>s can be distinguished. The first is<br />
“macro inventi<strong>on</strong>” <strong>and</strong> describes inventi<strong>on</strong>s of paramount importance for the ec<strong>on</strong>omy as a whole.<br />
The sec<strong>on</strong>d is “micro inventi<strong>on</strong>” <strong>and</strong> c<strong>on</strong>sists of small steps by which already extant techniques<br />
are ameliorated. Mokyr (1993) points out that macro inventi<strong>on</strong>s “… do not seem to obey obvious<br />
laws, do not necessarily resp<strong>on</strong>d to incentives, <strong>and</strong> defy most attempts to relate them to<br />
exogenous ec<strong>on</strong>omic variables. Many of them resulted from strokes of genius, luck or<br />
serendipity.” Mokyr notes that Britain was by no means in a better positi<strong>on</strong> in the realm of<br />
macro inventi<strong>on</strong>s compared with France but the instituti<strong>on</strong>al <strong>and</strong> ec<strong>on</strong>omic envir<strong>on</strong>ment were<br />
more c<strong>on</strong>ducive as regards micro inventi<strong>on</strong>s.<br />
51<br />
56<br />
274
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
3 STRATEGIES OF INNOVATIVE FIRMS<br />
This secti<strong>on</strong> sets out some of the strategies applied by companies to enter new markets <strong>and</strong><br />
introduce new products. This is addressed by looking at the strategies employed to gain <strong>and</strong><br />
protect such an advantage.<br />
The secti<strong>on</strong> then looks how the optimal timing of innovati<strong>on</strong> can be used as a strategy <strong>and</strong><br />
includes a discussi<strong>on</strong> of life cycles more generally. Finally, the secti<strong>on</strong> c<strong>on</strong>siders competiti<strong>on</strong> in<br />
network industries.<br />
Innovati<strong>on</strong> <strong>and</strong> Competitive Advantage<br />
In c<strong>on</strong>sidering the strategies employed by innovative firms, it is useful to draw a distincti<strong>on</strong><br />
between two broad categories:<br />
•<br />
•<br />
how a firm gains a competitive advantage over rivals; <strong>and</strong><br />
how a firm can protect <strong>and</strong> maintain that competitive advantage.<br />
In the following secti<strong>on</strong>s, we discuss further these strategies <strong>and</strong> their relati<strong>on</strong>ship with<br />
innovati<strong>on</strong>,although it is necessary to be aware that the distincti<strong>on</strong> between these comp<strong>on</strong>ents is not<br />
clear-cut.<br />
Value creati<strong>on</strong>: how firms gain a competitive advantage<br />
The first strategy focuses <strong>on</strong> how firms can create more value, which can be achieved in principle<br />
through:<br />
•<br />
•<br />
•<br />
cost reducti<strong>on</strong> underpinned by efficiency improvements;<br />
increasing quality <strong>and</strong> value of existing products <strong>and</strong> services via differentiati<strong>on</strong> relative to<br />
competitors; <strong>and</strong><br />
developing new products <strong>and</strong> services.<br />
At the centre of value creati<strong>on</strong> is the noti<strong>on</strong> of competitive advantage popularized by Porter<br />
(1985). Porter’s view is this:<br />
“Competitive advantage is at the heart of the firm’s performance in competitive<br />
markets.After several decades of vigorous expansi<strong>on</strong> <strong>and</strong> prosperity, however, many<br />
firms lost sight of competitive advantage in their scramble for growth <strong>and</strong> pursuit of<br />
diversificati<strong>on</strong>. Today the importance of competitive advantage could hardly be<br />
greater. Firms throughout the world face slower growth as well as domestic <strong>and</strong> global<br />
competitors that are no l<strong>on</strong>ger acting as if the exp<strong>and</strong>ing pie were big enough for all.”<br />
In accordance with Porter (1985), competiti<strong>on</strong> is a c<strong>on</strong>tinuous quest for competitive advantage.<br />
57<br />
275
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
External <strong>and</strong> internal sources of competitive advantage can be identified. External sources such<br />
as changing c<strong>on</strong>sumer patterns or changes in technology are exogenous to firms. Since firms<br />
are heterogeneous in their resources <strong>and</strong> capabilities, they are not equally able to adapt to such<br />
changes, <strong>and</strong> thus some firms are able to realize competitive advantage through more timely <strong>and</strong><br />
efficient resp<strong>on</strong>ses to exogenous factors.<br />
The internal source of competitive advantage is essentially the capability to innovate, by which we<br />
mean inventive effort, rather than adaptive or imitative behavior.<br />
Cost leadership<br />
As highlighted above, value to customers can be created by means of decreasing costs, through<br />
which a firm is able to reduce prices relative to competitors.<br />
At the extreme, focus <strong>on</strong> this strategy can be described as thriving for cost leadership, which<br />
refers to the situati<strong>on</strong> when a firm has the capability to produce similar products at significantly<br />
lower cost than its competitors. There are several routes through which a firm can achieve cost<br />
advantage, for example:<br />
•<br />
•<br />
Ec<strong>on</strong>omies of scale <strong>and</strong> scope yield benefits to the extent that an increase in producti<strong>on</strong><br />
volume will entail a reducti<strong>on</strong> in unit cost.<br />
Ec<strong>on</strong>omies of learning describe experience-based learning. The more complex the<br />
process technology or the product, especially in terms of tacit knowledge (i.e. that which<br />
cannot be codified <strong>and</strong> transferred) the larger the scope for benefits from learning-bydoing.<br />
This provides means for established companies to experience cost advantage<br />
over new firms.<br />
Enhancing process technology <strong>and</strong> process design allows for greater productive<br />
efficiency <strong>and</strong> lower producti<strong>on</strong> costs for the same output.<br />
Similarly, enhancing the organizati<strong>on</strong>al efficiency <strong>and</strong> routines (for a given technology) will<br />
increase productive efficiency.<br />
Reducing input costs allows for lower unit costs for the same level of productivity.<br />
Because of the use of different suppliers, geographical differences in input prices <strong>and</strong><br />
different bargaining power, different firms may pay a different price for the same set of<br />
inputs.<br />
Augmenting capacity utilizati<strong>on</strong>: In the presence of high fixed costs, unit costs are to be<br />
decreased by increasing capacity utilizati<strong>on</strong>. On the other h<strong>and</strong>, overcapacity resulting in<br />
overtime pay, premiums for night shifts <strong>and</strong> rising maintenance costs also increase unit<br />
costs.<br />
58<br />
276
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Product quality <strong>and</strong> differentiati<strong>on</strong><br />
Changes to product quality (in particular quality-price trade-offs) <strong>and</strong> product differentiati<strong>on</strong><br />
compared to rival products represent the sec<strong>on</strong>d facet of value creati<strong>on</strong>.<br />
A firm can set out to differentiate its product according to the tangible <strong>and</strong> intangible aspects of<br />
the products it manufactures. The former c<strong>on</strong>cerns physical characteristics <strong>and</strong> performance of<br />
the product, while the latter relates to perceived, mainly social <strong>and</strong> psychological faculties.<br />
The nature of competiti<strong>on</strong> in a market will be associated with whether firms strive to gain<br />
competitive advantage through innovati<strong>on</strong> in physical attributes of the products they supply, or the<br />
intangible characteristics. For example, where an industry is mature <strong>and</strong> there appears little<br />
scope for changes in product innovati<strong>on</strong> <strong>and</strong> cost leadership, competitive advantage may be<br />
achieved through a successful br<strong>and</strong>ing exercise that provides differentiati<strong>on</strong> relative to rivals.<br />
Product innovati<strong>on</strong><br />
While changes to the quality of a product represent some degree of product innovati<strong>on</strong>,<br />
innovative behavior is likely to be greater with regard to the introducti<strong>on</strong> of new products <strong>and</strong><br />
services that have features to mark them as distinct from current products.<br />
A strategy of product innovati<strong>on</strong> may have the effect of creating new markets (in competiti<strong>on</strong><br />
policy terms, the introducti<strong>on</strong> or products for which there are no good substitutes <strong>on</strong> the dem<strong>and</strong>side)<br />
but may also, though not necessarily, destroy old markets (as c<strong>on</strong>sumers cease<br />
c<strong>on</strong>sumpti<strong>on</strong> of previous generati<strong>on</strong> of products in favor of the new products).<br />
When it comes to assessing how to exploit an innovati<strong>on</strong> to the maximum, that is to maximize<br />
profits related to the innovati<strong>on</strong>, firms can choose from several opti<strong>on</strong>s depending <strong>on</strong><br />
•<br />
•<br />
How much risk can or are they willing to take, <strong>and</strong> thus how high a return do they expect;<br />
How many resources can or are they willing to put at disposal when exploiting the<br />
innovati<strong>on</strong>.<br />
Risks are high in emerging, innovative industries. First, firms have to face technological<br />
uncertainties (such as the directi<strong>on</strong> technology develops in <strong>and</strong> which technical st<strong>and</strong>ards are<br />
established in the industry). Sec<strong>on</strong>d, there are substantial market uncertainties. In other words, it<br />
is difficult to predict the potential size <strong>and</strong> expansi<strong>on</strong> of the market. N<strong>on</strong>etheless, these risks are<br />
manageable to some extent via:<br />
•<br />
•<br />
•<br />
Co-operati<strong>on</strong> with lead users<br />
Limiting financial risks related to financing innovati<strong>on</strong><br />
Maintaining a high degree of flexibility <strong>and</strong> reactivity.<br />
59<br />
277
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Grant (2002) enumerates five different strategies through which innovative firms can reap the<br />
benefits of their innovati<strong>on</strong>s. These strategies, shown in Table 3.2, can be associated with<br />
differing involvement of he firms, different risk-taking, resource requirements <strong>and</strong> firm size.<br />
Strategies in Innovative <strong>Markets</strong><br />
Strategy<br />
Internal<br />
commercializati<strong>on</strong><br />
Joint venture<br />
Strategic alliance<br />
Outsourcing<br />
Licensing<br />
Risk<br />
Necessary firm size Resource requirement<br />
High risk, high returns <strong>and</strong><br />
Medium-large High<br />
complete c<strong>on</strong>trol<br />
Sharing investment <strong>and</strong> risk but<br />
also risk of c<strong>on</strong>flict with the partner Medium-large Resource <strong>and</strong><br />
capability synergies<br />
company<br />
Medium am<strong>on</strong>g several firms<br />
Limits risk but increases<br />
dependence <strong>on</strong> the outside world<br />
Small risk <strong>and</strong> small returns<br />
Small<br />
Small<br />
Access to outside<br />
resources<br />
Low<br />
Source: Adapted from Grant (2002 p342).<br />
Each of these strategies is described below:<br />
• The strategy of internal commercializati<strong>on</strong> means that the firm commercializes the new<br />
product — that is, the output of the innovati<strong>on</strong> — without any external productive help.<br />
This is the highest level of involvement that goes h<strong>and</strong> in h<strong>and</strong> with the heaviest<br />
investments <strong>and</strong> therefore the highest risk. Nevertheless, the firms can have the whole<br />
process under its c<strong>on</strong>trol. In doing so, the resource requirements are substantial, which<br />
implies that the firm is expected to be fairly large to be able to bear the burdens.<br />
•<br />
•<br />
•<br />
•<br />
The decisi<strong>on</strong> to start a joint venture still implies a great deal of involvement. By c<strong>on</strong>trast, it<br />
also helps to share the cost <strong>and</strong> the risk of the investment. In additi<strong>on</strong> to that, the<br />
participating firms can pool their resources together <strong>and</strong> thus benefit from synergies. On<br />
the other h<strong>and</strong>, however, they run the risk of potential disagreement <strong>and</strong> possible<br />
differences in corporate culture.<br />
Strategic alliance represents a middle way between total <strong>and</strong> very small involvement.<br />
Risk <strong>and</strong> resources can be shared.<br />
Outsourcing is suited to both large <strong>and</strong> small companies so as to enable access to<br />
outside resources without too much commitment.<br />
Licensing is typically — but not exclusively — employed by small firms, which do not<br />
possess the necessary resources to exploit the innovati<strong>on</strong> <strong>on</strong> their own.<br />
Protecting competitive advantage<br />
The sec<strong>on</strong>d strategy can be understood as efforts to protect <strong>and</strong> maintain a competitive<br />
advantage that has been achieved through value creati<strong>on</strong>. In particular, it c<strong>on</strong>cerns efforts firms<br />
60<br />
278
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
may take to prevent imitati<strong>on</strong> by rivals, since in competitive envir<strong>on</strong>ments, imitati<strong>on</strong> is a crucial<br />
force that undermines the competitive advantage of a firm over time. Such behavior may be<br />
both legal (e.g. patenting) <strong>and</strong> illegal (e.g. certain types of anticompetitive behavior that inhibit<br />
rivals access to a market).<br />
Once a firm has been able to gain competitive advantage in some respect, it will have incentives<br />
to protect <strong>and</strong> maintain this. In particular, where a firm has benefited relative to competitors from<br />
an important innovati<strong>on</strong> it will be keen that rival firms do not undermine this advantage through<br />
imitati<strong>on</strong>.<br />
One way in which firms achieve such protecti<strong>on</strong> is through various legal instituti<strong>on</strong>s put in place<br />
for this purpose (e.g. patent policy). These are highlighted below before we turn to other ways in<br />
which a source of competitive advantage can be maintained.<br />
The use of legal instituti<strong>on</strong>s against imitati<strong>on</strong><br />
Intellectual property rights are legal instituti<strong>on</strong>s that may protect the competitive advantage a firm<br />
has realized through innovati<strong>on</strong> by restricting the degree of imitati<strong>on</strong> of the innovati<strong>on</strong> that can<br />
take place.<br />
Patent policy provides a firm with protecti<strong>on</strong> against direct imitati<strong>on</strong> of certain types of inventive<br />
innovati<strong>on</strong>, for a period of time, subject to disclosure of informati<strong>on</strong> relating to the innovati<strong>on</strong>.<br />
Similarly, copyright protecti<strong>on</strong> prevents the direct copying of new c<strong>on</strong>tent. In both cases the aim<br />
of the legal protecti<strong>on</strong> is to provide incentives for the innovati<strong>on</strong> to take place, exactly because in<br />
the absence of this protecti<strong>on</strong> such incentives for such innovati<strong>on</strong> would be mitigated by the<br />
threat of direct imitati<strong>on</strong> <strong>and</strong> risk of immediate erosi<strong>on</strong> of any competitive advantage accruing to<br />
the innovati<strong>on</strong>.<br />
N<strong>on</strong>etheless, there are ways in which patent policy may be used by firms against the interests of<br />
c<strong>on</strong>sumers. Pre-emptive patenting describes the case where firms gather a series of patent<br />
around an initial innovati<strong>on</strong>, not for the purposes of using those patents in the introducti<strong>on</strong> of new<br />
products, but rather to stop other firms form inventing around the original innovati<strong>on</strong>. Thus the<br />
new patented products or technologies are neither used by the patentee nor licensed to other<br />
firms, but their value derives from the fact that rivals cannot use them rather than the use the<br />
holder makes of them.<br />
Alternatively, where a particular product or service protected by patent represents a bottleneck to<br />
a range of related markets, the patent holder may be able to use its protecti<strong>on</strong> in the supply of the<br />
patented product to enjoy competitive advantage in the supply of products in these related<br />
markets.<br />
In some sense, there is overlap between creati<strong>on</strong> of competitive advantage through product<br />
differentiati<strong>on</strong> <strong>and</strong> maintenance of this advantage through br<strong>and</strong>ing. Where firms can link an<br />
innovative product to a successful br<strong>and</strong> image, <strong>and</strong> protect that br<strong>and</strong> from imitati<strong>on</strong> (e.g. via<br />
trademark law), the advantage from the initial innovati<strong>on</strong> can be maintained in the face of<br />
imitati<strong>on</strong> as to the products physical characteristics. For example, in the pharmaceutical industry,<br />
61<br />
279
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
successful creati<strong>on</strong> of goodwill, reputati<strong>on</strong> <strong>and</strong> br<strong>and</strong> recogniti<strong>on</strong> allow firms to enjoy some (albeit<br />
less) competitive advantage over a drug that they have created <strong>on</strong>ce the patent has expired.<br />
Other protecti<strong>on</strong> against innovati<strong>on</strong><br />
Besides the use of intellectual property rights, the competitive advantage associated with an<br />
innovati<strong>on</strong> can be obtained through other mechanisms.<br />
The first strategy is secrecy. Where an innovati<strong>on</strong> is hard to observe or hard to codify, the<br />
innovative firm may able to maintain competitive advantage simply by not facilitating the flow of<br />
informati<strong>on</strong> about the innovati<strong>on</strong> to competitors.<br />
A sec<strong>on</strong>d strategy to protect competitive advantage is to merge with rivals. This could provide<br />
protecti<strong>on</strong> where a particular rival is capable of reducing the competitive advantage a firm has<br />
gained through innovati<strong>on</strong>. Of course, the merger could bring other benefits that are associated<br />
with creati<strong>on</strong> rather than maintenance of competitive advantage. For example, the combinati<strong>on</strong><br />
of tacit knowledge in the merged parties may bring greater capability to innovate.<br />
Finally, a firm may be able to with sufficient market power may be able to protect its competitive<br />
advantage by using this power to restrict rivals’ ability to compete with it. Besides predati<strong>on</strong>, a<br />
firm may be able to use competitive advantage with regard to <strong>on</strong>e good or service, <strong>and</strong> spread<br />
this advantage to related services, for example:<br />
• denying the rival access to some important input service, ranging from necessary raw<br />
materials <strong>and</strong> comp<strong>on</strong>ents to services that allow supply of a service to particular<br />
customers, or even IP rights; or<br />
•<br />
bundling or tying products such that rivals are forced to compete against the bundle of<br />
services offered by a firm <strong>and</strong> the competitive advantage the firm holds is transferred<br />
across the range of products supplied together.<br />
Not <strong>on</strong>ly could such a strategy help maintain the competitive advantage from the original line of<br />
innovati<strong>on</strong>, but it may also serve to exp<strong>and</strong> the ec<strong>on</strong>omic area of which returns from the<br />
competitive advantage can be enjoyed. However, it is essential to recognize that such strategies<br />
may also be c<strong>on</strong>ducive to value generati<strong>on</strong>. For instance, with respect to bundling, this pricing<br />
opti<strong>on</strong> may be the <strong>on</strong>ly way in which sufficient revenue could be generated to cover the initial<br />
outlay required for an innovati<strong>on</strong>.<br />
3.2 Timing of Innovati<strong>on</strong> <strong>and</strong> Life Cycles<br />
Having set out firm strategies according to two broad categories, we discuss below issues of<br />
timing. The first sub-secti<strong>on</strong> c<strong>on</strong>siders how the success of innovati<strong>on</strong>, <strong>and</strong> thus the optimal<br />
innovati<strong>on</strong> strategy, may be dependant <strong>on</strong> timing. The sec<strong>on</strong>d sub-secti<strong>on</strong> c<strong>on</strong>siders product<br />
‘life-cycles’ more generally.<br />
62<br />
280
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
First-mover advantage<br />
An essential questi<strong>on</strong> in innovative markets is to choose a strategy whether to enter early in the<br />
market <strong>and</strong> invest heavily in research <strong>and</strong> development in the hope of future profits <strong>and</strong> market<br />
leadership or to adopt a wait-<strong>and</strong>-see policy <strong>and</strong> to follow the technology leader by copying the<br />
innovati<strong>on</strong>. The benefits of a first-mover strategy relies <strong>on</strong> three c<strong>on</strong>diti<strong>on</strong>s:<br />
•<br />
•<br />
•<br />
Length of the lead time (i.e. the time over which the innovati<strong>on</strong> is protected from imitators<br />
<strong>and</strong> followers) either through IP rights or less formally.<br />
The extent to which complementary resources are required while exploiting the<br />
innovati<strong>on</strong>. If the need for complementary resources is relatively low, the initial investment<br />
<strong>and</strong> therefore the risk is lower. This may stimulate firms to try to be the first-mover.<br />
The possibility to set a st<strong>and</strong>ard is also likely to give an impetus for first-movers.<br />
The reas<strong>on</strong> why st<strong>and</strong>ards are set is closely c<strong>on</strong>nected with positive network externalities<br />
(discussed further below in the c<strong>on</strong>text of the ec<strong>on</strong>omics of networks). The higher the number of<br />
users of the st<strong>and</strong>ardized product, the more valuable the product to individual user. Positive<br />
network externalities have three major sources:<br />
•<br />
•<br />
•<br />
users of the product are c<strong>on</strong>nected with each other through networks (such as<br />
transportati<strong>on</strong> <strong>and</strong> the internet);<br />
the ease with which complementary products are available (for example, software<br />
applicati<strong>on</strong>s for Windows); <strong>and</strong><br />
switching cost between different networks: the larger the network the customer uses, the<br />
lower the costs to switch to another network (for example, mobile teleph<strong>on</strong>y networks).<br />
Generally, technological st<strong>and</strong>ards tend to result in ever increasing positive network externalities.<br />
If a technological st<strong>and</strong>ard appears to be dominant in the industry, more <strong>and</strong> more users are likely<br />
to choose this <strong>on</strong>e that may easily lead to winner-takes-all situati<strong>on</strong>. This is <strong>on</strong>e reas<strong>on</strong> why<br />
innovative firms try hard to establish their technology as the industry st<strong>and</strong>ard.<br />
Shapiro <strong>and</strong> Varian (1999) identify three possible <strong>and</strong> highly complementary strategies how firms<br />
could establish their own technology as the st<strong>and</strong>ard. The first strategy builds <strong>on</strong> obtaining the<br />
support of other, rival companies <strong>and</strong> firms producing complementary products, possibly through<br />
tacit agreements. The sec<strong>on</strong>d approach c<strong>on</strong>sists of market pre-empti<strong>on</strong>. Finally, <strong>and</strong> importantly,<br />
the firm aiming at winning the st<strong>and</strong>ard war has to make the impressi<strong>on</strong> from the very outset that<br />
its st<strong>and</strong>ard will be accepted at the end of the st<strong>and</strong>ard war. Put it another way, marketing <strong>and</strong><br />
other communicati<strong>on</strong> techniques can be employed so as to create self-fulfilling expectati<strong>on</strong>s<br />
am<strong>on</strong>g customers <strong>and</strong> rivals.<br />
However, different firms may have different strategic windows. These are defined as the time<br />
period during which they are able to seize market opportunities in line with their resources <strong>and</strong><br />
capabilities. Smaller firms tend to have shorter strategic windows because they cannot wait for a<br />
63<br />
281
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
l<strong>on</strong>g time to introduce new innovati<strong>on</strong>s <strong>and</strong> procure competitive advantage as opposed to large<br />
corporati<strong>on</strong> that do not tend to rush in introducing new technologies <strong>and</strong> take unnecessary risks<br />
given their solid financial background.<br />
The role of the life-cycle of an industry<br />
The positi<strong>on</strong> in the life cycle of a particular product, market or industry is likely to have a role to<br />
play as regards the strategy adopted by any firm towards the creati<strong>on</strong> <strong>and</strong> protecti<strong>on</strong> of<br />
competitive advantage. Life cycle theory sets out a useful analytical framework for analyzing this.<br />
In this c<strong>on</strong>text, there is some ambiguity in the relevant literature regarding terminology between<br />
product life cycles <strong>and</strong> industry life cycles. In principle these are different c<strong>on</strong>cepts since industry<br />
usually refers to a wider set of entities that a product, but these terms refer broadly to the same<br />
approach <strong>and</strong> are sometimes used interchangeably. In some sense the c<strong>on</strong>cepts of an industry<br />
life cycle is associated with the industrial organizati<strong>on</strong> literature (B<strong>on</strong>accorsi <strong>and</strong> Giuri (2000),<br />
Horvath et al. (2000) <strong>and</strong> Klepper <strong>and</strong> Kenneth (2001)), while marketing literature may be more<br />
likely to use term product life-cycle. In this overview we refer to product life cycles.<br />
Based <strong>on</strong> a series of case studies <strong>and</strong> empirical papers, Klepper (1996, pp. 564-565) comes up<br />
with five major regularities that best characterize the product life cycle:<br />
•<br />
•<br />
•<br />
•<br />
•<br />
The number of new entrants rises very rapidly at the beginning <strong>and</strong> then starts declining<br />
<strong>and</strong> eventually stabilizes at a very low level.<br />
The number of incumbents increases in the initial phase <strong>and</strong> then starts falling<br />
accompanied by a steady increase in output.<br />
Initially, market shares change quickly, i.e. some firms grow fast <strong>and</strong> others loose ground.<br />
However, after a while, market shares stabilize.<br />
Simultaneously, there is an increase in the diversity of products competing in the market<br />
because of a rise in product innovati<strong>on</strong>.<br />
However, as the market gets more mature <strong>and</strong> with the c<strong>on</strong>solidati<strong>on</strong> of the firms,<br />
incumbents spill more energy <strong>on</strong> process innovati<strong>on</strong>.<br />
Werker (2003) provides two more stylized facts associated with the product life cycle complement<br />
the analysis above. First that prices fall sharply at the outset of the life cycle with st<strong>and</strong>ardizati<strong>on</strong><br />
<strong>and</strong> a growing number of customers. However, later <strong>on</strong>, prices decrease at a lower pace <strong>and</strong> the<br />
decrease finally comes to a halt. Sec<strong>on</strong>d, that the beginning of the life cycle favors entry<br />
whereas established firms are favored to new entrants at later stages of the life cycle.<br />
More generally, the life cycle of a product may typically have four phases: (i) introducti<strong>on</strong>; (ii)<br />
shakeout; (iii) maturity; <strong>and</strong> (iv) decline.<br />
During the first phase, i.e. the introducti<strong>on</strong>, a few companies produce a wide range of goods with<br />
substantial differences in product features partly <strong>on</strong> the grounds of differences in the underlying<br />
64<br />
282
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
technology. The competiti<strong>on</strong> am<strong>on</strong>g technologies yields rapid product innovati<strong>on</strong>. Werker (2003)<br />
notes that at this stage of the product cycle, c<strong>on</strong>sumers’ preferences are not well established.<br />
Therefore, firms exert effort to discover the unclear preferences through high rates of innovati<strong>on</strong>.<br />
The sec<strong>on</strong>d stage of the life cycle usually witnesses a shakeout <strong>and</strong> the emergence of a<br />
dominant technology, accompanied by st<strong>and</strong>ardizati<strong>on</strong> of around this technology. This lays the<br />
ground for further rapid innovati<strong>on</strong>. The competitive envir<strong>on</strong>ment forces unsuccessful firms to exit<br />
the market <strong>and</strong> leads to mergers <strong>and</strong> acquisiti<strong>on</strong>s. Product differentiati<strong>on</strong> in this phase can be<br />
achieved mainly through new product design <strong>and</strong> by c<strong>on</strong>tinually improving quality.<br />
For instance, Klepper <strong>and</strong> Sim<strong>on</strong>s (2001) provide empirical evidence for dramatic shakeouts that<br />
occurred at some point in the automobiles, tires, televisi<strong>on</strong>s <strong>and</strong> penicillin industries in the US.<br />
They suggest that earlier entrants are more likely to survive shakeouts. Shakeouts do not seem<br />
to be triggered by drastic events but rather the c<strong>on</strong>solidati<strong>on</strong> in these industries was a result of the<br />
process of competiti<strong>on</strong> during which early entrants gain high market shares through innovati<strong>on</strong>. A<br />
similar c<strong>on</strong>clusi<strong>on</strong> is drawn by Horvath et al. (2000), who find that the shakeout in the US beer<br />
brewing industry occurred from 1880 to 1890 was a c<strong>on</strong>sequence of earlier mass entry <strong>and</strong> the<br />
subsequent competiti<strong>on</strong>.<br />
When an industry reaches maturity, products tend to become increasingly homogeneous <strong>and</strong><br />
could be viewed as commodities. In this c<strong>on</strong>text, <strong>and</strong> to counter this tendency, br<strong>and</strong>ing <strong>and</strong> the<br />
introducti<strong>on</strong> of complementary services represent an effective way to differentiate products. Price<br />
competiti<strong>on</strong> also tends to become more intense in this phase of the life cycle. Hence, cost<br />
advantage is a crucial aspect of competitive advantage. Simultaneously, the technology<br />
developed in earlier stages becomes diffused <strong>and</strong> firms seek to improve this technology<br />
incrementally.<br />
The final phase can be referred to as the decline of the industry. During this phase fierce <strong>and</strong><br />
possibly destructive price wars become the norm given that product differentiati<strong>on</strong> is extremely<br />
difficult. The lack or at least the very low pace of product <strong>and</strong> process innovati<strong>on</strong> is <strong>on</strong>e reas<strong>on</strong><br />
for this. As a result, some firms decide to exit from the market <strong>and</strong> this further c<strong>on</strong>tributes to the<br />
decline.<br />
Different industries may be very different as regards the intensity of competiti<strong>on</strong> <strong>and</strong> the pace of<br />
innovati<strong>on</strong> in the same stage of the life cycle. Accordingly, three types of industries can be<br />
identified:<br />
• In local m<strong>on</strong>opoly markets, products are very specialized <strong>and</strong> target narrow market<br />
segments. Products are of high quality with high unit values, which implies that volume of<br />
producti<strong>on</strong> is low. Because of substantial differentiati<strong>on</strong>, competiti<strong>on</strong> is very limited.<br />
Examples for this kind of market are found am<strong>on</strong>gst the following areas: highway <strong>and</strong><br />
residential c<strong>on</strong>structi<strong>on</strong>, surgical appliances, defense industry.<br />
•<br />
Traditi<strong>on</strong>al industrial markets are characterized by homogenous products where<br />
competitive advantage can be usually achieved via cost leadership, br<strong>and</strong>ing or product<br />
65<br />
283
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
variety. Passenger cars <strong>and</strong> household equipment are examples of products bel<strong>on</strong>ging to<br />
this category<br />
• Schumpeterian markets are dominated by product innovati<strong>on</strong>, where incumbents can be<br />
completely displaced by entrants. These markets might be seen within the computer<br />
software industry <strong>and</strong> within c<strong>on</strong>sumer electr<strong>on</strong>ics.<br />
3.3 Issues Relating to Network Industries<br />
Network industries give rise to a number of issues that are likely to affect — or at least influence<br />
— the strategies used by firms to innovate. The characteristic features of these industries<br />
originate from network externalities. A positive network externality arises when “a good is more<br />
valuable to a user the more users adopt the same good or compatible <strong>on</strong>es”. A comm<strong>on</strong><br />
distincti<strong>on</strong> is that between direct <strong>and</strong> indirect network externalities.<br />
Direct externalities typically occur in a two-way (physical) network. This externality reflects the<br />
fact, for instance, that a teleph<strong>on</strong>e user benefits from others being c<strong>on</strong>nected to the same<br />
network: an additi<strong>on</strong>al teleph<strong>on</strong>e increases the number of potential communicati<strong>on</strong> within the<br />
system, <strong>and</strong> thus the value of membership.<br />
Indirect externalities arise from the fact that a network of users increases the incentives to<br />
produce compatible products that are complementary with the network good. These externalities<br />
arise in systems, which can be defined as “[…] collecti<strong>on</strong>s of two or more comp<strong>on</strong>ents together<br />
with an interface that allows the comp<strong>on</strong>ents to work together” (Katz <strong>and</strong> Shapiro, 1994, p93).<br />
The typical system comprises of a “platform” (or hardware) <strong>and</strong> of “applicati<strong>on</strong>s” (or software) that<br />
can <strong>on</strong>ly be used with a platform. Compatibility refers to the possibility that software can be used<br />
with a particular platform. The network externality arises when users make their purchase over<br />
time, because in the presence of ec<strong>on</strong>omies of scale a greater number of complementary<br />
products can be supplied at a lower price when the network grows.<br />
The existence of network externalities has a substantial impact <strong>on</strong> the way technologies are<br />
chosen <strong>and</strong> promoted. In this secti<strong>on</strong> we follow closely Katz <strong>and</strong> Shapiro (1994) overview of<br />
systems competiti<strong>on</strong>s <strong>and</strong> their identificati<strong>on</strong> of three main issues that have been addressed in<br />
the literature:<br />
•<br />
•<br />
technology adopti<strong>on</strong> decisi<strong>on</strong>s;<br />
product selecti<strong>on</strong> decisi<strong>on</strong>s; <strong>and</strong><br />
.<br />
66<br />
284
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
• compatibility decisi<strong>on</strong>s.<br />
<strong>Technology</strong> adopti<strong>on</strong> decisi<strong>on</strong>s<br />
A number of issues that arise in markets characterized by network externalities can be addressed<br />
by c<strong>on</strong>sidering a system <strong>on</strong>ly, setting aside competiti<strong>on</strong> between different systems.<br />
The existence of direct network externalities drives a wedge between private <strong>and</strong> social<br />
incentives to join the network. A single user does not c<strong>on</strong>sider the benefits that would accrue to<br />
others by his joining the network <strong>and</strong> as a result the market may lead to network of inefficient<br />
small size. In fact, the benefits that an individual would take into c<strong>on</strong>siderati<strong>on</strong> when choosing<br />
whether to join a network depends not <strong>on</strong>ly <strong>on</strong> the current size of the network, but also <strong>on</strong> its<br />
expected future size.<br />
C<strong>on</strong>sumers’ expectati<strong>on</strong>s play a central role in driving market outcomes where there are<br />
technology decisi<strong>on</strong>s in markets with important network effects. An important c<strong>on</strong>sequence is that<br />
multiple equilibria can easily arise. For instance, if all c<strong>on</strong>sumers expect no <strong>on</strong>e else to join the<br />
network, then its size would be zero, even if the network may be valuable for c<strong>on</strong>sumers; if all<br />
expect every<strong>on</strong>e to join, the network may achieve a large size.<br />
Indirect externalities may have a different impact <strong>on</strong> the size of the network. In a system market,<br />
<strong>on</strong>e c<strong>on</strong>sumer’s adopti<strong>on</strong> decisi<strong>on</strong> would does not affect other c<strong>on</strong>sumers, given the prices <strong>and</strong><br />
varieties of products available. The externalities arise indirectly through the impact of <strong>on</strong>e<br />
c<strong>on</strong>sumers’ adopti<strong>on</strong> decisi<strong>on</strong> of the future variety or prices of applicati<strong>on</strong>s. Katz <strong>and</strong> Shapiro<br />
(1994) note that in such an envir<strong>on</strong>ment, if cost c<strong>on</strong>diti<strong>on</strong>s are c<strong>on</strong>sistent with the existence of a<br />
competitive equilibrium, a perfectly competitive equilibrium arises in system markets. When, at<br />
the other extreme, a m<strong>on</strong>opolist is the sole supplier of the comp<strong>on</strong>ents of a system, there are<br />
issues regarding multi-product pricing <strong>and</strong> inter-temporal pricing.<br />
We follow Katz <strong>and</strong> Shapiro (1994) closely to discuss these issues. C<strong>on</strong>sider the case where the<br />
m<strong>on</strong>opolist can commit to the prices of both comp<strong>on</strong>ents of the system:<br />
•<br />
•<br />
with fixed proporti<strong>on</strong>s technology, the m<strong>on</strong>opolist simply sets the price that maximizes<br />
profits in the usual way.<br />
with a variable proporti<strong>on</strong>s technology, issues of bundling <strong>and</strong> price discriminati<strong>on</strong> arise<br />
but again the inefficiencies are attributable to m<strong>on</strong>opoly power rather than externalities.<br />
If the supplier cannot commit to prices in advance, buyer expectati<strong>on</strong>s are again crucial: a<br />
m<strong>on</strong>opolist would like to c<strong>on</strong>vince c<strong>on</strong>sumers applicati<strong>on</strong>s will be available at low price in the<br />
future. After c<strong>on</strong>sumers are locked in, however, the m<strong>on</strong>opolist may raise price to the m<strong>on</strong>opoly<br />
level.<br />
In the literature, it is usually assumed that c<strong>on</strong>sumers’ expectati<strong>on</strong>s <strong>on</strong> the future prices of<br />
applicati<strong>on</strong>s are affected by the quantity of hardware currently sold which may act as a signal: a<br />
larger base may lead to cheaper <strong>and</strong>/or more varied applicati<strong>on</strong>s. Hence, the m<strong>on</strong>opolist may<br />
want to lower the price of hardware to create a larger network <strong>and</strong> software aftermarket.<br />
67<br />
285
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Product selecti<strong>on</strong> decisi<strong>on</strong>s<br />
Competiti<strong>on</strong> in network industries may take the form of competiti<strong>on</strong> between incompatible<br />
systems.<br />
Network externalities may lead to a de facto st<strong>and</strong>ardizati<strong>on</strong> whereby every<strong>on</strong>e uses the same<br />
system. Due to positive feedback elements, a system may become dominant <strong>on</strong>ce it has gained<br />
an initial edge, a phenomen<strong>on</strong> which is usually referred to as “tipping”.<br />
Two main factors, however, may limit tipping <strong>and</strong> sustain multiple networks:<br />
•<br />
•<br />
network externalities being exhausted at a smaller network size;<br />
c<strong>on</strong>sumers’ heterogeneity <strong>and</strong> product differentiati<strong>on</strong>.<br />
In the latter case multiple networks would reflect c<strong>on</strong>sumers’ love for variety <strong>and</strong> there is a typical<br />
trade-off between variety <strong>and</strong> st<strong>and</strong>ardizati<strong>on</strong>.<br />
Competiti<strong>on</strong> between systems may be very intense, at least before a dominant system, if any,<br />
would emerge as dominant. For instance, firms may be engaged in very intense price<br />
competiti<strong>on</strong> at an early stage for seeking to establish an installed base <strong>and</strong> achieve leadership.<br />
This competiti<strong>on</strong> can be interpreted as firms bidding for future m<strong>on</strong>opoly profits.<br />
Competiti<strong>on</strong> may also be played in trying to affect c<strong>on</strong>sumers’ expectati<strong>on</strong>s about the dominant<br />
system that would emerge.<br />
An important implicati<strong>on</strong> of the network externalities literature is that the market may settle in an<br />
equilibrium where the dominant system is that with a lower social evaluati<strong>on</strong>. <strong>Markets</strong> can exhibit<br />
“excess inertia” <strong>and</strong> remain locked into an obsolete st<strong>and</strong>ard, even though a better <strong>on</strong>e is<br />
available.<br />
However, markets can also exhibit “excess momentum” whereby the market tips inefficiently to<br />
new technologies. This could, for instance, arise when competiti<strong>on</strong> is between an older<br />
technology which is competitively supplied <strong>and</strong> a new technology which is sp<strong>on</strong>sored. The new<br />
technology that is sp<strong>on</strong>sored may have an advantage over an older technology that is more<br />
competitively supplied because the sp<strong>on</strong>sor of the technology may engage in pricing below costs<br />
or other types of investments (with the hope of recouping these later <strong>on</strong>ce the technology is<br />
established). N<strong>on</strong>etheless, this perceived risk is not straightforward — pricing very low from the<br />
outset may be the <strong>on</strong>ly way of introducing an efficient new technology, <strong>and</strong> tipping may be<br />
inevitable.<br />
Compatibility choices<br />
Compatibility between different networks/systems is often a choice variable of firms. Two types of<br />
compatibility can be c<strong>on</strong>sidered:<br />
• horiz<strong>on</strong>tal compatibility: between two comparable rival systems; <strong>and</strong><br />
68<br />
286
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
vertical compatibility: between successive generati<strong>on</strong>s of a technology.<br />
Compatibility entails both social benefits <strong>and</strong> costs:<br />
•<br />
•<br />
•<br />
•<br />
•<br />
compatibility exp<strong>and</strong>s the size of both networks thereby avoiding the cost of participating<br />
to two different networks (e.g. duplicate equipment);<br />
compatibility in systems may lead to lower producti<strong>on</strong> costs for ec<strong>on</strong>omics of scale,<br />
learning effects, etc;<br />
compatibility enhances variety by allowing c<strong>on</strong>sumers to combine comp<strong>on</strong>ents from<br />
various systems;<br />
the risks of adopting a particular technology are lower; <strong>and</strong><br />
costs derive from the mechanism by which compatibility is achieved. St<strong>and</strong>ardizati<strong>on</strong><br />
may lead to a loss of variety <strong>and</strong> may prevent the development of new incompatible<br />
systems; adapters that allow interfacing have a cost themselves <strong>and</strong> may work<br />
imperfectly.<br />
The nature of competiti<strong>on</strong> in the market is affected by compatibility decisi<strong>on</strong>s. For systems that<br />
are compatible, competiti<strong>on</strong> is essentially at the level of each comp<strong>on</strong>ent. For incompatible<br />
systems, competiti<strong>on</strong> is at the network or system level.<br />
Katz <strong>and</strong> Shapiro (1986) c<strong>on</strong>sider the impact of compatibility <strong>on</strong> pricing competiti<strong>on</strong> over time.<br />
Price competiti<strong>on</strong> is relaxed at earlier stages of the product life cycle because the market loses its<br />
winner-take-all feature. For the same reas<strong>on</strong>, however, competiti<strong>on</strong> may be intensified in later<br />
stages of the product life cycle.<br />
Compatibility decisi<strong>on</strong>s may be affected by a number of factors:<br />
•<br />
•<br />
•<br />
asymmetries in the probability of being the winner, e.g. reputati<strong>on</strong>;<br />
product differentiati<strong>on</strong>; <strong>and</strong><br />
installed base<br />
Side payments may help firms to reach an agreement <strong>on</strong> compatibility. When they are not<br />
feasible, it is useful to distinguish markets where a firm can unilaterally impose compatibility <strong>and</strong><br />
those where a firm can unilaterally impose incompatibility.<br />
In general, there may be different mechanisms <strong>and</strong> instituti<strong>on</strong>s that govern compatibility choices:<br />
st<strong>and</strong>ards committees, unilateral acti<strong>on</strong> <strong>and</strong> hybrid mechanisms. (Farrell <strong>and</strong> Sal<strong>on</strong>er, 1988)<br />
69<br />
287
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
4<br />
4.1<br />
BIBLIOGRAPHY<br />
Industrial Organizati<strong>on</strong><br />
Arrow, K J (1962) “Ec<strong>on</strong>omic Welfare <strong>and</strong> the Allocati<strong>on</strong> of Resources for Inventi<strong>on</strong>”, in<br />
Universities Nati<strong>on</strong>al Bureau Committee for Ec<strong>on</strong>omic Research. The rate <strong>and</strong> directi<strong>on</strong> of<br />
inventive activity Princet<strong>on</strong>, Princet<strong>on</strong> University Press.<br />
Baumol,William (2002) The Free Market Innovati<strong>on</strong> Machine,Princet<strong>on</strong>,N.J.,Princet<strong>on</strong> Univ.Press<br />
Beath, J, Poyago-Theotoki, J, Ulph, D, (1998), “Organizati<strong>on</strong> design <strong>and</strong> informati<strong>on</strong>-sharing in a<br />
research joint venture with spillovers”, Bulletin of Ec<strong>on</strong>omic Research, 50:1, pp.47-59.<br />
Beath, John, Katsoulacos, Yannis <strong>and</strong> Ulph, David (1987) “Sequential product innovati<strong>on</strong> <strong>and</strong><br />
industry evoluti<strong>on</strong>”, Ec<strong>on</strong>omic Journal (Supplement), Vol 99, p32-43.<br />
Beath, John, Katsoulacos, Yannis <strong>and</strong> Ulph, David (1994) “Strategic R&D innovati<strong>on</strong>” in Cable,<br />
John (ed) Current Issues in Industrial Ec<strong>on</strong>omics, Basingstoke: Macmillan.<br />
Beath, John, Katsoulacos, Yannis <strong>and</strong> Ulph, David (1995) “Game-theoretical approaches to the<br />
modelling of technological change” in St<strong>on</strong>eman, Paul (ed) H<strong>and</strong>book of the Ec<strong>on</strong>omics of<br />
Innovati<strong>on</strong> <strong>and</strong> Technological Change, Oxford UK <strong>and</strong> Cambridge USA: Blackwell.<br />
Bester, H. <strong>and</strong> Petrakis, E. (1993) “The incentives for cost reducti<strong>on</strong> in a differentiated industry”,<br />
Internati<strong>on</strong>al Journal of Industrial Organizati<strong>on</strong>, No16, p519-534.<br />
Bo<strong>on</strong>e, Jan (2000) “Competitive pressure: the effects <strong>on</strong> investments in product <strong>and</strong> process<br />
innovati<strong>on</strong>”, RAND Journal of Ec<strong>on</strong>omics, Vol 31, No 3, p549-569.<br />
Bo<strong>on</strong>e, Jan (2001) “Intensity of competiti<strong>on</strong> <strong>and</strong> the incentive to innovate”, Internati<strong>on</strong>al Journal of<br />
Industrial Organizati<strong>on</strong>, Vol 19, p705-726.<br />
Cohen, W M <strong>and</strong> Levinthal, D A (1989a) “The Implicati<strong>on</strong>s of Spillovers for R&D Investment <strong>and</strong><br />
Welfare: A New Perspective” in a Link <strong>and</strong> K. Smith eds, Advances in applied micro-ec<strong>on</strong>omics<br />
vol. 5: The factors affecting technological change. Greenwich, C<strong>on</strong>n: JAI Press.<br />
Cohen, W M <strong>and</strong> Levinthal, D A (1989b) “Innovati<strong>on</strong> <strong>and</strong> learning: the two faces of R&D –<br />
implicati<strong>on</strong>s for the analysis of R&D investment”, Ec<strong>on</strong>omic Journal, Vol 99,p569-96.<br />
d’Asprem<strong>on</strong>t, C, Jacquemin, A, (1988), “Cooperative <strong>and</strong> n<strong>on</strong>-cooperative R&D in a duopoly with<br />
spillovers”, American Ec<strong>on</strong>omic Review, 78, pp.1133-1137.<br />
Dasgupta, P <strong>and</strong> Stiglitz, J E (1980a) “Industrial structure <strong>and</strong> the nature of innovative activity”,<br />
Ec<strong>on</strong>omic Journal, 90:266-293.<br />
Dasgupta, P <strong>and</strong> Stiglitz, J E (1980b) “Uncertainty, industrial structure <strong>and</strong> the speed of R&D”,<br />
Bell Journal of Ec<strong>on</strong>omics, 11:1 28.<br />
Dasgupta, P. (1986) “The theory of technological competiti<strong>on</strong>”, in J.E. Stiglitz <strong>and</strong> F. Mathews<strong>on</strong><br />
eds., New <strong>Development</strong>s in the Analysis of Market Structure, L<strong>on</strong>d<strong>on</strong>: Macmillan.<br />
70<br />
288
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Delb<strong>on</strong>o, F. <strong>and</strong> Denicolò, Vincenzo (1990) “R&D investment in a symmetric <strong>and</strong> homogeneous<br />
oligopoly: Bertr<strong>and</strong> vs Cournot”, Internati<strong>on</strong>al Journal of Industrial Organizati<strong>on</strong>, p297-313<br />
Fudenberg, D & Tirole, J (1983) “Learning-by-doing <strong>and</strong> market performance”, Bell Journal,<br />
14:522-530.<br />
Fudenberg, D., Gilbert, R, Stiglitz, J. <strong>and</strong> Tirole, Jean (1983) “Preempti<strong>on</strong>, leapfrogging <strong>and</strong><br />
competiti<strong>on</strong> in patent races”, European Ec<strong>on</strong>omic Review, Vol 22, p3-31<br />
Gilbert, R J <strong>and</strong> Newbery, D M G (1982) “Preemptive patenting <strong>and</strong> the persistence of m<strong>on</strong>opoly”,<br />
American Ec<strong>on</strong>omic Review, 72:514-526.<br />
Greenlee, P, Cassiman, B, (1999), “Product market objectives <strong>and</strong> the formati<strong>on</strong> of research joint<br />
ventures”, Managerial <strong>and</strong> Decisi<strong>on</strong> Ec<strong>on</strong>omics, 20(3), pp.115-130<br />
Grossman, G M <strong>and</strong> Shapiro, C (1987) “Dynamic R&D competiti<strong>on</strong>”, Ec<strong>on</strong>omic Journal,<br />
97:372-387.<br />
Hagedoorn, J, Schakenraad, J, (1992), “Leading companies <strong>and</strong> networks of strategic alliances<br />
in informati<strong>on</strong> technologies”, Research Policy, 21, pp. 163-190.<br />
Harris. Christopher <strong>and</strong> Vickers, John (1987) “Racing with uncertainty”, Review of Ec<strong>on</strong>omic<br />
Studies, Vol 54, No 1, p1-21<br />
Jacquemin, A, Soete, L, (1994), “Cooperati<strong>on</strong> in R&D, efficiency <strong>and</strong> European antitrust policy”,<br />
European Review, 2, pp.65-72.<br />
Judd, K L (1985b) “On The Performance of Patents”, Ec<strong>on</strong>ometrica, 53:567-585.<br />
Kamien, M, Muller, E, <strong>and</strong> Zang, I, (1992), “Research joint ventures <strong>and</strong> R&D cartels”, American<br />
Ec<strong>on</strong>omic Review, 82, pp.1293-1306.<br />
Katsoulacos, Y, Ulph, D, (1998), “Endogenous spillovers <strong>and</strong> the performance of research joint<br />
ventures”, The Journal of Industrial Ec<strong>on</strong>omics, 3, pp.333-357.<br />
Katz, L M <strong>and</strong> Shapiro, C (1986) “<strong>Technology</strong> Adopti<strong>on</strong> in the Presence of Network Externalities”,<br />
Journal of Political Ec<strong>on</strong>omy, 94: 822-841.<br />
Katz, M L <strong>and</strong> Shapiro, C (1986) “How to License Intangible Property”, Quarterly Journal of<br />
Ec<strong>on</strong>omics, 101:567-590.<br />
Katz, M L <strong>and</strong> Shapiro, C (1986) “<strong>Technology</strong> Adopti<strong>on</strong> in the Presence of Network Externalities”,<br />
Journal of Political Ec<strong>on</strong>omy, 94: 822-841.<br />
Katz, M L <strong>and</strong> Shapiro, C (1987) “R&D Rivalry with Licensing or Imitati<strong>on</strong>”, American Ec<strong>on</strong>omic<br />
Review, 77:402-420.<br />
71<br />
289
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Katz, M L, (1986), “An analysis of cooperative research <strong>and</strong> development”, RAND Journal of<br />
Ec<strong>on</strong>omics, 17 (Winter), pp.527-543.<br />
Katz, M L, Ordover, J A, (1990), “R&D cooperati<strong>on</strong> <strong>and</strong> competiti<strong>on</strong>”, Brookings Papers <strong>on</strong><br />
Ec<strong>on</strong>omic Activity, Special Issue, pp. 137-91.<br />
Kleinknecht, A, Reijnen, J, (1992), “Why do firms cooperate <strong>on</strong> R&D? an empirical study”,<br />
Research Policy, 21, 347-360.<br />
Lee, T <strong>and</strong> Wilde, I (1980) “Market Structure <strong>and</strong> Innovati<strong>on</strong>: A Reformulati<strong>on</strong>”, Quarterly Journal<br />
of Ec<strong>on</strong>omics, 94:429, 436.<br />
Levin R, Klevorick, A K, Nels<strong>on</strong>, RR <strong>and</strong> Winter, S G (1987) “Appropriating the returns from<br />
industrial research <strong>and</strong> development”, Brookings Papers <strong>on</strong> Ec<strong>on</strong>omic Activity, vol. 3:783-820.<br />
Loury, G C (1979) “Market Structure <strong>and</strong> Innovati<strong>on</strong>”. Quarterly Journal of Ec<strong>on</strong>omics. 93(3):<br />
395-410.<br />
Perloff, Jeffrey , Karp, Larry <strong>and</strong> Golan, Amos (2007) Estimating Market Power <strong>and</strong> Strategies,<br />
Cambridge,Ma.: Cambridge Univ. Press<br />
Poyago-Theotoki, J, (1999), “A note <strong>on</strong> endogenous spillovers in a n<strong>on</strong>-tournament R&D<br />
duopoly”, Review of Industrial Organizati<strong>on</strong>, 15(3), November, pp.253-262.<br />
Qiu, L. (1997) “On the dynamic efficiency of Bertr<strong>and</strong> <strong>and</strong> Cournot equilibria”, Journal of<br />
Ec<strong>on</strong>omic Theory, NO 75, p213-229<br />
Reinganum, J F (1981a) “Dynamic Games of Innovati<strong>on</strong>”, Journal of Ec<strong>on</strong>omic Theory, 25:21-41;<br />
1985 “Corrigendum” 35:196-197.<br />
Reinganum, J F (1981b) “On the Diffusi<strong>on</strong> of New <strong>Technology</strong>: A Game Theoretic Approach”,<br />
Review of Ec<strong>on</strong>omic Studies, 48:395-405.<br />
Reinganum, J F (1981c) “Market Structure <strong>and</strong> the Diffusi<strong>on</strong> of New <strong>Technology</strong>”, Bell Journal,<br />
12:618-624.<br />
Reinganum, J F (1982a) “A Dynamic Game of R <strong>and</strong> D: Patent Protecti<strong>on</strong> <strong>and</strong> Competitive<br />
Behaviour”, Ec<strong>on</strong>ometrica, 50:671-688.<br />
Reinganum, J F (1982b) “Strategic Search Theory”, Internati<strong>on</strong>al Ec<strong>on</strong>omic Review, 23:1-15.<br />
Reinganum, Jennifer F. (1983) “Uncertain innovati<strong>on</strong> <strong>and</strong> the persistence of m<strong>on</strong>opoly”, American<br />
Ec<strong>on</strong>omic Review, Vol 73, p741-48<br />
Reinganum, Jennifer F. (1985) “Innovati<strong>on</strong> <strong>and</strong> industry evoluti<strong>on</strong>”, Quarterly Journal of<br />
Ec<strong>on</strong>omics, p81-99<br />
Segerstrom, P., 2007. Intel ec<strong>on</strong>omics. Internati<strong>on</strong>al Ec<strong>on</strong>omic Review 48 (1), 247–280<br />
Simps<strong>on</strong>, R D, V<strong>on</strong>ortas, N S, (1994), “Cournot equilibrium with imperfectly appropriable R&D”,<br />
The Journal of Industrial Ec<strong>on</strong>omics, 1, pp.79-92<br />
72<br />
290
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Spence, A M (1984) “Cost Reducti<strong>on</strong>, Competiti<strong>on</strong>, <strong>and</strong> Industry Performance”, Ec<strong>on</strong>ometrica<br />
52:101-121.<br />
Sutt<strong>on</strong>, John (1998) <strong>Technology</strong> <strong>and</strong> Market Structure, Cambridge, Massachusetts <strong>and</strong> L<strong>on</strong>d<strong>on</strong>,<br />
Engl<strong>and</strong>: The MIT Press.<br />
Suzumura, K, (1992), “Cooperative <strong>and</strong> n<strong>on</strong>-cooperative R&D in an oligopoly with spillovers”,<br />
American Ec<strong>on</strong>omic Review, 82, pp.1307-1320.<br />
Vickers, John (1986) “The evoluti<strong>on</strong> of industry structure when there is a sequence of<br />
innovati<strong>on</strong>s”, Journal of Industrial Ec<strong>on</strong>omics, Vol 35,p1-12<br />
4.2 Endogenous Growth<br />
Aghi<strong>on</strong>,Philippe <strong>and</strong> Giffith,Rachel (2005) Competiti<strong>on</strong> <strong>and</strong> Growth, Cambridge,Ma., The MIT<br />
Press<br />
Aghi<strong>on</strong>, Philippe <strong>and</strong> Howitt, Peter (1988) “Growth <strong>and</strong> Cycles through Creative Destructi<strong>on</strong>”, MIT<br />
Mimeo.<br />
Aghi<strong>on</strong>, Philippe <strong>and</strong> Howitt, Peter (1992) “A model of growth through creative destructi<strong>on</strong>”,<br />
Ec<strong>on</strong>ometrica, Vol 60,p323-51.<br />
Aghi<strong>on</strong>, Philippe <strong>and</strong> Howitt, Peter (1998) Endogenous Growth Theory, Cambridge,<br />
Massachusetts <strong>and</strong> L<strong>on</strong>d<strong>on</strong>, Engl<strong>and</strong>: The MIT Press.<br />
Aghi<strong>on</strong>, Philippe <strong>and</strong> Shankerman, Mark (2000), “A model of market enhancing infrastructure”<br />
CEPR Discussi<strong>on</strong> Paper, NO.2461.<br />
Aghi<strong>on</strong>, Philippe, Bloom, Nicholas, Blundell, Richard, Griffith, Rachel <strong>and</strong> Howitt, Peter (2002)<br />
“Competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong>: an inverted U relati<strong>on</strong>ship” NBER Working Paper, no.9269.<br />
Aghi<strong>on</strong>, Philippe, Dewatrip<strong>on</strong>t, M. <strong>and</strong> Rey, P. (1997) “Competiti<strong>on</strong>, financial discipline <strong>and</strong><br />
growth”, Review of Ec<strong>on</strong>omic Studies, Vol 66, No 4, p825-852.<br />
Aghi<strong>on</strong>, Philippe, Harris, Christopher <strong>and</strong> Vickers, John (1995) “Competiti<strong>on</strong> <strong>and</strong> Growth with<br />
Step-by-Step Technological Progress”, mimeo.<br />
Aghi<strong>on</strong>, Philippe, Harris, Christopher <strong>and</strong> Vickers, John (1997) “Competiti<strong>on</strong> <strong>and</strong> growth with<br />
step-by-step innovati<strong>on</strong>: an example”, European Ec<strong>on</strong>omic Review, Vol 41, p771-782.<br />
Aghi<strong>on</strong>, Philippe, Harris, Christopher, Howitt, Peter <strong>and</strong> Vickers, John (2001) “Competiti<strong>on</strong>,<br />
imitati<strong>on</strong> <strong>and</strong> growth with step-by-step innovati<strong>on</strong>”, Review of Ec<strong>on</strong>omic Studies, Vol 68, p467-<br />
492.<br />
Aghi<strong>on</strong>,Philippe <strong>and</strong> Griffith,Rachel (2005) Competiti<strong>on</strong> <strong>and</strong> Growth, Cambridge,Ma., The MIT Pr<br />
Ahn, Sangho<strong>on</strong>, “Competiti<strong>on</strong>, innovati<strong>on</strong> <strong>and</strong> productivity growth: a review of theory <strong>and</strong><br />
evidence”, OECD Ec<strong>on</strong>omic Department Working Paper, no 317.<br />
Arrow, K J (1962) “Ec<strong>on</strong>omic Welfare <strong>and</strong> the Allocati<strong>on</strong> of Resources for Inventi<strong>on</strong>”, in<br />
Universities Nati<strong>on</strong>al Bureau Committee for Ec<strong>on</strong>omic Research. The rate <strong>and</strong> directi<strong>on</strong> of<br />
inventive activity Princet<strong>on</strong>, Princet<strong>on</strong> University Press.<br />
73<br />
291
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Blundell, Richard, Griffith, Rachel <strong>and</strong> Van Reenen, John (1999) “Market share, market value <strong>and</strong><br />
innovati<strong>on</strong> in a panel of British manufacturing firms”, Review of Ec<strong>on</strong>omic Studies, Vol 66, No 3, p<br />
529-554.<br />
Chen,Derek H.C. <strong>and</strong> Kee,H.L. (2005),”A Model <strong>on</strong> Knowledge <strong>and</strong> Endogeneous Growth”, World<br />
Bank Policy Research Working Paper 3539, March<br />
D’Asprem<strong>on</strong>t, Claude, Dos Santos Ferreira, Rodolphe <strong>and</strong> Gérard-Varet, Louis-André (2002)<br />
“Strategic R&D investment, competiti<strong>on</strong> toughness <strong>and</strong> growth”, mimeo.<br />
Gilbert, Richard J. <strong>and</strong> Newbery, David M.G. (1982) “Preemptive Patenting <strong>and</strong> the Persistence<br />
of M<strong>on</strong>opoly”, American Ec<strong>on</strong>omic Review, 72:514-526.<br />
Grossman, Gene M. <strong>and</strong> Helpman, Elhanan. (1991) “Quality ladders <strong>and</strong> product cycles”,<br />
Quarterly Journal of Ec<strong>on</strong>omics, Vol 106, p557-586.<br />
Lucas, R E, (1988). “On the Mechanics of Ec<strong>on</strong>omic <strong>Development</strong>”, Journal of M<strong>on</strong>etary<br />
Ec<strong>on</strong>omics 22(1): 3-42.<br />
Nickell, Stephen, J. (1996) “Competiti<strong>on</strong> <strong>and</strong> corporate performance”, The Journal of Political<br />
Ec<strong>on</strong>omy, Vol 104, No 4, p724-746.<br />
Romer, Paul M. (1986) “Increasing Returns <strong>and</strong> L<strong>on</strong>g Run Growth”, The Journal of Political<br />
Ec<strong>on</strong>omy 94(5): 1002-1037.<br />
Romer, Paul M. (1990) “Endogenous Technological Change”, The Journal of Political Ec<strong>on</strong>omy<br />
98(5) part 2: 71-102.<br />
Solow, R M (1956) “A C<strong>on</strong>tributi<strong>on</strong> to the Theory of Ec<strong>on</strong>omic Growth”, Quarterly Journal of<br />
Ec<strong>on</strong>omics 70(1): 65-94.<br />
Solow,Robert (1997) Learning from ‘Learning by Doing’, Stanford: Stanford Univ. Press<br />
Stokey, Nancy L. (1988) “Learning by doing <strong>and</strong> the introducti<strong>on</strong> of new goods”, Journal of<br />
Political Ec<strong>on</strong>omy, Vol 96, No 4 (August), p701-171.<br />
Young, Alwyn (1991) “Learning by Ding <strong>and</strong> the Dynamic Effects of Internati<strong>on</strong>al Trade”, Quarterly<br />
Journal of Ec<strong>on</strong>omics 106(2): 369-406.<br />
Young, Alwyn (1993) “Inventi<strong>on</strong> <strong>and</strong> Bounded Learning by Doing”, Journal of Political Ec<strong>on</strong>omy<br />
101(3): 443-472.<br />
4.3 New Ec<strong>on</strong>omic Geography<br />
Davis, R<strong>on</strong>ald R. <strong>and</strong> David E.Weinstein (2002) B<strong>on</strong>es, Bombs, <strong>and</strong> Break Points: The<br />
Geography of Ec<strong>on</strong>omic Activity, 92(5), pp. 1269-1289.<br />
Fujita, Masahisa, Paul Krugman <strong>and</strong> Anth<strong>on</strong>y J. Venables (2001): The Spatial Ec<strong>on</strong>omy: Cities,<br />
Regi<strong>on</strong>s <strong>and</strong> Internati<strong>on</strong>al Trade, MIT Press.<br />
74<br />
292
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Fujita, Masahisa <strong>and</strong> Jacques-François Thisse, (2002) Agglomerati<strong>on</strong> <strong>and</strong> growth with migrati<strong>on</strong><br />
<strong>and</strong> knowledge externalities, mimeo, April.<br />
Kim, S. (1995) “Expansi<strong>on</strong> of markets <strong>and</strong> the geographic distributi<strong>on</strong> of ec<strong>on</strong>omic activities. The<br />
trends in US regi<strong>on</strong>al manufacturing structure: 1860-1987” Quarterly Journal of Ec<strong>on</strong>omics,<br />
110(4), pp. 881-908.<br />
Krugman, Paul Anth<strong>on</strong>y J. Venables (1995) “Globalizati<strong>on</strong> <strong>and</strong> the Inequality of Nati<strong>on</strong>s”<br />
Quarterly Journal of Ec<strong>on</strong>omics, 110(4), pp. 857-880.<br />
Krugman, Paul (1980) “Scale Ec<strong>on</strong>omies, Product Differentiati<strong>on</strong>, <strong>and</strong> the Pattern of Trade”<br />
American Ec<strong>on</strong>omic Review, 70(5), pp. 950-959.<br />
Krugman, Paul (1981) “Intraindustrial Specialisati<strong>on</strong> <strong>and</strong> the Gains from Trade” Journal of Political<br />
Ec<strong>on</strong>omy, 89(5), pp. 959-973.<br />
Krugman, Paul (1991) “Increasing Returns <strong>and</strong> Ec<strong>on</strong>omic Geography” Journal of Political<br />
Ec<strong>on</strong>omy, 99(3), pp. 483-499.<br />
Martin, R<strong>on</strong> <strong>and</strong> Peter Sunley (1996) “Paul Krugman’s Geographical Ec<strong>on</strong>omics <strong>and</strong> Its<br />
Implicati<strong>on</strong>s for Regi<strong>on</strong>al <strong>Development</strong> Theory. A Critical Assessment” Ec<strong>on</strong>omic Geography,<br />
72(3), pp. 259-292.<br />
Midelfart-Knarvik K.H., H.G. Overman, S.J. Redding <strong>and</strong> A.J. Venables, (2000) The Locati<strong>on</strong> of<br />
European Industry, European Commissi<strong>on</strong>, DG ECFIN, Ec<strong>on</strong>omic Papers No. 142, April.<br />
Neary, Peter (2001) “Of Hypes <strong>and</strong> Hyperbolas: Introducing the New Ec<strong>on</strong>omic Geography”<br />
Journal of Ec<strong>on</strong>omic Literature, 39, June, pp. 536-561.<br />
Overman, Henry G., Stephen Redding <strong>and</strong> Anth<strong>on</strong>y J. Venables (2001) “The Ec<strong>on</strong>omic<br />
Geography of Trade, Producti<strong>on</strong> <strong>and</strong> Income. A Survey of Empirics” L<strong>on</strong>d<strong>on</strong> School of<br />
Ec<strong>on</strong>omics, Centre for Ec<strong>on</strong>omic Performance, September.<br />
Robert-Nicoud, Frédéric (2002) “The structure of the simple ‘New Ec<strong>on</strong>omic Geography’ models”<br />
L<strong>on</strong>d<strong>on</strong> School of Ec<strong>on</strong>omic, Centre for Ec<strong>on</strong>omic Performance, August.<br />
Samuels<strong>on</strong>, Paul A. (1954) “The Transfer Problem <strong>and</strong> Transport Costs II: Analysis of Effects of<br />
Trade Impediments” The Ec<strong>on</strong>omic Journal, 64, pp. 264-289.<br />
4.4 Evoluti<strong>on</strong>ary Ec<strong>on</strong>omics<br />
Baumol,William (1993) Entrepreneurship, Management <strong>and</strong> the Structure of Payoff, Cambridge,Ma.<br />
The MIT Press Chapts. 1-7<br />
Bergek, Anna <strong>and</strong> Staffan Jacobss<strong>on</strong> (2002) The emergence of a growth industry: A comparative<br />
analysis of the German, Dutch <strong>and</strong> Swedish wind turbine industry, in Stan Metcalfe <strong>and</strong> Uwe<br />
Cantner: Change, transformati<strong>on</strong> <strong>and</strong> development, Springer pp. 197-228.<br />
Denicolò, Vincenzo <strong>and</strong> Zanchettin, Piercarlo (2002) “Competiti<strong>on</strong> <strong>and</strong> growth in a neo-<br />
Schumpeterian model”, mimeo Università di Bologna.<br />
75<br />
293
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Dosi, G (1988) “Sources, procedures, <strong>and</strong> microec<strong>on</strong>omic effects of innovati<strong>on</strong>”, Journal of<br />
Ec<strong>on</strong>omic Literature, Vol 26, No 3, p 1120-1171.<br />
Edquist, Charles (2001) The Systems of Innovati<strong>on</strong> Approach <strong>and</strong> Innovati<strong>on</strong> Policy: An account<br />
of the state of the art, lead paper presented at the DRUID C<strong>on</strong>ference, Aalborg, June 12-15,<br />
2001.<br />
Etro, Federico (2008), “Growth Leaders”, Journal of Macroec<strong>on</strong>omics 30, 1148-1172<br />
Fagerberg, J (2002) “A layman’s guide to evoluti<strong>on</strong>ary ec<strong>on</strong>omics”, mimeo, September 2002.<br />
Foster, J (2001) “Competiti<strong>on</strong>, competitive selecti<strong>on</strong> <strong>and</strong> ec<strong>on</strong>omic evoluti<strong>on</strong>”, in Garrouste, P <strong>and</strong><br />
Ioannides, S (eds) Evoluti<strong>on</strong> <strong>and</strong> Path Dependence in Ec<strong>on</strong>omic Ideas, L<strong>on</strong>d<strong>on</strong>: Routledge.<br />
Freeman (1987) <strong>Technology</strong>, policy <strong>and</strong> ec<strong>on</strong>omic performance: less<strong>on</strong>s from Japan,<br />
L<strong>on</strong>d<strong>on</strong>: Pinter.<br />
Freeman (1994) Critical survey: The ec<strong>on</strong>omics of technical change, Cambridge Journal of<br />
Ec<strong>on</strong>omic, 18, pp. 463-514.<br />
Hayek, F. A. "The Use of Knowledge in Society" American Ec<strong>on</strong>omic Review, 35,<br />
(September 1945), p 519-530.<br />
Hauknes, Johan (1999) “Innovati<strong>on</strong> systems <strong>and</strong> capabilities” STEP Working Paper, Oslo.<br />
Levin, R.C. – Klevorick, A.K. – Nels<strong>on</strong>, R.R. – Winter, S.G. (1987) Appropriating the returns from<br />
industrial research <strong>and</strong> development, Brookings Papers <strong>on</strong> Ec<strong>on</strong>omic Activity, 3, pp. 783-820.<br />
Liu, Xieling – White, Steven (2001) Comparing innovati<strong>on</strong> systems: A framework <strong>and</strong> applicati<strong>on</strong><br />
to China`s transiti<strong>on</strong>al c<strong>on</strong>text, Research Policy, 30, pp. 1091-1114.<br />
Lundvall, Bengt-Åke (1992) Nati<strong>on</strong>al systems of innovati<strong>on</strong>: Towards a theory of innovati<strong>on</strong> <strong>and</strong><br />
interactive learning, L<strong>on</strong>d<strong>on</strong>: Pinter.<br />
Metcalfe, J S (1995), “The Ec<strong>on</strong>omic Foundati<strong>on</strong>s of <strong>Technology</strong> Policy: Equilibrium <strong>and</strong><br />
Evoluti<strong>on</strong>ary Perspectives”, in Paul St<strong>on</strong>eman ed., H<strong>and</strong>book of the Ec<strong>on</strong>omics of Innovati<strong>on</strong> <strong>and</strong><br />
Technological Change, Oxford UK <strong>and</strong> Cambridge USA: Blackwell.<br />
Metcalfe, J S (1998) Evoluti<strong>on</strong>ary Ec<strong>on</strong>omics <strong>and</strong> Creative Destructi<strong>on</strong>, L<strong>on</strong>d<strong>on</strong>: Routledge.<br />
Nels<strong>on</strong>, Richard R. (1996) The sources of ec<strong>on</strong>omic growth, Harvard University Press, Chapter<br />
10: Nati<strong>on</strong>al innovati<strong>on</strong> systems: A retrospective <strong>on</strong> a study, pp. 274-301.<br />
Nels<strong>on</strong>, R <strong>and</strong> Winter, G (2002) “Evoluti<strong>on</strong>ary theorizing in ec<strong>on</strong>omics”, Journal of Ec<strong>on</strong>omic<br />
Perspectives, Vol 16, No 2, p 23-46.<br />
Nels<strong>on</strong>, R <strong>and</strong> Winter, S (1982) An Evoluti<strong>on</strong>ary Theory of Ec<strong>on</strong>omic Change, L<strong>on</strong>d<strong>on</strong>: Belknap<br />
Press.<br />
76<br />
294
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Pavitt, K (2002) “Knowledge about knowledge since Nels<strong>on</strong> & Winter: a mixed record”, SPRU<br />
Electr<strong>on</strong>ic Working Paper Series, No. 83.<br />
Saviotti, Pier Paolo (2000) Networks, nati<strong>on</strong>al innovati<strong>on</strong> systems <strong>and</strong> self-organisati<strong>on</strong>,<br />
University of Pierre Mendes-France, mimeo.<br />
Segerstrom, P., 2007. Intel ec<strong>on</strong>omics. Internati<strong>on</strong>al Ec<strong>on</strong>omic Review 48 (1), 247–280<br />
St<strong>on</strong>eman, Paul (ed) (1998) H<strong>and</strong>book of the Ec<strong>on</strong>omics of Innovati<strong>on</strong> <strong>and</strong> Technological<br />
Change, Oxford: Blackwell Publisher.<br />
Winter, S (1987) “Knowledge <strong>and</strong> competence as strategic assets” in Teece, D (ed) The<br />
Competitive Challenge, New York: Harper <strong>and</strong> Row.<br />
Zeira, J.(2003) Innovati<strong>on</strong>s, Patent Races <strong>and</strong> Endogenous Growth, mimeo. Harvard University.<br />
4.5 Cliometrics<br />
Atkens<strong>on</strong>, Andrew <strong>and</strong> Patrick Kehoe (2001) “The Transiti<strong>on</strong> to a New Ec<strong>on</strong>omy After the<br />
Sec<strong>on</strong>d Industrial Revoluti<strong>on</strong>”, NBER Working Paper, No. 8676, December<br />
Coase, R<strong>on</strong>ald (1960): The Problem of Social Cost, Journal of Law <strong>and</strong> Ec<strong>on</strong>omics, 3(1), pp. 1-<br />
44<br />
Coleman, D.C. <strong>and</strong> Christine MacLeod (1986) “Attitude to New Techniques: British Businessmen”<br />
1800-1950, Ec<strong>on</strong>omic History Review, 39(4), pp. 588-611<br />
C<strong>on</strong>rad, Alfred H. <strong>and</strong> John R. Meyer (1957) “Ec<strong>on</strong>omic Theory, Statistical Inference, <strong>and</strong><br />
Ec<strong>on</strong>omic History” Journal of Ec<strong>on</strong>omic History, 17(4), pp. 524-44<br />
C<strong>on</strong>rad, Alfred H. <strong>and</strong> John R. Meyer (1958) “The Ec<strong>on</strong>omics of Slavery in the Ante Bellum<br />
South” Journal of Political Ec<strong>on</strong>omy, 66(2), pp. 95-130.<br />
Crafts, N.F.R (1987) “Cliometrics, 1971-1986: a survey” Journal of Applied Ec<strong>on</strong>ometrics, 2(3),<br />
pp. 171-192<br />
Crafts, Nicholas (2002) “Productivity growth in the industrial revoluti<strong>on</strong>: A new growth accounting<br />
perspective”, mimeo, L<strong>on</strong>d<strong>on</strong> School of Ec<strong>on</strong>omics, January.<br />
Crafts, Nicholas F.R <strong>and</strong> C.K. Harley (1992) “Output growth <strong>and</strong> British industrial revoluti<strong>on</strong>: a<br />
restatement of the Crafts-Harley view” Ec<strong>on</strong>omic History Review, 45(4), pp. 703-730<br />
Crafts, Nicholas F.R <strong>and</strong> Terence C. Mills (1997) “Endogenous innovati<strong>on</strong>, trend growth, <strong>and</strong> the<br />
British industrial revoluti<strong>on</strong>: reply to Greasley <strong>and</strong> Oxley” Journal of Ec<strong>on</strong>omic History, 57(4), pp.<br />
950-956.<br />
Crafts, Nicholas F.R. (1995a) “Exogenous or Endogenous Growth? The Industrial Revoluti<strong>on</strong><br />
Rec<strong>on</strong>sidered” Journal of Ec<strong>on</strong>omic History, 55(4), pp. 745-772<br />
Crafts, Nicholas F.R. (1995b). “Macroinventi<strong>on</strong>s, Ec<strong>on</strong>omic Growth, <strong>and</strong> `Industrial Revoluti<strong>on</strong>` in<br />
Britain <strong>and</strong> France” Ec<strong>on</strong>omic History Review, 48(3), pp. 591-598.<br />
77<br />
295
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Crafts, Nicholas F.R. (1996) “The First Industrial Revoluti<strong>on</strong>: A Guided Tour for Growth<br />
Ec<strong>on</strong>omists” American Ec<strong>on</strong>omic Review, 86(2), pp. 197-201.<br />
Fischlow, Albert (1965) American Railroads <strong>and</strong> the Transformati<strong>on</strong> of the Ante-Bellum Ec<strong>on</strong>omy,<br />
Harvard University Press, Chapter 2.<br />
Fogel, Robert W. (1962): “A quantitative approach to the study of railroads in American ec<strong>on</strong>omic<br />
growth: a report of some preliminary findings” Journal of Ec<strong>on</strong>omic History, 22(2), pp. 163-197.<br />
Fogel, Robert W. (1964): Railroads <strong>and</strong> the American Ec<strong>on</strong>omic Growth: <str<strong>on</strong>g>Essays</str<strong>on</strong>g> in Ec<strong>on</strong>ometric<br />
History, Baltimore: Johns Hopkins University Press.<br />
Fogel, Robert W. (1965) “The reunificati<strong>on</strong> of ec<strong>on</strong>omic history with ec<strong>on</strong>omic theory” American<br />
Ec<strong>on</strong>omic Review, 55 (1/2), pp. 92-98.<br />
Fogel, Robert W. (1966a) “The new ec<strong>on</strong>omic history: its findings <strong>and</strong> methods” The Ec<strong>on</strong>omic<br />
History Review, 19 (3), pp. 642-656.<br />
Fogel, Robert W. (1966b) “Railroads as an analogy to the space effort: some ec<strong>on</strong>omic aspects”<br />
Ec<strong>on</strong>omic Journal, 76(301), pp. 16-43.<br />
Goldin, Claudia (1995) “Cliometrics <strong>and</strong> the Nobel” Journal of Ec<strong>on</strong>omic Perspectives, 9 (2), pp.<br />
191-208.<br />
Greasley, David <strong>and</strong> Les Oxley (1994) “Rehabilitati<strong>on</strong> sustained: the industrial revoluti<strong>on</strong> as a<br />
macroec<strong>on</strong>omic epoch” Ec<strong>on</strong>omic History Review, 47(4), pp. 760-768.<br />
Greasley, David <strong>and</strong> Les Oxley (1997a) “Endogenous growth or “Big Bang”: two views of the first<br />
industrial revoluti<strong>on</strong>” Journal of Ec<strong>on</strong>omic History, 57(4), pp. 935-949.<br />
Greasley, David <strong>and</strong> Les Oxley (1997b) “Endogenous growth, trend output, <strong>and</strong> the industrial<br />
revoluti<strong>on</strong>: reply to Crafts <strong>and</strong> Mills” Journal of Ec<strong>on</strong>omic History, 57(4), pp. 957-960.<br />
Gregers<strong>on</strong>, Brigitte <strong>and</strong> Björn Johns<strong>on</strong>, (1998) Nati<strong>on</strong>al Systems of Innovati<strong>on</strong> as a Framework<br />
for Innovati<strong>on</strong> Policy, The United Nati<strong>on</strong>s University, April.<br />
Heckman, James J. (1997) “The value of quantitative evidence <strong>on</strong> the effect of the past <strong>and</strong> the<br />
present” American Ec<strong>on</strong>omic Review, 87(2), pp. 404-408.<br />
J<strong>on</strong>es, Charles I. (1999) “Was an industrial revoluti<strong>on</strong> inevitable? Ec<strong>on</strong>omic growth over the very<br />
l<strong>on</strong>g run” NBER Working Paper, No. 7375, October.<br />
Khan, Zorina B. (1995) “Property rights <strong>and</strong> patent litigati<strong>on</strong> in early nineteenth-century America”<br />
Journal of Ec<strong>on</strong>omic History, 55(1), pp. 58-97.<br />
L<strong>and</strong>es, David (1994) “What room for accident in history? Explaining big changes by small<br />
events” Ec<strong>on</strong>omic History Review, 47(4), pp. 637-656.<br />
L<strong>and</strong>es,David (1998), The Wealth <strong>and</strong> Poverty of Nati<strong>on</strong>s, W.W.Nort<strong>on</strong>, New York<br />
78<br />
296
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Lorenz, Edward H. (1991) “An evoluti<strong>on</strong>ary explanati<strong>on</strong> for competitive decline: the British<br />
shipbuilding industry, 1890-1970” Journal of Ec<strong>on</strong>omic History, 51(4), pp. 911-935.<br />
MacLeod, Christine (1992) “Strategies for innovati<strong>on</strong>: the diffusi<strong>on</strong> of new technology in<br />
nineteenth-century British industry” Ec<strong>on</strong>omic History Review, 45(2), pp. 285-307.<br />
Maddis<strong>on</strong>, Angus (2001) The World Ec<strong>on</strong>omy: A Millennial Perspective, OECD, <strong>Development</strong><br />
Centre Studies.<br />
McCloskey, D<strong>on</strong>ald N. (1978) “The achievements of the cliometric school” Journal of Ec<strong>on</strong>omic<br />
History, 38(1), pp. 13-28<br />
Mokyr (1992) “Technological inertia in ec<strong>on</strong>omic history” Journal of Ec<strong>on</strong>omic History, 52(2), pp.<br />
325-338.<br />
Mokyr, Joel (1993) Introducti<strong>on</strong>: The New Ec<strong>on</strong>omic History <strong>and</strong> the Industrial Revoluti<strong>on</strong>, in Joel<br />
Mokyr (ed.): The British Industrial Revoluti<strong>on</strong>: An Ec<strong>on</strong>omic perspective, Oxford.<br />
Mokyr, Joel (1997) Inventi<strong>on</strong> <strong>and</strong> Rebelli<strong>on</strong>: Why Do innovati<strong>on</strong> Occur at All? An Evoluti<strong>on</strong>ary<br />
Approach, paper presented at the C<strong>on</strong>ference <strong>on</strong> Minorities <strong>and</strong> Ec<strong>on</strong>omic growth, Bar Ilan<br />
University, June 2-5 1997.<br />
Mokyr, Joel (1999a) Innovati<strong>on</strong> <strong>and</strong> Rebelli<strong>on</strong>: Why do Innovati<strong>on</strong>s Occur at all? An Evoluti<strong>on</strong>ary<br />
Approach, in Elise Brezis <strong>and</strong> Peter Temin, eds., Minorities <strong>and</strong> Ec<strong>on</strong>omic Growth. Amsterdam:<br />
Elsevier Publishers, pp. 179-203.<br />
Mokyr, Joel (1999b) Knowledge, <strong>Technology</strong>, <strong>and</strong> Industrial growth During the Industrial<br />
Revoluti<strong>on</strong>, Northwestern University, mimeo.<br />
North, Douglass C. (1968) “Sources of productivity change <strong>on</strong> ocean shipping, 1600-1850”<br />
Journal of Political Ec<strong>on</strong>omy, 76 (5), pp. 953-970.<br />
North, Douglass C. (1990) Instituti<strong>on</strong>s, instituti<strong>on</strong>al change, <strong>and</strong> ec<strong>on</strong>omic performance,<br />
Cambridge University Press.<br />
North, Douglass C. (1991) “Instituti<strong>on</strong>s” Journal of Ec<strong>on</strong>omic Perspectives, 5(1), pp. 97-112.<br />
North, Douglass C. (1994) “Ec<strong>on</strong>omic performance through time” American Ec<strong>on</strong>omic Review,<br />
84(3), pp. 359-368.<br />
North, Douglass C. (1997) “Cliometrics – 40 years later” American Ec<strong>on</strong>omic Review, 87(2), pp.<br />
412-414.<br />
North, Douglass (2005), Underst<strong>and</strong>ing the Process of Ec<strong>on</strong>omic Change, Princet<strong>on</strong> Univ. Press<br />
O’Rourke, Kevin H. <strong>and</strong> Jeffrey G. Williams<strong>on</strong>, (2002) “From Malthus to Ohlin, trade, growth <strong>and</strong><br />
distributi<strong>on</strong> since 1500” NBER Working Paper, No. 8955, May<br />
Ols<strong>on</strong>, Mancur (1996) “Big bills left <strong>on</strong> the sidewalk: why some nati<strong>on</strong>s are rich, <strong>and</strong> other poor”<br />
Journal of Ec<strong>on</strong>omic Perspectives, 10(2), pp. 3-24.<br />
Ols<strong>on</strong>,Mancur (1982),The Rise <strong>and</strong> Decline of Nati<strong>on</strong>s, Yale Univ. Press<br />
79<br />
297
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Pearce, Robert (2002): Nati<strong>on</strong>al systems of innovati<strong>on</strong> <strong>and</strong> the internati<strong>on</strong>al technology strategies<br />
of multinati<strong>on</strong>als, University of Reading.<br />
Phillips, William H. (1992) “Patent growth in the old domini<strong>on</strong>: the impact of railroad integrati<strong>on</strong><br />
before 1880” Journal of Ec<strong>on</strong>omic History, 52(2), pp. 389-400.<br />
Posner, Richard (1980) “A theory of primitive society, with special reference to the law” Journal of<br />
Law <strong>and</strong> Ec<strong>on</strong>omics, 23, pp. 1-54.<br />
Schoser, Christof (1999) “The instituti<strong>on</strong>s defining nati<strong>on</strong>al systems of innovati<strong>on</strong>: a new<br />
tax<strong>on</strong>omy to analyse the Impact of globalisati<strong>on</strong>”, University of Freiburg, Discussi<strong>on</strong> Paper.<br />
Shiue, Carol H (2002) “Transport costs <strong>and</strong> the geography of arbitrage in eighteenth-century<br />
China” American Ec<strong>on</strong>omic Review, 92 (5), pp. 1406-1419.<br />
Smith, Chauncey (1890) “A century of patent law” Quarterly Journal of Ec<strong>on</strong>omics, 5(1), pp. 44-<br />
69.<br />
Sokoloff, Kenneth L. (1988) “Incentive activity in early industrial America: evidence from patent<br />
records, 1790-1846” Journal of Ec<strong>on</strong>omic History, 48(4), pp. 813-850.<br />
Sullivan, Richard J. (1994) “Estimates of the value of patent rights in Great Britain <strong>and</strong> Irel<strong>and</strong>,<br />
1852-1876” Ec<strong>on</strong>omica, 61(241), pp. 37-58.<br />
Triplett, Jack E. (1999) “The Solow productivity paradox: what do computers do to productivity?”<br />
Canadian Journal of Ec<strong>on</strong>omics, 32(2), pp. 309-334.<br />
Wallis, John J. – North, Douglass C. (1986): Measuring the Transacti<strong>on</strong> Sector in the American<br />
Ec<strong>on</strong>omy, in Engerman, S.L. – Gallman, R.E. (eds), L<strong>on</strong>g-term factors in American ec<strong>on</strong>omic<br />
growth, University of Chicago Press, pp. 95-148.<br />
4.6 Strategic Management <strong>and</strong> Networks<br />
Audretsch, D, Baumol, W <strong>and</strong> Burke, A (2001) “Competiti<strong>on</strong> policy in dynamic markets”,<br />
Internati<strong>on</strong>al Journal of Industrial Organizati<strong>on</strong>, Vol 19, p 665-693.<br />
B<strong>on</strong>accorsi, Andrea <strong>and</strong> Paola Giuri (2000) “Industry life cycle <strong>and</strong> the evoluti<strong>on</strong> of an industry<br />
network” LEM Working Paper Series, No. 4, December<br />
Ellig, Jerry <strong>and</strong> Daniel Lin (2001): A tax<strong>on</strong>omy of dynamic competiti<strong>on</strong> theories, in Jerry Ellig (ed):<br />
Dynamic Competiti<strong>on</strong> <strong>and</strong> Public Policy: <strong>Technology</strong>, Innovati<strong>on</strong> <strong>and</strong> Antitrust Issues, Cambridge<br />
University Press.<br />
Evans, D <strong>and</strong> R. Schmalensee (2001) “Some ec<strong>on</strong>omic aspects of antitrust analysis in<br />
dynamically competitive industries“ NBER working paper, No.8268.<br />
Farrel, Joseph <strong>and</strong> Garth Sal<strong>on</strong>er, “Coordinati<strong>on</strong> through committees <strong>and</strong> markets”, RAND<br />
Journal of Ec<strong>on</strong>omics, Vol 16 (Spring), p70-83<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>,H.W. (2003), Ec<strong>on</strong>omies of Network Industries, L<strong>on</strong>d<strong>on</strong>: Routledge, 2003<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>,H.W. (2006), Innovati<strong>on</strong>, <strong>Technology</strong> <strong>and</strong> Hypercompetiti<strong>on</strong>, L<strong>on</strong>d<strong>on</strong>: Routledge,<br />
2006<br />
80<br />
298
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Grant, Robert M. (2002): C<strong>on</strong>temporary Strategy Analysis, 4th editi<strong>on</strong>, Blackwell Publishing<br />
Katz, Michael L. <strong>and</strong> Shapiro, Carl (1986) “Product compatibility choice in a market with<br />
technological progress”, Oxford Ec<strong>on</strong>omic Papers, Special Issue <strong>on</strong> the New Industrial<br />
Ec<strong>on</strong>omics, November.<br />
Katz, Michael L. <strong>and</strong> Shapiro, Carl (1994) “Systems competiti<strong>on</strong> <strong>and</strong> network effects”, Journal of<br />
Ec<strong>on</strong>omic Perspectives, Vol 8, No 2 (Spring), p93-115<br />
Klein, Jeremy (2002): Bey<strong>on</strong>d Competitive Advantage, Strategic Change, 11(6), pp. 317-326.<br />
Klepper, Steven (1996) “Entry, exit, growth <strong>and</strong> innovati<strong>on</strong> over the product life cycle” American<br />
Ec<strong>on</strong>omic Review, Vol. 86, No 3, pp. 562-583.<br />
Klepper, Steven <strong>and</strong> Kenneth Sim<strong>on</strong>s (2001), “Industry shakeouts <strong>and</strong> technological change”<br />
mimeo.<br />
Jovanovic, Boyan (1982) “Selecti<strong>on</strong> <strong>and</strong> the evoluti<strong>on</strong> of industry” Ec<strong>on</strong>ometrica, Vol. 50 No 3, pp.<br />
649-670.<br />
Le<strong>on</strong>ard-Bart<strong>on</strong>, D (1992) “Core capabilities <strong>and</strong> core rigidities: a paradox in managing new<br />
product development” Strategic Management Journal, Summer, pp. 111-125.<br />
Liebowitz, S.J. <strong>and</strong> Stephen E. Margolis (1994) “Network externality: an uncomm<strong>on</strong> tragedy”<br />
Journal of Ec<strong>on</strong>omic Perspectives, p133-150.<br />
Liebowitz, S.J. <strong>and</strong> Stephen E. Margolis, (1995) “Are network externalities a new source of<br />
market failure?” Research in Law <strong>and</strong> Ec<strong>on</strong>omics¸ Vol 17, p.41-46<br />
March, J.G. (1991) “Explorati<strong>on</strong> <strong>and</strong> exploitati<strong>on</strong> in organizati<strong>on</strong>al learning” Organizati<strong>on</strong>al<br />
Science, 2, pp. 71-87.<br />
Porter, Michael (1990): The Competitive Advantage of Nati<strong>on</strong>s, The Free Press, New York<br />
Porter, Michael E (1985): Competitive Advantage, The Free Press, New York<br />
Posner, R (2000) “Antitrust in the new ec<strong>on</strong>omy“ John M. Olin Law <strong>and</strong> Ec<strong>on</strong>omics Working<br />
Paper, No 106, University of Chicago.<br />
Shapiro, C <strong>and</strong> H.R. Varian (1999) Informati<strong>on</strong> Rules: A Strategic Guide to the Network<br />
Ec<strong>on</strong>omy, Harvard Business School Press, Bost<strong>on</strong>.<br />
Sim<strong>on</strong>s, Kenneth (2001) “Product market characteristics <strong>and</strong> the industry life cycle”, mimeo.<br />
Tushman, M L <strong>and</strong> E. Romanelli (1994) “Organizati<strong>on</strong>al transformati<strong>on</strong> as punctuated equilibrium:<br />
an empirical test” Academy of Management Journal, 37, pp. 1141-1166.<br />
81<br />
299
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Wecker, Claudia (2003) “Market performance <strong>and</strong> competiti<strong>on</strong>: a product life cycle model”,<br />
Technovati<strong>on</strong>, Vol 23, No 4, pp. 281-290.<br />
Comprehensive treatments of Reviewer(<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>)<br />
Ec<strong>on</strong>omies of Network Industries, Routledge : L<strong>on</strong>d<strong>on</strong>, New York 2003<br />
Innovati<strong>on</strong>, <strong>Technology</strong> <strong>and</strong> Hyper-Competiti<strong>on</strong> Routledge, L<strong>on</strong>d<strong>on</strong>, New York 2006<br />
Strategic Ec<strong>on</strong>omics of Network Industries, NovaScience, New York 2009<br />
Strategic Alliances in Biotech <strong>and</strong> Pharmaceuticals, NovaScience, New York 2010 (with F.Floether<br />
<strong>and</strong> C.Umali)<br />
Strategies of Ec<strong>on</strong>omic Growth <strong>and</strong> Catch-Up, NovaScience, New York 2012 (with M. Goosen, <strong>and</strong><br />
M.Takashima)<br />
82<br />
300
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
HANS W. GOTTINGER<br />
301
HANS W. GOTTINGER<br />
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
302
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
HANS W. GOTTINGER<br />
303
HANS W. GOTTINGER<br />
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
304
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
HANS W. GOTTINGER<br />
305
HANS W. GOTTINGER<br />
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
306
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
HANS W. GOTTINGER<br />
307
HANS W. GOTTINGER<br />
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
308
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
HANS W. GOTTINGER<br />
309
HANS W. GOTTINGER<br />
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
310
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
HANS W. GOTTINGER<br />
311
HANS W. GOTTINGER<br />
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
312
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
HANS W. GOTTINGER<br />
313
HANS W. GOTTINGER<br />
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
314
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
CATCH‐UP AND CONVERGENCE:<br />
MECHANISM DESIGN FOR<br />
ECONOMIC DEVELOPMENT*<br />
*Paper presented at the 1st Internati<strong>on</strong>al C<strong>on</strong>gress Jean Baptiste Say, Research Network <strong>on</strong> Innovati<strong>on</strong>,<br />
Boulogne‐sur‐Mer, France, August 27‐30, 2014<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
STRATEC Munich, Germany<br />
gottingerhans@gmail.com<br />
Abstract<br />
This paper identifies factors that promote l<strong>on</strong>g‐run sustainable ec<strong>on</strong>omic growth <strong>and</strong> industrial catchup<br />
in achieving superior performance. An informati<strong>on</strong>ally efficient mechanism design of the Hayek‐<br />
Hurwicz type would c<strong>on</strong>tain as essential elements allocative efficiency with a limited state role,<br />
fostering of private entrepreneurship for high intensity dynamic competiti<strong>on</strong> through industry<br />
specific ‘technological racing’ within a supporting market structure, free trade <strong>and</strong> intellectual<br />
property rights (IPRs), flexible labour markets, low taxes <strong>on</strong> labour, capital income <strong>and</strong> profits.<br />
Emphasis <strong>on</strong> high value‐added network industries directed toward increasing returns. If resource<br />
endowment am<strong>on</strong>g countries are similar these factors would eventually lead to c<strong>on</strong>vergence through<br />
catch‐up performance, otherwise they would lead to divergence.<br />
Key words<br />
Catch‐Up; C<strong>on</strong>vergence; Mechanism Design; Ec<strong>on</strong>omic <strong>Development</strong><br />
Taking a philosophical point of view, this may be seen as the mechanics for the<br />
implementati<strong>on</strong> of Adam Smith’s invisible h<strong>and</strong>: despite private informati<strong>on</strong> <strong>and</strong> pure selfish<br />
behavior, social welfare is achieved. All the field of Mechanism Design is just a generalizati<strong>on</strong><br />
of this possibility. – Noam Nisan (2007)<br />
INTRODUCTION<br />
What are the major factors for ec<strong>on</strong>omies to succeed <strong>on</strong> catching‐up in terms of GDP<br />
or GDP per capita (of purchasing power parity PPP) as an indicator of prosperity?<br />
As has been broadly covered in the ec<strong>on</strong>omic growth <strong>and</strong> development literature<br />
(<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g> & Goosen, 2012), there have been c<strong>on</strong>vergence theories <strong>on</strong> advanced<br />
(developed) ec<strong>on</strong>omies that have been partially verified for some OECD ec<strong>on</strong>omies,<br />
67<br />
315
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
but equally there have been observati<strong>on</strong>s of divergence am<strong>on</strong>g some developed <strong>and</strong><br />
developing (transiti<strong>on</strong>al) ec<strong>on</strong>omies that appear to show a growing gap. <strong>Technology</strong><br />
adopti<strong>on</strong> within regi<strong>on</strong>s of a nati<strong>on</strong>al ec<strong>on</strong>omy (say , China) were a major factor of<br />
less developed regi<strong>on</strong>s to catch up. Catch‐up metrics could also be identified in<br />
disaggregated form next to aggregates such as GDP (total), GDP per head, Human<br />
<strong>Development</strong> Index (HDI), for example, industry/technology index, infrastructure<br />
indicator (transportati<strong>on</strong>, communicati<strong>on</strong> networks), R&D expenditure<br />
(government, private industry) <strong>and</strong> competitiveness (global market share of key<br />
industries). A process of catching up induced by industrial races may tend to<br />
c<strong>on</strong>verge over time within a bloc of similar countries if technological <strong>and</strong><br />
educati<strong>on</strong>al endowment is similar as covered by Abramovitz’ (1986) famous ‘social<br />
capabilities’, as key industries engage in more incremental <strong>and</strong> complementary<br />
innovati<strong>on</strong> that through internati<strong>on</strong>al trade <strong>and</strong> foreign direct investment (FDI)<br />
spread to emerging industries in likewise developing ec<strong>on</strong>omies. In the post World<br />
War II history what was the mechanism that induced Japan to be <strong>on</strong> a path of catchup<br />
growth in the 1960s <strong>and</strong> 1970s in terms of per capita income growth? As<br />
suggested in the Asian miracle (Krugman, 1994), <strong>on</strong>e factor was input growth in key<br />
value‐added industries through capital expansi<strong>on</strong>, the other cumulative<br />
technological advancement through largely incremental innovati<strong>on</strong> leading to<br />
superiority in some industries. But as so<strong>on</strong> as the technological fr<strong>on</strong>tier was close to<br />
be reached it became increasingly difficult to dominate the market. Also in a wide<br />
array of high‐tech markets it became increasingly decisive to have integrative<br />
technologies to catch complementary <strong>and</strong> increasing returns markets.<br />
The paper shows that catch‐up processes should be primarily understood as<br />
technological races, establishing new industries that allow free flow of informati<strong>on</strong><br />
through entrepreneurial activity <strong>and</strong> innovati<strong>on</strong>. Under these circumstances, it is<br />
more likely that an ec<strong>on</strong>omic mechanism design generating more informati<strong>on</strong>,<br />
choices, ec<strong>on</strong>omic freedom <strong>and</strong> market transparency supported by democratic<br />
instituti<strong>on</strong>s, will have a better ec<strong>on</strong>omic performance record, in a sustainably l<strong>on</strong>g<br />
run, than any other that fails <strong>on</strong> the informati<strong>on</strong> distributi<strong>on</strong> <strong>and</strong> strategic incentive<br />
side, which would coincide with various types of socialist systems.<br />
The c<strong>on</strong>tent of the paper is as follows. Secti<strong>on</strong> 2 gives a brief historical outline <strong>on</strong> the<br />
interrelati<strong>on</strong> between ec<strong>on</strong>omic development <strong>and</strong> catching‐up paths. Secti<strong>on</strong> 3<br />
sketches the basic structure of algorithmic mechanism design (AMD) as it pertains to<br />
computati<strong>on</strong>al ec<strong>on</strong>omic systems, in particular to superior performance of<br />
decentralized (distributed) dynamic systems. Secti<strong>on</strong> 4 shows the c<strong>on</strong>necti<strong>on</strong><br />
between industrial competiti<strong>on</strong> <strong>and</strong> ist aggregati<strong>on</strong> to macro competiti<strong>on</strong> <strong>on</strong> a<br />
nati<strong>on</strong>al or regi<strong>on</strong>al level – in view of technological racing. Secti<strong>on</strong> 5 identifies<br />
technological fr<strong>on</strong>tiers <strong>on</strong> the firm level (FTF) as embedded <strong>on</strong> the industry level<br />
(ITF), amenable to a statistical profiling of technological evoluti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong>. It<br />
shows some rules of technological race behaviour resulting in statistical indicators<br />
68 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
316
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
<strong>on</strong> an industry level. Sec. 6 looks at industrial (in)efficiencies in catch‐up countries as<br />
benchmarked against the industrial leader. C<strong>on</strong>clusi<strong>on</strong>s follow <strong>and</strong> open problems<br />
are discussed in the last secti<strong>on</strong>.<br />
A BRIEF HISTORICAL REVIEW<br />
In a historical c<strong>on</strong>text, a catch‐up process becomes multidimensi<strong>on</strong>al, emerging with<br />
Engl<strong>and</strong>, moving to the US <strong>and</strong> a few European ec<strong>on</strong>omies, Germany, France, later<br />
to Japan <strong>and</strong> the newly industrialized ec<strong>on</strong>omies (NIEs) in Asia overlapping the<br />
BRICs. The argument is that due to expansi<strong>on</strong> of trade through globalizati<strong>on</strong> the<br />
catch‐up process is multidimensi<strong>on</strong>al <strong>and</strong> multispeed (Wan, 2004). Even for a few<br />
increasing in overall speed the benchmark of catch‐up keeps shifting from <strong>on</strong>e or a<br />
few to many more. For example, in the past WW II period, industrial <strong>and</strong> ec<strong>on</strong>omic<br />
growth of Japan was driven toward the US ec<strong>on</strong>omy less so toward Europe; the<br />
pattern of growth <strong>and</strong> catch‐up of China hits many more emerging ec<strong>on</strong>omies with<br />
significant potentials. We have modelled this by a more complex differential game as<br />
in <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g> <strong>and</strong> Goosen (2012).<br />
As a precursor of ec<strong>on</strong>omic mechanism design, from an ec<strong>on</strong>omic history<br />
perspective, we first identify Gershenkr<strong>on</strong> (1962) who emphasized state acti<strong>on</strong> <strong>on</strong><br />
industrializati<strong>on</strong> setting free guided entrepreneurial activities through targeted<br />
industrial policy (al<strong>on</strong>g the line of Japan <strong>and</strong> later South Korea, Taiwan) (Lee, 2013).<br />
This compares to Abramovitz’ (1986) <strong>and</strong> Abramovitz <strong>and</strong> David’s (1992) catching<br />
up process where self‐reinforcing, industry specific, competitive market forces,<br />
rather than state acti<strong>on</strong>, would initiate <strong>and</strong> sustain a technological race resulting in<br />
leadership positi<strong>on</strong>s across a range of industries. Here the roots of growth rest in the<br />
microec<strong>on</strong>omic industry structure <strong>and</strong> new industry creating high technology<br />
entrepreneurship as c<strong>on</strong>cepti<strong>on</strong>ally <strong>and</strong> empirically explored by Scherer (1992, 1999).<br />
The World Bank in the past takes a middle ground, propagating the state’s role to<br />
develop social capital which then by itself creates social capability to induce an<br />
upward potential through market forces. A more activistic role has recently been<br />
favoured by Justus Lin (2013), the former Chinese Chief Ec<strong>on</strong>omist of the World<br />
Bank. Amsden’s development statism is a further extensi<strong>on</strong> of Gershenkr<strong>on</strong>,<br />
catching‐up as a process of learning how to compete through ’Late Industrializati<strong>on</strong>’<br />
(Amsden, 1989). Other than the main catchup processes listed here have been<br />
reviewed by Burkett <strong>and</strong> Hart‐L<strong>and</strong>sberg (2003), <strong>and</strong> Fagerberg <strong>and</strong> Godinho (2004).<br />
In ’Achieving Rapid Growth’ Sachs <strong>and</strong> Warner (1996), in comparing growing<br />
middle income countries identify major factors such as allocative efficiency (with<br />
government interacti<strong>on</strong> low, markedly less regulati<strong>on</strong>), high degree of competiti<strong>on</strong>,<br />
free trade. flexible labour markets, low taxes <strong>on</strong> labour, capital income <strong>and</strong> profits.<br />
While c<strong>on</strong>vergence am<strong>on</strong>g developing ec<strong>on</strong>omies may be facilitated by becoming<br />
69<br />
317
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
more alike in terms of sources of growth such as technological progress, innovati<strong>on</strong>,<br />
levels of physical capital, labour productivity, quality of human capital, extent of<br />
trade openess, industrial structure <strong>and</strong> instituti<strong>on</strong>al framework, <strong>on</strong>e clear<br />
distinguishing factor may involve increasing returns mechanisms (IRMs). IRMs<br />
eliminate any kind of c<strong>on</strong>vergence <strong>and</strong> allow for <strong>on</strong>going quasi‐linear growth.<br />
Anecdotal observati<strong>on</strong>s as put forward by Easterly <strong>and</strong> Levine (2002) can be<br />
summarized as stylized facts <strong>on</strong> growth mechanisms.<br />
(1) Factor accumulati<strong>on</strong> does not account for growth differences but total factor<br />
productivity does for a substantial amount;<br />
(2) There are hugely growing differences in GDP per capita. On a global scale<br />
divergence <strong>and</strong> not c<strong>on</strong>diti<strong>on</strong>al c<strong>on</strong>vergence is the major c<strong>on</strong>cern in<br />
development policy; <strong>and</strong><br />
(3) Am<strong>on</strong>g developing ec<strong>on</strong>omies growth is not persistent over time.<br />
If movable through market forces all factors of producti<strong>on</strong> flow to the same places<br />
suggesting important externalities.<br />
A SIMPLE MECHANISM DESIGN FOR CATCHUP AND<br />
DEVELOPMENT<br />
A simple mechanism design emanates from the following paradigmatic situati<strong>on</strong>:<br />
Let there be k agents in an (Internet) ec<strong>on</strong>omy that collectively generate dem<strong>and</strong><br />
competing for resources from a supplier. The supplier herself announces prices<br />
entering a bulletin board accessible to all agents (as a sort of transparent market<br />
instituti<strong>on</strong>). In a simple form of a trading process we could exhibit a “tat<strong>on</strong>nement<br />
process” <strong>on</strong> a graph where the agents set up a dem<strong>and</strong> to the suppliers who<br />
advertise prices <strong>on</strong> a bulletin board which are c<strong>on</strong>verted to new prices in interacti<strong>on</strong><br />
with the agents.<br />
The tat<strong>on</strong>nement process in ec<strong>on</strong>omics is a simple form of an algorithmic<br />
mechanism design (AMD) (Nisan & R<strong>on</strong>en, 2001), which in modern computer<br />
science (CS) emerges as an offspring to algorithmic game theory (Nisan et al, 2007).<br />
Algorithmic ingredients apply to rati<strong>on</strong>al <strong>and</strong> selfish agents having well defined<br />
utility functi<strong>on</strong>s repesenting preferences over possible outputs of the algorithm. A<br />
payment ingredient motivates the agents. Mechanism Design Theory (MDT) aims to<br />
show how privately known preferences of the entire populati<strong>on</strong> can be aggregated<br />
towards a ’social choice’ that drives the mechanism of the whole ec<strong>on</strong>omy.<br />
(i) Each agent or group of agents have some some private input represented<br />
by its type. Its type is embedded in public knowledge or resource<br />
endowment as the ’social envir<strong>on</strong>ment’;<br />
(ii) There is a producti<strong>on</strong> functi<strong>on</strong> (or output specificati<strong>on</strong>) that associates<br />
each type t = t 1 , ..., t n with a set of socially allowable outputs (producti<strong>on</strong>s)<br />
oO; <strong>and</strong><br />
70 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
318
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
(iii)<br />
Each agent’s preferences are codified as real‐valued utility functi<strong>on</strong>s noted<br />
by u i (t i ,o).<br />
The universal functi<strong>on</strong> is specified in linear terms as u i = r i + v i (t i ,o) where r i 0 is the<br />
initial endowment (or resource), or the agent’s wealth, which the agent attempts to<br />
optimize.<br />
If we assume that the mechanism is truthful to the extent that the agents report their<br />
real type then truth‐telling is the <strong>on</strong>ly dominant strategy. This is c<strong>on</strong>sidered a<br />
truthful implementati<strong>on</strong>. The societal objective functi<strong>on</strong> is simply the aggregati<strong>on</strong> of<br />
all agents valuati<strong>on</strong>s. A ’maximizing mechanism design process’ (MDP) is called<br />
utilitarian or ’Hayekian’ if its objective functi<strong>on</strong> is the sum of all agents’ utilities.<br />
In the c<strong>on</strong>text of the internet ec<strong>on</strong>omy an MDP would enable users of network<br />
applicati<strong>on</strong>s to present their ’quality of service’ dem<strong>and</strong>s via utility functi<strong>on</strong>s<br />
defining the system performance requirements (<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, 2013). The resource<br />
allocati<strong>on</strong> process involves ec<strong>on</strong>omic actors to perform ec<strong>on</strong>omic optimizati<strong>on</strong> given<br />
scheduling policies, load balancing <strong>and</strong> service provisi<strong>on</strong>ing.<br />
Distributed algorithmic mechanism design (DMD) for internet resource allocati<strong>on</strong> in<br />
distributed systems is akin to an equilibrium c<strong>on</strong>verging market based ec<strong>on</strong>omy<br />
where selfish agents maximize utility <strong>and</strong> firms seek to maximize profits <strong>and</strong> the<br />
state keeps an ec<strong>on</strong>omic order providing basic public goods <strong>and</strong> public safety<br />
(Feigenbaum et al, 2007). A distributed algorithmic mechanism design thus c<strong>on</strong>sists<br />
of three comp<strong>on</strong>ents: a feasible strategy space at the network nodes for each agent<br />
(or aut<strong>on</strong>omous system), an aggregated outcome functi<strong>on</strong> computed by the<br />
mechanism <strong>and</strong> a set of multi‐agent prescribed strategies induced by the<br />
mechanism.<br />
For such DMDP in place an internet ec<strong>on</strong>omy can be shown computati<strong>on</strong>ally <strong>and</strong><br />
informati<strong>on</strong>ally efficient in the sense of Hurwicz <strong>and</strong> Reiter (2006), corroborated<br />
from a CS view by C<strong>on</strong>itzer <strong>and</strong> S<strong>and</strong>holm (2002). Furthermore, for efficient macromanagement<br />
it would satisfy ’moral hazard’ <strong>and</strong> incentive based c<strong>on</strong>cerns, <strong>and</strong> in<br />
view of ’informati<strong>on</strong>al c<strong>on</strong>straints’, e.g. adverse selecti<strong>on</strong>, it may also be superior in<br />
performance since in a market based decentralized capitalistic system due to keen<br />
competiti<strong>on</strong> am<strong>on</strong>g operating managers ‐ there will be more of higher type<br />
performing managers keeping truthful payoff‐relevant informati<strong>on</strong> than in a<br />
comparative socialist planning system with plenty of lower type performing<br />
managers providing lower quality or less payoff relevant informati<strong>on</strong>.<br />
A distributed algorithmic mechanism design (Feigenbaum et al, 2007) being<br />
computati<strong>on</strong>ally efficient in a large decentralized internet ec<strong>on</strong>omy is a powerful<br />
paradigm to substantiate claims by Hayek (1945) that an industrialized ec<strong>on</strong>omy<br />
71<br />
319
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
based <strong>on</strong> market principles has an overall better performance than socialist type<br />
ec<strong>on</strong>omies of a similar nature <strong>and</strong> scale. It is a paradigm that even c<strong>on</strong>temporary<br />
theorists of MDT seem to have partially missed (Myers<strong>on</strong>, 2008) <strong>and</strong> that puts<br />
historically the socialist planning debate in a new light which ir<strong>on</strong>ically, by some<br />
proposals, has been c<strong>on</strong>ducted <strong>on</strong> the basis of computati<strong>on</strong>al feasibility <strong>and</strong><br />
superiority.<br />
Such a Hayekian MDP also extends to a dynamic real ec<strong>on</strong>omy which also invokes<br />
highly desirable properties of incentive structures (Myers<strong>on</strong>, 2008) <strong>and</strong> knowledge<br />
creati<strong>on</strong> through hi‐tech entrepreneurship. This suggests that an MDP of the<br />
Hayekian‐Hurwicz type should be more likely to generate a l<strong>on</strong>g‐run sustainable<br />
growth <strong>and</strong> development process with comparative greater welfare benefits than<br />
what any socialist type planning could achieve. This is in compliance with Hayekian<br />
development ideas as put forward recently by Easterly (2013).<br />
The focus of this paper would be to explore growth <strong>and</strong> development generating<br />
structures <strong>and</strong> factors that are compatible with a Hayek‐Hurwicz design scheme for<br />
the developing world.<br />
INDUSTRIAL AND MACRO COMPETITION<br />
The striking pattern that emerges in the innovative activities of firms is their rivalries<br />
for a technological leadership positi<strong>on</strong> in situati<strong>on</strong>s that are best described as races<br />
or hypercompetiti<strong>on</strong> (Harris & Vickers, 1987; <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, 2006b). A race is an<br />
interactive pattern characterized by firms or nati<strong>on</strong>s c<strong>on</strong>stantly trying to get ahead of<br />
their rivals, or trying not to fall too far behind. In high‐technology industries, where<br />
customers are willing to pay a premium for advanced technology, leadership<br />
translates into increasing returns in the market through positive network<br />
externalities. Abramovitz (1986), in explaining the catch‐up hypothesis, lays stress<br />
<strong>on</strong> a country’s social capability in terms of years of educati<strong>on</strong> as a proxy of technical<br />
competence <strong>and</strong> its instituti<strong>on</strong>s. Competing behaviour is also a dynamic story of<br />
how technology unfolds in an industry. In c<strong>on</strong>trast to any existing way of looking at<br />
the evoluti<strong>on</strong> of technology, racing behaviour, though in character more ’a<br />
productivity race than a runner’s race (Abramovitz & David, 1997), recognizes the<br />
fundamental importance of strategic interacti<strong>on</strong>s between competing firms. Thus<br />
firms take their rivals’ acti<strong>on</strong>s into account when formulating their own decisi<strong>on</strong>s.<br />
The importance of this characterizati<strong>on</strong> is at least twofold. At <strong>on</strong>e level, racing<br />
behaviour has implicati<strong>on</strong>s for appreciating technology strategy at the level of the<br />
individual firm; at the other level, for underst<strong>and</strong>ing the impact of policies that aim<br />
to spur technological innovati<strong>on</strong> in an industry or country.<br />
On a nati<strong>on</strong>al scale, simple catch‐up hypotheses have put emphasis <strong>on</strong> the great<br />
potential of adopting unexploited technology in the early stage <strong>and</strong> the increase of<br />
self‐limiting power in the later stage. However, the actual growth path of the<br />
72 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
320
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
technological trajectory of a specific ec<strong>on</strong>omy may be overwhelmingly c<strong>on</strong>strained<br />
by social capability. The capability endogenously changes as states of the ec<strong>on</strong>omy<br />
<strong>and</strong> technology evolve. The success of ec<strong>on</strong>omic growth due to diffusi<strong>on</strong> of<br />
advanced technology or the possibility of leapfrogging is mainly attributable to how<br />
the social capability evolves. In other words, which effects become more influential:<br />
growing resp<strong>on</strong>siveness to competiti<strong>on</strong> or growing obstacles to it <strong>on</strong> account of<br />
vested interests <strong>and</strong> established positi<strong>on</strong>s?<br />
Observati<strong>on</strong>s <strong>on</strong> industrial patterns in Europe, the US or Asia point to which type of<br />
racing behaviour is prevalent in global high‐ technology industries. The pattern<br />
evolving from such c<strong>on</strong>duct could be benchmarked against the fr<strong>on</strong>tier pursuit type<br />
of the global technological leaders. Another observati<strong>on</strong> relates to policy inferences<br />
<strong>on</strong> market structure, entrepreneurship, innovati<strong>on</strong> activity, industrial policy <strong>and</strong><br />
regulatory frameworks in promoting <strong>and</strong> hindering industry fr<strong>on</strong>tier races in a<br />
global industrial c<strong>on</strong>text. Does lagging behind <strong>on</strong>e’s closest technological rivals’<br />
cause a firm to increase its innovative effort? The term ‘race’ suggests that no single<br />
firm would want to fall too far behind, <strong>and</strong> that every<strong>on</strong>e would like to get ahead. If<br />
a company tries to innovate more when it is behind than when it is ahead, then<br />
‘catch‐up’ behaviour will be the dominant effect. Once a firm gets far enough ahead<br />
of its rivals, then the latter will step up their efforts to get closer. The leading<br />
company will slow down its innovative efforts until its competitors have drawn<br />
uncomfortably close or have surpassed it. Of course, the process of getting closer to<br />
may be much easier than surpassing the rival This process repeats itself every time a<br />
firm gets far enough ahead of its rivals. Of course, catch‐up may <strong>on</strong>ly c<strong>on</strong>sistently<br />
apply to the next rivals but will not impact the leader. This is called ’persistent<br />
leadership’. On a nati<strong>on</strong>al level catchup processes like this may not lead to<br />
c<strong>on</strong>vergence.<br />
An alternative behaviour pattern would corresp<strong>on</strong>d to a business increasing its<br />
innovative effort if it gets far enough ahead, thus making catch‐up by the lagging<br />
companies increasingly difficult. For any of these businesses there appears to be a<br />
clear link to market <strong>and</strong> industry structure, as termed ‘intensity of rivalry’.<br />
We investigated two different kinds of races: <strong>on</strong>e that is a fr<strong>on</strong>tier race between itself<br />
<strong>and</strong> the technological leader at any point in time (‘fr<strong>on</strong>tier‐ sticking’ behaviour), or it<br />
might try to actually usurp the positi<strong>on</strong> of the leader by ‘leapfrogging’ it. When there<br />
are disproporti<strong>on</strong>ately large payoffs to being in the technical lead (relative to the<br />
payoffs that a firm can realize if it is simply close enough to the technical fr<strong>on</strong>tier),<br />
then <strong>on</strong>e would expect that leapfrogging behaviour would occur more frequently<br />
than fr<strong>on</strong>tier‐sticking behaviour. Alternatively, racing toward the fr<strong>on</strong>tier creates the<br />
reputati<strong>on</strong> of being an innovati<strong>on</strong> leader hoping to maintain <strong>and</strong> increase market<br />
73<br />
321
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
share in the future. All attempts to leapfrog the current technological leader might<br />
not be successful since many lagging firms might be attempting to leapfrog the<br />
leader simultaneously <strong>and</strong> the leader might be trying to get further ahead<br />
simultaneously. Corresp<strong>on</strong>dingly, <strong>on</strong>e could distinguish between attempted<br />
leapfrogging <strong>and</strong> realized leapfrogging.<br />
Am<strong>on</strong>g the key issues to be addressed is the apparent inability of technologyoriented<br />
corporati<strong>on</strong>s to maintain leadership in fields that they pi<strong>on</strong>eered. There is a<br />
presumpti<strong>on</strong> that firms fail to remain competitive because of agency problems or<br />
other suboptimal managerial behaviour within these organizati<strong>on</strong>s. An alternative<br />
explanati<strong>on</strong> is that technologically trailing firms, in symmetric competitive<br />
situati<strong>on</strong>s, will devote greater effort to innovati<strong>on</strong>, so that a failure of technological<br />
leaders to maintain their positi<strong>on</strong> is an appropriate resp<strong>on</strong>se to the competitive<br />
envir<strong>on</strong>ment. In asymmetric situati<strong>on</strong>s, with entrants challenging incumbents,<br />
research does dem<strong>on</strong>strate that startup firms show a str<strong>on</strong>ger endeavor to close up<br />
to or leapfrog the competitors. Such issues highlight the dynamics of the race within<br />
the given market structure in any of the areas c<strong>on</strong>cerned.<br />
We observe two different kinds of market asymmetries with bearing <strong>on</strong> racing<br />
behaviour: risk‐driven <strong>and</strong> resource‐based. When the incumbents’ profits are large<br />
enough <strong>and</strong> do not vary much with the product characteristics, the entrant is likely<br />
to choose the faster opti<strong>on</strong> in each stage as l<strong>on</strong>g as he has not fallen behind in the<br />
c<strong>on</strong>test. In view of resource‐based asymmetries, as a firm’s stage resource<br />
endowment increases, it could use the additi<strong>on</strong>al resources to either choose more<br />
aggressive targets or to attempt to finish the stage so<strong>on</strong>er, or both. Previous work<br />
has suggested that a firm that surges ahead of its rival increases its investment in<br />
R&D <strong>and</strong> speeds up, while a lagging firm reduces its investment <strong>and</strong> slows down.<br />
C<strong>on</strong>sequently, preceding effort suggests that the lead c<strong>on</strong>tinues to increase.<br />
However, based <strong>on</strong> related work for the US <strong>and</strong> Japanese telecommunicati<strong>on</strong>s<br />
industry when duopolistic <strong>and</strong> m<strong>on</strong>opolistic competiti<strong>on</strong> <strong>and</strong> product system<br />
complexity for new products are accounted for, the speeding up of a leading firm<br />
occurs <strong>on</strong>ly under rare circumstances. For example, a company getting far enough<br />
ahead such that the (temporary) m<strong>on</strong>opoly term dominates its payoff expressi<strong>on</strong>,<br />
will always choose the fast strategy, while a company that gets far enough behind<br />
will always choose the aggressive approach. Under these c<strong>on</strong>diti<strong>on</strong>s, the lead is<br />
likely to c<strong>on</strong>tinue to increase. If, <strong>on</strong> the other h<strong>and</strong>, both m<strong>on</strong>opoly <strong>and</strong> duopoly<br />
profits increase substantially with increased aggressiveness then even large leads<br />
can vanish with significant probability.<br />
Overall, this characterizati<strong>on</strong> highlights two forces that influence a firm’s choices in<br />
the various stages: proximity to the finish line <strong>and</strong> distance between the firms. This<br />
probability of reaping m<strong>on</strong>opoly profits is higher the farther ahead a firm is of its<br />
rival <strong>and</strong> even more so the closer the firm is to the finish line. If the lead company is<br />
74 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
322
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
far from the finish line, even a sizeable lead may not translate into the dominance of<br />
the m<strong>on</strong>opoly profit term, since there is plenty of time for the lead situati<strong>on</strong> to be<br />
reversed, <strong>and</strong> failure to finish first remains a probable outcome. In c<strong>on</strong>trast, the<br />
probability that the lagging company will get to be a m<strong>on</strong>opolist becomes smaller as<br />
it falls behind the leader. This raises the following questi<strong>on</strong>: what kind of acti<strong>on</strong>s<br />
cause a firm to get ahead? Intuitively, <strong>on</strong>e would expect that a firm that is ahead of<br />
its rival at any time t, in the sense of having completed more stages by time t, is<br />
likely to have chosen the faster strategy more often. We will c<strong>on</strong>struct numerical<br />
estimates of the probability that a leading firm is more likely to have chosen a<br />
strategy faster to verify this intuiti<strong>on</strong>.<br />
Moving away from the firm‐led race patterns revolving in a particular industry to a<br />
clustering of racing <strong>on</strong> an industry level is putting industry in different geoec<strong>on</strong>omics<br />
z<strong>on</strong>es against each other <strong>and</strong> becoming dominant in strategic<br />
product/process technologies. Here racing patterns am<strong>on</strong>g industries in a relatively<br />
free‐trade envir<strong>on</strong>ment could lead to competitive advantages <strong>and</strong> more wealth<br />
creating <strong>and</strong> accumulating dominance in key product/process technologies in <strong>on</strong>e<br />
regi<strong>on</strong> at the expense of others. There appears to be a link that individual races <strong>on</strong><br />
the firm level induce similar races <strong>on</strong> the industry level <strong>and</strong> will be a c<strong>on</strong>tributing<br />
factor to the globalizati<strong>on</strong> of network industries.<br />
Thus similar catch‐up processes are taking place between leaders <strong>and</strong> followers<br />
within a group of industrialized countries in pursuit of higher levels of productivity.<br />
Supposing that the level of labour productivity were governed entirely by the level<br />
of technology embodied in capital stock, <strong>on</strong>e may c<strong>on</strong>sider that the differentials in<br />
productivities am<strong>on</strong>g countries are caused by the ‘technological age’ of the stock<br />
relative to its ‘chr<strong>on</strong>ological age’. The technological age of capital is the age of<br />
expertise at the time of investment plus years elapsing from that time. Since a<br />
leading state may be supposed to be furnished with the capital stock embodying, in<br />
each vintage, technology which was ‘at the very fr<strong>on</strong>tier’ at the time of investment,<br />
the technological age of the stock is, so to speak, the same as its chr<strong>on</strong>ological age.<br />
While a leader is restricted in increasing its productivity by the advance of new<br />
technology, trailing countries have the potential to make a larger leap as they are<br />
provided with the privilege of exploiting the backlog in additi<strong>on</strong> of the newly<br />
developed technology. Hence, followers being behind with a larger gap in<br />
technology will have a str<strong>on</strong>ger potential for growth in productivity. The potential,<br />
however, will be reduced as the catch‐up process goes <strong>on</strong> because the unexploited<br />
stock of technology becomes smaller <strong>and</strong> smaller. However, as new technologies<br />
arise <strong>and</strong> are rapidly adopted in a Schumpeterian process of ‘creative destructi<strong>on</strong>’,<br />
their network effects induce rapid accelerating <strong>and</strong> cumulative growth potentials<br />
75<br />
323
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
which are catalyzed through industry competiti<strong>on</strong>. In the absence of such a process<br />
we can explain the tendency to c<strong>on</strong>vergence of productivity levels of follower<br />
countries. Historically, however, it fails to answer alleged puzzles as to why a<br />
country, such as the United States, has preserved the st<strong>and</strong>ing of the technological<br />
leader for a l<strong>on</strong>g time since taking over leadership from Britain in around the end of<br />
the nineteenth century <strong>and</strong> why the shifts have taken place in the ranks of follower<br />
states in their relative levels of productivity (i.e. technological gaps between them<br />
<strong>and</strong> the leader). Abramovitz (1986) poses some extensi<strong>on</strong>s <strong>and</strong> qualificati<strong>on</strong>s <strong>on</strong> this<br />
simple catch‐up hypothesis in an attempt to explain these facts. Am<strong>on</strong>g other factors<br />
than technological backwardness, he lays stress <strong>on</strong> a country’s social capability in<br />
terms of years of educati<strong>on</strong> as a proxy of technical competence <strong>and</strong> its political,<br />
commercial, industrial, <strong>and</strong> financial instituti<strong>on</strong>s. To become effective social<br />
capability may also include or exp<strong>and</strong> to ’deep craft’, a ’set of knowings’ <strong>on</strong><br />
technological performance <strong>and</strong> industrial techniques (Arthur, 2009). The social<br />
capacity of a state may become str<strong>on</strong>ger or weaker as technological gaps close or<br />
grow <strong>and</strong> thus Abramovitz argues that the actual catch‐up process does not provide<br />
itself to simple formulati<strong>on</strong>. This view has a comm<strong>on</strong> underst<strong>and</strong>ing to what another<br />
ec<strong>on</strong>omist, Ols<strong>on</strong> (1996), expresses to be ‘public policies <strong>and</strong> instituti<strong>on</strong>s’ as his<br />
explanati<strong>on</strong> of the great differences in per capita income across countries, stating<br />
that any poorer states that implement relatively good ec<strong>on</strong>omic policies <strong>and</strong><br />
instituti<strong>on</strong>s enjoy rapid catch‐up growth.<br />
The suggesti<strong>on</strong> should be taken seriously when we wish to underst<strong>and</strong> the<br />
technological catching‐up to American leadership by Japan, in particular during the<br />
post‐war period, <strong>and</strong> explore the possibility of a shift in st<strong>and</strong>ing between these two<br />
countries. This c<strong>on</strong>siderati<strong>on</strong> will directly bear <strong>on</strong> the future trend of the state of the<br />
art which exerts a crucial influence <strong>on</strong> the development of the world ec<strong>on</strong>omy (Juma<br />
& Clark, 2002; Fagerberg & Godinho, 2004). These explanati<strong>on</strong>s notwithst<strong>and</strong>ing, we<br />
venture as a major factor for divergent growth processes the level of intensity of the<br />
racing process within the most prevalent value‐added industries with cross‐secti<strong>on</strong>al<br />
spillovers. These are the communicati<strong>on</strong>s <strong>and</strong> informati<strong>on</strong> industries which have<br />
been shaped <strong>and</strong> led by leading American firms <strong>and</strong> where the rewards benefited<br />
their industries <strong>and</strong> country. Although European <strong>and</strong> Japanese companies were part<br />
of the race they were left behind in core markets reaping lesser benefits. (Since ICT<br />
investment relative to GDP is <strong>on</strong>ly less than half in states such as Japan, Germany<br />
<strong>and</strong> France compared to the US, 2% vs. more than 4% in 1999, this does not bode<br />
well for a rapid catch‐up in those countries <strong>and</strong> a fortiori, for the EU as a whole).<br />
Steering or guiding the process of racing through the pursuit of industrial policies<br />
aiming to increase competitive advantage of respective industries, as having been<br />
practiced in Japan, would stimulate catch‐up races but appears to be less effective in<br />
promoting fr<strong>on</strong>tier racing. Another profound reas<strong>on</strong> lies in the phenomen<strong>on</strong> of<br />
network externalities affecting ICT industries. That is, racing ahead of rivals in<br />
76 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
324
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
respective industries may create external ec<strong>on</strong>omies to the effect that such ec<strong>on</strong>omies<br />
within dominant industries tend to improve their internati<strong>on</strong>al market positi<strong>on</strong> <strong>and</strong><br />
therefore pull ahead in competitiveness vis‐à‐vis their (trading) partners. As<br />
Krugman (1991) observed: ‘It is probably true that external ec<strong>on</strong>omies are a more<br />
important determinant of internati<strong>on</strong>al trade in high technology sectors than<br />
elsewhere’. The point is that racing behaviour in leading high‐growth network<br />
industries by generating fr<strong>on</strong>tier positi<strong>on</strong>s, create critical cluster <strong>and</strong> network<br />
externalities pipelining through other sectors of the ec<strong>on</strong>omy <strong>and</strong> create competitive<br />
advantages elsewhere, as supported by the increasing returns debate (Arthur, 1996).<br />
In this sense we can speak of positive externalities endogenizing growth of these<br />
ec<strong>on</strong>omies <strong>and</strong> c<strong>on</strong>tributing to competitive advantage. All these characteristics lay<br />
the foundati<strong>on</strong>s of the ‘Network Ec<strong>on</strong>omy’.<br />
The Network Ec<strong>on</strong>omy is formed through an ever‐emerging <strong>and</strong> interacting set of<br />
increasing returns industries; it is about high‐intensity, technology driven‐racing,<br />
dynamic entrepreneurship, <strong>and</strong> focused risk‐taking through (free) venture capital<br />
markets endogenized by societal <strong>and</strong> instituti<strong>on</strong>al support. With the excepti<strong>on</strong> of<br />
pockets of activity in some parts of Europe (the UK <strong>and</strong> Sc<strong>and</strong>inavia), <strong>and</strong> in specific<br />
areas such as mobile communicati<strong>on</strong>s, these ingredients for the Network Ec<strong>on</strong>omy<br />
are <strong>on</strong>ly in the early stage of emerging in C<strong>on</strong>tinental Europe, <strong>and</strong> the political<br />
mindset in support of the Network Ec<strong>on</strong>omy is anything but prevalent. As l<strong>on</strong>g as<br />
we do not see a significant shift toward movements in this directi<strong>on</strong>, Europe will not<br />
see the full benefits of the Network Ec<strong>on</strong>omy within a Global Ec<strong>on</strong>omy.<br />
Racing behaviour <strong>on</strong> technological positi<strong>on</strong>s am<strong>on</strong>g firms in high‐ technology<br />
industries, as exemplified by the globally operating telecommunicati<strong>on</strong>s <strong>and</strong><br />
computer industries, produce spillover benefits in terms of increasing returns <strong>and</strong><br />
widespread productivity gains. Due to relentless competiti<strong>on</strong> am<strong>on</strong>g technological<br />
leaders the network effects result in significant advantages in the value added to this<br />
industry c<strong>on</strong>tributing to faster growth of GDP, <strong>and</strong> through a flexible labour market,<br />
also to employment growth. This c<strong>on</strong>stitutes a new paradigm in ec<strong>on</strong>omic thinking<br />
through network ec<strong>on</strong>omies <strong>and</strong> is a major gauge to compare the wealth‐creating<br />
power of the US ec<strong>on</strong>omy over the past decade against the European <strong>and</strong> advanced<br />
Asian ec<strong>on</strong>omies. It is interesting to speculate <strong>on</strong> the implicati<strong>on</strong>s of the way<br />
companies in major high‐technology markets, such as telecommunicati<strong>on</strong>s, split<br />
clearly into the two major technology races, with <strong>on</strong>e group of firms clearly lagging<br />
the other.<br />
The trajectories of technological evoluti<strong>on</strong> certainly seem to suggest that firms from<br />
<strong>on</strong>e fr<strong>on</strong>tier cannot simply jump to another trajectory. Witness, in this regard, the<br />
gradual process necessary for a firm in the catch‐up race to approach those in the<br />
77<br />
325
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
fr<strong>on</strong>tier race. There appears to be a fr<strong>on</strong>tier ‘lock‐in’, in that <strong>on</strong>ce a company is part<br />
of a race, the group of rivals within that same race are the <strong>on</strong>es whose acti<strong>on</strong>s<br />
influence that company’s strategy the most. Advancing technological capability is a<br />
cumulative process. The ability to advance to a given level of technical capability<br />
appears to be a functi<strong>on</strong> of existing technical capability. Given this path dependence,<br />
the questi<strong>on</strong> remains: why do some firms apparently choose a path of technological<br />
evoluti<strong>on</strong> that is less rapid than others? Two sets of possible explanati<strong>on</strong>s could be<br />
derived from our case analysis, which need not be mutually exclusive. The first<br />
explanati<strong>on</strong> lingers primarily <strong>on</strong> the expensive nature of R&D in industries like<br />
telecommunicati<strong>on</strong>s <strong>and</strong> computers which rely <strong>on</strong> novel discovery for their<br />
advancement. Firms choosing the catch‐up race will gain access to a particular<br />
technical level later than those choosing the fr<strong>on</strong>tier, but will do so at a lower cost.<br />
TECHNOLOGICAL FRONTIERS<br />
The evoluti<strong>on</strong> of a cross secti<strong>on</strong> of high technology industries reflects repetitive<br />
strategic interacti<strong>on</strong>s between companies in a c<strong>on</strong>tinuous quest to dominate the<br />
industry or at least to improve its competitive positi<strong>on</strong> through company level <strong>and</strong><br />
industry level technological evoluti<strong>on</strong>. We can observe several racing patterns across<br />
industries, each of which is the result of a subset of firms jockeying for a positi<strong>on</strong><br />
either as a race leader or for a positi<strong>on</strong> near the leader c<strong>on</strong>stituting a leadership club.<br />
The identificati<strong>on</strong> <strong>and</strong> interpretati<strong>on</strong> of the races relies <strong>on</strong> the fact that different firms<br />
take very different technological paths to target a superior performance level with<br />
the reward of increasing market shares, maintaining higher productivity <strong>and</strong><br />
profitability. In a Schumpeterian framework such races cannot be interpreted in a<br />
free‐riding situati<strong>on</strong> where <strong>on</strong>e firm exp<strong>and</strong>s resources in advancing the state of<br />
technology <strong>and</strong> the others follow closely behind. Such spillover interpretati<strong>on</strong>s are<br />
suspect when products are in the domain of high complexity, of high risk in<br />
succeeding, <strong>and</strong> different firms typically adopt different procedural <strong>and</strong><br />
architectural approaches.<br />
The logic underlying this evoluti<strong>on</strong> holds in any industry in which two broad sets of<br />
c<strong>on</strong>diti<strong>on</strong>s are satisfied. First, it pays for a firm to have a technological lead over its<br />
rival; it also boosts its market image <strong>and</strong> enhances its reputati<strong>on</strong>al capital. Sec<strong>on</strong>d,<br />
for various levels of technological complexity am<strong>on</strong>g the products introduced by<br />
various firms, technological complexity can be represented by a multi‐criteria<br />
performance measure, that is, by a vector‐valued distance measure. The collecti<strong>on</strong> of<br />
performance indicators, parameters, being c<strong>on</strong>nected with each other for individual<br />
companies form an envelope that shapes a ‘technological fr<strong>on</strong>tier’. The technological<br />
fr<strong>on</strong>tier is in fact a reas<strong>on</strong>able indicator of the evolving state of knowledge (technical<br />
expertise) in the industry. At any point in time the industry technology fr<strong>on</strong>tier (ITF)<br />
indicates the degree of technical sophisticati<strong>on</strong> of the most advanced products<br />
carried by companies in that industry in view of comparable performance st<strong>and</strong>ards.<br />
78 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
326
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
Firm level technology fr<strong>on</strong>tiers (FTF) are c<strong>on</strong>structed analogously <strong>and</strong> indicate, at<br />
any point in time, the extent of technical sophisticati<strong>on</strong> achieved by the firm until<br />
that point in time. The evoluti<strong>on</strong> of company <strong>and</strong> industry level fr<strong>on</strong>tiers is highly<br />
interactive. Groups of company fr<strong>on</strong>tiers are seen to co‐evolve in a manner that<br />
suggests that the respective firms are racing to catch up with, <strong>and</strong> get ahead of each<br />
other.<br />
A data set could focus <strong>on</strong> a given set of products (systems) by major European,<br />
American or Asian enterprises in those industries for a sufficiently representative<br />
period of market evoluti<strong>on</strong>. In principle, we can identify at least two races in<br />
progress in the industries throughout a given period of durati<strong>on</strong>. One comprises the<br />
world fr<strong>on</strong>tier race in each of those industries, the other, for example, the European<br />
fr<strong>on</strong>tier race which technically would c<strong>on</strong>stitute a subfr<strong>on</strong>tier to the worldwide race.<br />
The aggregate technology fr<strong>on</strong>tier of the firms in a particular race (that is, ITF) is<br />
c<strong>on</strong>structed in a manner similar to the individual FTFs. Essentially, the maximal<br />
envelope of the FTFs in a particular race c<strong>on</strong>stitutes the ITF for that race. The ITF<br />
indicates, as a functi<strong>on</strong> of calendar time, the best achievable performance by any<br />
firm in the race at a given date.<br />
A statistical profiling of technological evoluti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> relates to<br />
competitive racing am<strong>on</strong>g rival companies. Am<strong>on</strong>g the (n<strong>on</strong>‐exclusive) performance<br />
criteria to be assessed are: (1) frequency of fr<strong>on</strong>tier pushing; (2) technological<br />
dominati<strong>on</strong> period; (3) innovati<strong>on</strong>s vs. imitati<strong>on</strong>s in the race; (4) innovati<strong>on</strong><br />
frequency when behind or ahead; (5) nature of jumps, leapfrogging or fr<strong>on</strong>tiersticking;<br />
(6) inter‐jump times <strong>and</strong> jump sizes; (7) race closeness measures; (8) interfr<strong>on</strong>tier<br />
distance; (9) market leading through ‘market making’ innovati<strong>on</strong>s; <strong>and</strong> (10)<br />
leadership in ‘innovati<strong>on</strong> markets’.<br />
A race may or may not have different firms in the leadership positi<strong>on</strong> at different<br />
times. It may be a tighter race at some times than at others, <strong>and</strong> in general, may<br />
exhibit a variety of forms of interesting behaviour. While analysis of racing<br />
behaviour is left to various interpretati<strong>on</strong>s, it is appropriate to ask why the firms are<br />
motivated to keep <strong>on</strong> racing at all. As access to superior technology exp<strong>and</strong>s the<br />
scope of opportunities available to the firms, the technology can be applied in a<br />
range of markets. However, leading edge technology is acquired at a cost. It seems<br />
unlikely that all the companies would find it profitable to compete to be at the<br />
leading edge all the time. Also not every firm has access to equal capabilities in<br />
leveraging a given level of technological resources. Firms may, for example, be<br />
expected to differ in their access to complementary assets that allows them to<br />
appropriately reap the benefits from their innovati<strong>on</strong>. It is reas<strong>on</strong>able to assume that<br />
whatever the level of competence of a company in exploiting its resources it will be<br />
79<br />
327
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
better off the more advanced the technology. Based <strong>on</strong> this procedure an analysis<br />
will show how dynamic competiti<strong>on</strong> evolved in the past.<br />
Unlike other (statistical) indicators (such as patent statistics) referring to the degree<br />
of competitiveness am<strong>on</strong>g industries, regi<strong>on</strong>s <strong>and</strong> countries c<strong>on</strong>cerned, the proposed<br />
measures cover behavioral dynamic movements in respective industries, <strong>and</strong><br />
therefore are able to lend intrinsic predictive value to crucial ec<strong>on</strong>omic variables<br />
relating to ec<strong>on</strong>omic growth <strong>and</strong> wealth creati<strong>on</strong>. The results are likely to provide<br />
strategic support for industrial <strong>and</strong> technology policy in a regi<strong>on</strong>al or nati<strong>on</strong>al<br />
c<strong>on</strong>text <strong>and</strong> will enable policy makers to identify strengths <strong>and</strong> weaknesses of<br />
relevant players <strong>and</strong> their envir<strong>on</strong>ments in those markets. While this process looks<br />
like a micro representati<strong>on</strong> of dynamic technological evoluti<strong>on</strong> driving companies<br />
<strong>and</strong> industries into leadership positi<strong>on</strong>s, we may c<strong>on</strong>strue an analogous process that<br />
drives a regi<strong>on</strong> or a nati<strong>on</strong> into advancement <strong>on</strong> a macro scale in order to achieve a<br />
higher level pecking order am<strong>on</strong>g its peers. This may allow using the micro<br />
foundati<strong>on</strong>s of racing as a basis for identifying clubs of nati<strong>on</strong>s or regi<strong>on</strong>s am<strong>on</strong>g<br />
them to achieve higher levels <strong>and</strong> rates of growth.<br />
Catch‐up or Leapfrogging<br />
It was Schumpeter (1947) who observed that it is the expectati<strong>on</strong> of supernormal<br />
profits from a temporary m<strong>on</strong>opoly positi<strong>on</strong> following an innovati<strong>on</strong> that is the chief<br />
driver of R & D investment. Al<strong>on</strong>g this line, the simplest technology race model can<br />
be explained as follows: A number of firms invest in R&D. Their investment results<br />
in an innovati<strong>on</strong> with the time spent in R&D subject to some varying level of<br />
uncertainty. However, a greater investment reduces the expected time to completi<strong>on</strong><br />
of R&D. The model investigates how many firms will choose to enter such a c<strong>on</strong>test,<br />
<strong>and</strong> how much they will invest. However, despite some extensive theoretical<br />
examinati<strong>on</strong> of technological races there have been very few empirical studies <strong>on</strong><br />
this subject (Lerner, 1997) <strong>and</strong> virtually n<strong>on</strong>e in the c<strong>on</strong>text of major global<br />
industries, <strong>and</strong> <strong>on</strong> a comparative basis.<br />
Technological fr<strong>on</strong>tiers at the firm <strong>and</strong> industry race levels offer a powerful tool<br />
through which to view evolving technologies within an industry. By providing a<br />
benchmarking roadmap that shows where an individual firm is relative to the other<br />
firms in the industry, they highlight the importance of strategic interacti<strong>on</strong>s in the<br />
firm’s technology decisi<strong>on</strong>s. From the interactive process of racing could emerge<br />
various behavioural patterns. Does lagging behind <strong>on</strong>e’s closest technological rivals<br />
cause a firm to increase its innovative effort? The term ‘race’ suggests that no single<br />
company would want to fall too far behind, <strong>and</strong> that every<strong>on</strong>e would like to get<br />
ahead. If a firm tries to innovate more when it is behind than when it is ahead, then<br />
‘catch‐up’ behaviour will be the dominant effect. Once a firm gets ahead of its rivals<br />
noticeably, then rivals will step up their efforts to catch up. The leader will slow<br />
down its innovative efforts until its rivals have drawn uncomfortably close or have<br />
80 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
328
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
surpassed it. This process repeats itself every time a company gets far enough ahead<br />
of its rivals. An alternative behaviour pattern would corresp<strong>on</strong>d to a firm increasing<br />
its innovative effort if it gets far enough ahead, thus making catch‐up by the lagging<br />
firms increasingly difficult. This looks like the ‘Intel Model’ where <strong>on</strong>ly the paranoid<br />
survives (Grove, 1992). For any of these forms there appears to be a clear link to<br />
market <strong>and</strong> industry structure, as termed ‘intensity of rivalry’ by Kamien <strong>and</strong><br />
Schwarz (1982).<br />
We group two different kinds of races: <strong>on</strong>e that is a fr<strong>on</strong>tier race am<strong>on</strong>g leaders <strong>and</strong><br />
would‐be leaders (first league) <strong>and</strong> another that is a catch‐up race am<strong>on</strong>g laggards<br />
<strong>and</strong> imitators (sec<strong>on</strong>d league). Though both leagues may play their own game, in a<br />
free market c<strong>on</strong>test, it would be possible that a member of the sec<strong>on</strong>d league may<br />
penetrate into the first, as <strong>on</strong>e in the first league may fall back into the sec<strong>on</strong>d.<br />
Another aspect of innovati<strong>on</strong> speed has been addressed by Kessler <strong>and</strong> Bierly (2002).<br />
As a general rule they found that the speed to racing ahead may be less significant<br />
the more “radical” (drastic) the innovati<strong>on</strong> appears to be <strong>and</strong> the more likely it leads<br />
to a dominant design. These two forms have been applied empirically to the<br />
development of the early Japanese computer industry (<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, 2006a), that is, a<br />
fr<strong>on</strong>tier racing model regarding the struggle for technological leadership in the<br />
global industry between IBM <strong>and</strong> ‘Japan Inc.’ guided by MITI (now METI), <strong>and</strong> a<br />
catch‐up racing model relating to competiti<strong>on</strong> am<strong>on</strong>g the leading Japanese<br />
mainframe manufacturers as laggards.<br />
It is also interesting to distinguish between two sub‐categories of catch‐up<br />
behaviour. A lagging firm might simply try to close the gap between itself <strong>and</strong> the<br />
technological leader at any point in time (‘fr<strong>on</strong>tier‐sticking’ behaviour), or it might<br />
try to actually usurp the positi<strong>on</strong> of the leader by ‘leapfrogging’ it. When there are<br />
disproporti<strong>on</strong>ately large payoffs to being in the technical lead (relative to the payoffs<br />
that a firm can realize if it is simply close enough to the technical fr<strong>on</strong>tier), then <strong>on</strong>e<br />
would expect that leapfrogging behaviour would occur more frequently than<br />
fr<strong>on</strong>tier‐sticking behaviour (Owen & Ulph, 1994). Alternatively, racing toward the<br />
fr<strong>on</strong>tier creates the ‘reputati<strong>on</strong>’ of being an innovati<strong>on</strong> leader facilitating to maintain<br />
<strong>and</strong> increase market share in the future (Albach, 1997). All attempts to leapfrog the<br />
current technological leader might not be successful since many lagging firms might<br />
be attempting to leapfrog the leader simultaneously <strong>and</strong> the leader might be trying<br />
to get further ahead simultaneously. Corresp<strong>on</strong>dingly, <strong>on</strong>e should distinguish<br />
between attempted leapfrogging <strong>and</strong> realized leapfrogging. This phenomen<strong>on</strong><br />
(though dependent <strong>on</strong> industry structure) appears as the predominant behaviour<br />
pattern in the US <strong>and</strong> Japan fr<strong>on</strong>tier races (Brezis et al, 1991). Albach (1993) cites<br />
studies for Germany that show otherwise.<br />
81<br />
329
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
Leapfrogging behaviour influenced by the expected size of payoffs as suggested by<br />
Owen <strong>and</strong> Ulph (1994) might be revised in compliance with the characteristics of<br />
industrial structure of the local (regi<strong>on</strong>al) markets, the amount of R&D efforts for<br />
leapfrogging <strong>and</strong> the extent of globalizati<strong>on</strong> of the industry. Even in the case where<br />
the payoffs of being in the technological lead are expected to be disproporti<strong>on</strong>ately<br />
large, the lagging companies might be satisfied to remain close enough to the leader<br />
so as to gain or maintain a share in the local market. This could occur when the<br />
amount of R&D efforts (expenditures) required for leapfrogging would be too large<br />
for a lagging firm to be viable in the industry <strong>and</strong> when the local market has not<br />
been open enough for global competiti<strong>on</strong>: the local market might be protected for<br />
the lagging local companies under the auspices of measures of regulati<strong>on</strong> by the<br />
government (e.g. government purchasing, c<strong>on</strong>trols <strong>on</strong> foreign capital) <strong>and</strong> the<br />
c<strong>on</strong>diti<strong>on</strong>s preferable for these firms (e.g. language, marketing practices).<br />
When the industrial structure is composed of multi‐product companies, as for<br />
example it used to be in the Japanese computer industry, sub‐fr<strong>on</strong>tier firms may<br />
derive spill over benefits in developing new products in other technologically<br />
related fields (e.g. communicati<strong>on</strong>s equipment, c<strong>on</strong>sumer electr<strong>on</strong>ic products). These<br />
companies may prefer an R&D strategy just to keep up with the technological<br />
fr<strong>on</strong>tier level (catch‐up) through realizing a greater profit stream over a whole range<br />
of products.<br />
What are the implicati<strong>on</strong>s of the way firms split cleanly into the two technology<br />
races, with <strong>on</strong>e group clearly lagging the other technologically? The trajectories of<br />
technological evoluti<strong>on</strong> certainly seem to suggest that firms from <strong>on</strong>e fr<strong>on</strong>tier cannot<br />
simply jump to another trajectory. Witness, in this regards the gradual process<br />
necessary for the companies in the Japanese fr<strong>on</strong>tier to catch up with those at the<br />
global fr<strong>on</strong>tier. There appears to be a fr<strong>on</strong>t line ‘lock‐in’ in that <strong>on</strong>ce a firm is part of<br />
a race, the group of rivals within that same race are the <strong>on</strong>es whose acti<strong>on</strong>s influence<br />
the firm’s strategy the most.<br />
Advancing technological capability is a cumulative process. The ability to advance to<br />
a given level of technical capability appears to be a functi<strong>on</strong> of existing technical<br />
potential. Given this ‘path dependence’, the questi<strong>on</strong> remains: why do some firms<br />
apparently choose a path of technological evoluti<strong>on</strong> that is less rapid than others<br />
are? We propose two sets of possible explanati<strong>on</strong>s, which need not to be mutually<br />
exclusive. The first explanati<strong>on</strong> hinges primarily <strong>on</strong> the expensive nature of R & D in<br />
industries like the computer industry, which rely <strong>on</strong> novel scientific discovery for<br />
their advancement. Firms choosing the subfr<strong>on</strong>tier will gain access to a particular<br />
technical level later than those choosing the fr<strong>on</strong>tier, but will do so at a lower cost.<br />
Expending fewer resources <strong>on</strong> R & D ensures a slower rate of technical evoluti<strong>on</strong>.<br />
The sec<strong>on</strong>d explanati<strong>on</strong> relates mainly to technological spillovers. Following the<br />
success of the fr<strong>on</strong>tier firms in achieving a certain performance level, these become<br />
82 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
330
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
known to the subfr<strong>on</strong>tier firms. In fact, leading edge research in the computer<br />
industry is usually reported in patent applicati<strong>on</strong>s <strong>and</strong> scientific journals <strong>and</strong> is<br />
widely disseminated throughout the industry. The hypothesis is that partial<br />
spillover of knowledge occurs to the subfr<strong>on</strong>tier firms, whose task is then simplified<br />
to some extent. Notice that the subfr<strong>on</strong>tier firms still need to race to be technological<br />
leaders, as evidenced by the analysis above. This implies that the spillovers are<br />
nowhere near perfect. Company specific learning is still the norm. However, it is<br />
possible that knowing something about what research avenues have proved<br />
successful (for the fr<strong>on</strong>tier firms) could greatly ease the task for the firms that follow<br />
<strong>and</strong> try to match the technical level of the fr<strong>on</strong>tier company.<br />
Statistical Metrics of Industrial Racing Patterns<br />
Statistically descriptive measures of racing behaviour can be established that reflect<br />
the richness of the dynamics of ec<strong>on</strong>omic growth am<strong>on</strong>g competing nati<strong>on</strong>s. The<br />
point of departure for a statistical analysis of industrial racing patterns is the<br />
aggregate technological fr<strong>on</strong>tier represented by the nati<strong>on</strong>al producti<strong>on</strong> functi<strong>on</strong> as a<br />
reas<strong>on</strong>able indicator of the evolving state of knowledge (technical expertise) in a<br />
nati<strong>on</strong> or regi<strong>on</strong> which is the weighted aggregate of all industries or activities that<br />
themselves are represented by industry technology fr<strong>on</strong>tier (ITF). Firm level<br />
technology fr<strong>on</strong>tiers (FTF) are c<strong>on</strong>structed analogously <strong>and</strong> indicate, at any point in<br />
time, the weighted c<strong>on</strong>tributi<strong>on</strong> of that firm to the industry <strong>on</strong> st<strong>and</strong>ard industry<br />
classificati<strong>on</strong>.<br />
In this c<strong>on</strong>text we define ‘race’ as a c<strong>on</strong>tinual c<strong>on</strong>test for technological superiority<br />
am<strong>on</strong>g nati<strong>on</strong>s or regi<strong>on</strong>s with key industries. Under this c<strong>on</strong>ceptualisati<strong>on</strong> a race is<br />
characterised by a number of countries whose ITF’s remain ‘close’ together over a<br />
period (T) of, say, 25 to 50 years. The distinctive element is that countries engaging<br />
in a competiti<strong>on</strong> have ITF’s substantially closer together than those of any company<br />
not in the race. A statistical analysis should reflect that a race, as defined, may or<br />
may not have different countries in the leadership positi<strong>on</strong> at different times. It may<br />
be a tighter c<strong>on</strong>test at some times than at others, <strong>and</strong> in general, may exhibit a<br />
variety of forms of industrial behaviour. We look for clusters of firms who’s ITFs<br />
remain close enough throughout the durati<strong>on</strong> (formal measures of closeness are<br />
defined <strong>and</strong> measured). We identify races to take place at any level of industrial<br />
performance between the very top <strong>and</strong> the very bottom throughout 50 years<br />
durati<strong>on</strong> that is racing from the bottom to racing to the top.<br />
One comprises the world fr<strong>on</strong>tier race in each of those industries, the other a<br />
subfr<strong>on</strong>tier race (say, North America, Europe, East Asia, China, India, Latin<br />
America, Africa) which technically would c<strong>on</strong>stitute a subfr<strong>on</strong>tier to the world,<br />
allowing under the best of circumstances for the subfr<strong>on</strong>tier to be the fr<strong>on</strong>tier. Since<br />
83<br />
331
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
the level <strong>and</strong> breadth of industrial activity is reflected as an indicator for ec<strong>on</strong>omic<br />
welfare, racing to the top would go parallel with ec<strong>on</strong>omic growth <strong>and</strong> welfare<br />
enhancing, whereas racing from the bottom would corresp<strong>on</strong>d to poverty reducti<strong>on</strong><br />
<strong>and</strong> avoiding stati<strong>on</strong>ary (under)development traps.<br />
Characterizati<strong>on</strong> of Statistical Indicators of Industrial Racing<br />
While a variety of situati<strong>on</strong>s are possible, the extremes are the following: (a) <strong>on</strong>e<br />
country may push the fr<strong>on</strong>tier at all times, with the others following closely behind,<br />
(b) some countries share more or less equally in the task of advancing the most value<br />
generating industry technology fr<strong>on</strong>tiers (ITFs). Depending <strong>on</strong> the situati<strong>on</strong> the most<br />
value generating industries may be high technology based increasing returns or<br />
network industries that are able to induce complementary emerging industries with<br />
high potentials. Extreme situati<strong>on</strong> (a) corresp<strong>on</strong>ds to the existence of a unique<br />
technological leader for a particular race, <strong>and</strong> a number of quick followers. Situati<strong>on</strong><br />
(b), <strong>on</strong> the other h<strong>and</strong>, corresp<strong>on</strong>ds to the existence of multiple technological leaders.<br />
Assessment of Fr<strong>on</strong>tier Pushing: .The relevant statistics for the races relate to counting<br />
the times the ITFs are pushed forward by countries or regi<strong>on</strong>s at large within a<br />
global or regi<strong>on</strong>al fr<strong>on</strong>tier. Fr<strong>on</strong>tier pushing can be triggered through industrial<br />
policy by governments or well fostered entrepreneurship in an advanced capitalistic<br />
system<br />
Dominati<strong>on</strong> Period Statistics: Accepting the view that a country/regi<strong>on</strong> has greater<br />
potential to earn income <strong>and</strong> build wealth from its technological positi<strong>on</strong> if it is<br />
ahead of its race suggests that it would be interesting to examine the durati<strong>on</strong> of<br />
time for which a country can expect to remain ahead <strong>on</strong>ce it finds itself pushing its<br />
ITF. We statistically define the ‘dominati<strong>on</strong> period’ to be the durati<strong>on</strong> of time for<br />
which a country leads its particular race. It is interesting to note that the mean<br />
dominati<strong>on</strong> period is virtually indistinguishable for the three races, <strong>and</strong> lies between<br />
three <strong>and</strong> four years. A difference of means test cannot reject the hypothesis that the<br />
mean years of dominati<strong>on</strong> tend to cluster but hardly c<strong>on</strong>verge. So countries in each<br />
of the races can expect to remain ahead approximately in proporti<strong>on</strong> to their<br />
technological capability <strong>and</strong> more than the amount of time after they have propelled<br />
themselves to the fr<strong>on</strong>t of their respective races. However, the dominati<strong>on</strong> period<br />
tends to be a more uncertain quantity in the world fr<strong>on</strong>tier race, to a lesser degree in<br />
the EU fr<strong>on</strong>tier race than in any smaller regi<strong>on</strong>al races (as evidenced by the lower<br />
dominati<strong>on</strong> period st<strong>and</strong>ard deviati<strong>on</strong>).<br />
Catch‐up Statistics: If key industries of a country push to innovate more when they<br />
are behind than when they are ahead, then ‘catch‐ up’ behaviour will be the<br />
dominant effect. For each country/regi<strong>on</strong>, these statistics compare the fracti<strong>on</strong> of the<br />
total innovati<strong>on</strong>s carried out by industries in that country (i.e. the fracti<strong>on</strong> of the total<br />
number of times that their ITFs advance) when it was engaging in its race when<br />
lagging, with the fracti<strong>on</strong> of times that the country actually led its race. In the<br />
84 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
332
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
absence of catch‐up behaviour, or behaviour leading to a country increasingly<br />
dominating its rivals, we would expect to see no difference in these fracti<strong>on</strong>s. Then<br />
the fracti<strong>on</strong> of time that a country is ahead of its race could be an unbiased estimator<br />
of the fracti<strong>on</strong> of innovati<strong>on</strong>s in its key industries that it engages in when it is ahead.<br />
Relevant data, however, suggest that this is usually not the case. They appear to<br />
show that the fracti<strong>on</strong> of times a state leads its race at any development level in a<br />
group or club is larger than the fracti<strong>on</strong> of innovati<strong>on</strong>s that occur when the country<br />
is ahead, i.e. more innovati<strong>on</strong>s occur when the country is lagging than would be<br />
expected in the absence of catch‐up or increasing dominance behaviour. A major<br />
excepti<strong>on</strong> would arise if the country would act like an ‘Intel Ec<strong>on</strong>omy’, where<br />
unchallenged leadership in key industries creates incentives to increase the lead to<br />
its rivals. Catch‐up behaviour is supported by additi<strong>on</strong>al observati<strong>on</strong>s, as derivable<br />
from c<strong>on</strong>vergence <strong>and</strong> c<strong>on</strong>diti<strong>on</strong>al c<strong>on</strong>vergence in the ec<strong>on</strong>omic growth process that<br />
countries make larger jumps (i.e. the ITFs advance more) when they are behind than<br />
when they are leading the race<br />
Leapfrogging Statistics: From this, the distincti<strong>on</strong> emerges between two kinds of catchup.<br />
A lagging country might simply try to close the gap between itself <strong>and</strong> the<br />
technological leader at any point in time (fr<strong>on</strong>tier‐sticking behaviour), or it might try<br />
to actually usurp the positi<strong>on</strong> of the leader by ‘leapfrogging’ it. When there are<br />
disproporti<strong>on</strong>al larger incomes per head when being in the technical lead (relative to<br />
a situati<strong>on</strong> that a country can realize if it is simply close enough to the technological<br />
fr<strong>on</strong>tier), then <strong>on</strong>e would expect that leapfrogging behaviour would make it a more<br />
attractive incentive than fr<strong>on</strong>tier‐sticking behaviour.<br />
All attempts to leapfrog the current technological leader might not be successful<br />
since many lagging firms/industries might be attempting to leapfrog the leader<br />
simultaneously. Corresp<strong>on</strong>dingly, we observe both the attempted leapfroggings <strong>and</strong><br />
the realized leapfroggings. It appears likely that the leapfrogging phenomen<strong>on</strong><br />
would be more predominant in the premier league than in following up leagues.<br />
Interfr<strong>on</strong>tier Distance: How l<strong>on</strong>g does ‘knowledge’ take to spillover from fr<strong>on</strong>tier to<br />
subfr<strong>on</strong>tier industries? This requires investigating “interfr<strong>on</strong>tier distance”. One<br />
measure of how much subfr<strong>on</strong>tier industries’ technology lags the fr<strong>on</strong>tier industries’<br />
technology could be graphed as “subfr<strong>on</strong>tier lag” in terms of calendar time. At each<br />
point in time, this is simply the absolute difference in the subfr<strong>on</strong>tier performance<br />
<strong>and</strong> the fr<strong>on</strong>tier performance time. The graph would clearly indicate that this<br />
measure has been declining or increasing more or less m<strong>on</strong>ot<strong>on</strong>ically over the past<br />
50 years to the extent that the subfr<strong>on</strong>tier industries have been able/unable to catch<br />
up with the fr<strong>on</strong>tier industries. A complementary measure would be to assess the<br />
difficulty of bridging the lag. That is, how much l<strong>on</strong>ger does it take the subfr<strong>on</strong>tier to<br />
85<br />
333
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
reach a certain level of technical achievement after the fr<strong>on</strong>tier has reached that<br />
level? Thus it might very well turn out that the interfr<strong>on</strong>tier distance may be<br />
decreasing though the difficulty in bridging the gap is increasing.<br />
Race Closeness Measure (RCM): N<strong>on</strong>e of the previous analyses tell us how close any of<br />
the overall races are over a period of time. The races are all distant/close by<br />
c<strong>on</strong>structi<strong>on</strong>, however, some might be closer than others, We define ‘a measure of<br />
closeness’ of a race (RCM) at a particular time as follows:<br />
RCM (t) = 0 N Fi (t) – Fj (t) 2 /N (t) (1)<br />
where Fi (t) is country’s i ITF at time t, Fj(t) is country’s j comparable ITF at time<br />
t = max ITF(t) for each i, j <strong>and</strong> N(t) is the number of active key value‐generating<br />
industries at time t.<br />
The measure (Equati<strong>on</strong> 1) thus c<strong>on</strong>structed has a lowest value of 0, which<br />
corresp<strong>on</strong>ds to a ‘dead heat’ race. Higher values of the measure corresp<strong>on</strong>d to races<br />
that are less close. Unlike the earlier characteristics (dominati<strong>on</strong> period length,<br />
innovati<strong>on</strong> when ahead versus when behind, leapfrogging versus fr<strong>on</strong>tier‐sticking)<br />
which investigate the behaviour of a particular feature of the race <strong>and</strong> of a particular<br />
industry in relati<strong>on</strong> to the race fr<strong>on</strong>tier, the RCM is more of an aggregate statistic of<br />
how close the various racing parties are at a point in time. The closeness measure is<br />
simply an indicati<strong>on</strong> of parity, <strong>and</strong> not <strong>on</strong>e that says anything per se about the<br />
evoluti<strong>on</strong> of the technological fr<strong>on</strong>tier. To see this, note that if n<strong>on</strong>e of the fr<strong>on</strong>tiers<br />
were evolving, the closeness measure would be 0, as it would be if all the fr<strong>on</strong>tiers<br />
were advancing in perfect lock‐step with <strong>on</strong>e another.<br />
TABLE 1. ITF SHIFTS ACROSS AGGREGATED INDUSTRIES<br />
Aggregate Industries 1980‐ ITF (max = 100) 2010 GDP (%)<br />
US 80 85 70<br />
EU 60 75 60<br />
China 15 60 50<br />
USSR 30 35 40<br />
India 25 40 30<br />
Brazil 20 30 25<br />
Japan 70 70 65<br />
We talk about value‐added increasing returns industries over a period of 30 years.<br />
The industries comprise ICT, C<strong>on</strong>sumer Electr<strong>on</strong>ics, Chemicals <strong>and</strong> Materials,<br />
Automobiles, Pharma/Biotech, Machine Tools, Medical Instruments,<br />
Aerospace/Defense, Energy Technologies, <strong>and</strong> HT Transportati<strong>on</strong> Systems. Industry<br />
sectors can be assigned to various countries/regi<strong>on</strong>s such as US, EU, China, Russia,<br />
India, Brazil, Japan (Table 1). We benchmark the industry technology fr<strong>on</strong>tiers (ITFs)<br />
accordingly, that is, highest ‘state of knowledge’ at time t is 100 pc. The countries’<br />
rank to the max ITFs is assessed as the share of the max ITF. The assessment<br />
intervals are spaced in five year intervals starting in 1980 until 2010.<br />
86 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
334
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
Efficiency<br />
Let’s explore the inefficiency of the follower nati<strong>on</strong>s; i.e., the negative effect <strong>on</strong> the<br />
potential technology gap stemming from inefficient social <strong>and</strong> instituti<strong>on</strong>al factors. A<br />
good example of cross industrial inefficiencies over a historically representative<br />
period (1810‐2000) is Russia that was hardly advancing ec<strong>on</strong>omically against<br />
underdeveloped benchmark countries <strong>and</strong> falling behind leading ec<strong>on</strong>omies,<br />
reinforced through the bolshevik revoluti<strong>on</strong> <strong>and</strong> its underperforming ec<strong>on</strong>omic<br />
mechanism design (Gaidar, 2012). Increasing efficiencies deblock catch‐up in lagging<br />
countries (Juma & Clark, 2002). Efficiency is found by dividing a nati<strong>on</strong>’s estimated<br />
fixed effect by the regi<strong>on</strong>al adopti<strong>on</strong> rate. As defined here, it is quite robust to<br />
different estimati<strong>on</strong>s <strong>and</strong> samples. The relative efficiencies of the nati<strong>on</strong>s within<br />
regi<strong>on</strong>s appear to c<strong>on</strong>form to comm<strong>on</strong> beliefs. For example, in Europe, the<br />
Netherl<strong>and</strong>s, Belgium <strong>and</strong> Switzerl<strong>and</strong> are the most efficient while Turkey, Portugal<br />
<strong>and</strong> Greece are the least efficient. In East Asia, H<strong>on</strong>g K<strong>on</strong>g is the most efficient while<br />
Ind<strong>on</strong>esia <strong>and</strong> Thail<strong>and</strong> are the least efficient. Finally, in Latin America, Mexico <strong>and</strong><br />
Argentina are at the top <strong>and</strong> H<strong>on</strong>duras <strong>and</strong> Bolivia at the bottom. Another way to<br />
discuss the findings is to c<strong>on</strong>sider the time required to catch‐up. Previously, Parente<br />
<strong>and</strong> Prescott (2004) showed that countries with lower levels of relative efficiency will<br />
adopt modern technologies at much later dates. C<strong>on</strong>versely, <strong>on</strong>e could argue that if<br />
those countries adopt modern technologies c<strong>on</strong>currently with their low level of<br />
relative efficiency then their rates of growth would stay at a subpar level of their<br />
potential.<br />
One major source of efficiency generati<strong>on</strong> for a country, according to Parente <strong>and</strong><br />
Prescott (2004), is bel<strong>on</strong>ging to a ‘free trade club’ that improves efficiency through<br />
greater industrial competiti<strong>on</strong>. We calculate the required time period until the<br />
nati<strong>on</strong>s reach their fr<strong>on</strong>tier when <strong>on</strong>ly the catch‐up term <strong>and</strong> inefficiency are allowed<br />
to vary across regi<strong>on</strong>s <strong>and</strong> countries. Two fr<strong>on</strong>tiers are c<strong>on</strong>sidered: nati<strong>on</strong>s’<br />
inefficiency fr<strong>on</strong>tier <strong>and</strong> the leader nati<strong>on</strong>’s fr<strong>on</strong>tier. The latter requires that the<br />
inefficiency levels fade away in time which we assume occurs at the rate of . The<br />
European countries, with the excepti<strong>on</strong> of Turkey, all seem to have reached their<br />
inefficiency reduced fr<strong>on</strong>tier. The same is true for most of the East Asian countries.<br />
Thus, these nati<strong>on</strong>s will not catch‐up with the US without higher accumulati<strong>on</strong> rates<br />
or improved efficiency. For Latin America, most countries are still catching‐up with<br />
their inefficiency fr<strong>on</strong>tier, so that if accumulati<strong>on</strong> rates were the same catch‐up<br />
would still take place through diffusi<strong>on</strong> of disembodied technology. Of course, if<br />
inefficiency levels remain then a follower could never completely catch‐up with the<br />
leader by taking advantage of the technology gap al<strong>on</strong>e. As an illustrative example,<br />
for the required time to catch‐up with the leader if inefficiency levels were<br />
improving at the rate much of Europe <strong>and</strong> Latin America could then approach the<br />
87<br />
335
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
fr<strong>on</strong>tier faster than East Asia <strong>on</strong> account of East Asia’s lower rate of technology<br />
adopti<strong>on</strong>. This begs the questi<strong>on</strong> of what determines these. (in)efficiencies?<br />
It is reas<strong>on</strong>able to expect a tradeoff between a general technology level (GTL) of a<br />
nati<strong>on</strong>’s leading industries <strong>and</strong> its instituti<strong>on</strong>al efficiencies (IE). Thus, using an<br />
aggregate score, (GTL, IE) , say, a country may be in the top rank of GTL but weak<br />
<strong>on</strong> IE which may be surpassed in growth by <strong>on</strong>e which is lower in GTL rank but<br />
str<strong>on</strong>g <strong>on</strong> IE.<br />
CONCLUSION AND FURTHER DISCUSSION<br />
Ec<strong>on</strong>omic growth over the l<strong>on</strong>g‐run can <strong>on</strong>ly be achieved in the course of a real,<br />
sustainable value‐creating process through industrial performance <strong>and</strong> open<br />
markets in which technology <strong>and</strong> innovati<strong>on</strong> are the key facilitators. Nati<strong>on</strong>s with<br />
their industries engage in rival c<strong>on</strong>tests in what we term industrial races within a<br />
given internati<strong>on</strong>al trade regime. This reflects a micro‐ec<strong>on</strong>omic based behavioral<br />
focus <strong>on</strong> ec<strong>on</strong>omic growth (positive or negative). It builds a deeper foundati<strong>on</strong> to<br />
explanati<strong>on</strong>s of ec<strong>on</strong>omic growth than c<strong>on</strong>venti<strong>on</strong>al macro‐ec<strong>on</strong>omic texts . It also<br />
uncovers the true sources of growth as a tool for growth diagnostics (Rodrik, 2007)<br />
allowing to embrace other observati<strong>on</strong>s <strong>on</strong> urban growth <strong>and</strong> n<strong>on</strong> primarily<br />
ec<strong>on</strong>omic factors. In an influential paper in Foreign Affairs entitled ‘Can India<br />
overtake China’ Huang <strong>and</strong> Khanna (2003) first looked at macro‐ec<strong>on</strong>omic factors,<br />
which favor China. They then c<strong>on</strong>sidered micro‐ec<strong>on</strong>omic structures <strong>and</strong> behaviors<br />
such as competent indigenous entrepreneurship, a sound capital market, an<br />
independent legal system, property rights <strong>and</strong> a grass roots approach to<br />
development. The latter all favor India in the l<strong>on</strong>g run, say over the next fifty years.<br />
In a widely covered empirical investigati<strong>on</strong> <strong>on</strong> global growth patterns we c<strong>on</strong>cur<br />
with Easterly <strong>and</strong> Levins’ (2002) finding that it is not factor accumulati<strong>on</strong>, per se, but<br />
total factor productivity that explains cross‐country differences in the level of GDP<br />
growth rates. This total productivity in turn is derived from technology (innovati<strong>on</strong>)<br />
transfer <strong>and</strong> diffusi<strong>on</strong>, its’ supporting instituti<strong>on</strong>al characteristics <strong>and</strong> cultural<br />
dependence. Of course, <strong>on</strong> a deeper level, c<strong>on</strong>siderati<strong>on</strong>s of merely formal<br />
instituti<strong>on</strong>s may not suffice for explanati<strong>on</strong>s but instead forms of ec<strong>on</strong>omic<br />
mechanism design may be called for that effectively deal with (enforce rules <strong>on</strong>)<br />
‘moral hazard’ <strong>and</strong> ‘adverse selecti<strong>on</strong>’ issues (Myers<strong>on</strong>, 2006). Ec<strong>on</strong>omic growth in a<br />
decentralized system would be fully supported by a a Hayek‐Hurwicz mechanism<br />
design.<br />
Observati<strong>on</strong>s <strong>on</strong> firm‐led racing patterns emerging in oligopolistic market structures<br />
of particular high tech industries, <strong>and</strong> the clustering of racing <strong>on</strong> an industry level<br />
are putting companies in different geo‐ec<strong>on</strong>omic z<strong>on</strong>es against each other, becoming<br />
dominant in strategic product/process technologies. Here racing patterns am<strong>on</strong>g<br />
industries in a relatively free trade envir<strong>on</strong>ment could lead to competitive<br />
88 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
336
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
advantages <strong>and</strong> more wealth creating <strong>and</strong> accumulating dominance in key product /<br />
process technologies in <strong>on</strong>e regi<strong>on</strong> at the expense of others. The questi<strong>on</strong> is whether<br />
individual c<strong>on</strong>tests <strong>on</strong> a firm level induce similar effects <strong>on</strong> an industry level <strong>and</strong> if<br />
so, what c<strong>on</strong>trolling effects may be rendered by regi<strong>on</strong>al or multilateral policies <strong>on</strong><br />
regulatory, trade <strong>and</strong> investment matters? The point is that racing behaviour in<br />
leading high technology industries by generating fr<strong>on</strong>tier positi<strong>on</strong>s create cluster<br />
<strong>and</strong> network externalities pipelining through other sectors of the ec<strong>on</strong>omy <strong>and</strong><br />
creating competitive advantages elsewhere, as supported by the ‘increasing returns’<br />
debate. In this sense we can speak of positive externalities endogenizing growth of<br />
these ec<strong>on</strong>omies <strong>and</strong> c<strong>on</strong>tributing to competitive advantage.<br />
We are about to show in the upcoming chapters how technological racing, rivalry<br />
<strong>and</strong> competiti<strong>on</strong> instigates a process of innovati<strong>on</strong>, industrial <strong>and</strong> market evoluti<strong>on</strong><br />
<strong>and</strong> how it extends to larger entities than firms <strong>and</strong> industries to regi<strong>on</strong>s <strong>and</strong><br />
nati<strong>on</strong>al ec<strong>on</strong>omies or ec<strong>on</strong>omy networks. It will show what drives ec<strong>on</strong>omic<br />
growth <strong>and</strong> globalizati<strong>on</strong>, which industries are most significantly affected <strong>and</strong> how<br />
technological racing results in value generati<strong>on</strong> in increasing returns <strong>and</strong> network<br />
industries. Furthermore, we c<strong>on</strong>sider how the emergence of selective managerial<br />
strategies is most likely to carry success in the pursuit of corporate <strong>and</strong> industrial<br />
policies.<br />
Welfare enhancing technology racing as a c<strong>on</strong>stituent element of the capitalist<br />
process reinforced by globalizati<strong>on</strong> provides social benefits far exceeding the costs.<br />
Even more important, any alternative path, other than the competitive, would likely<br />
be inferior given the costs in that it would generate a less valued <strong>and</strong> less welfare<br />
producing technology portfolio. That is, even if the competitive process is wasteful,<br />
(for example, in parallel or correlated technology development) its unique high<br />
value innovati<strong>on</strong> outcome far exceeds the benefits of any alternative path. There is<br />
historical, observati<strong>on</strong>al <strong>and</strong> analytical evidence given in <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g> <strong>and</strong> Goosen<br />
(2012).<br />
On a nati<strong>on</strong>al scale simple catch‐up hypotheses have put emphasis <strong>on</strong> the great<br />
potential of adopting unexploited technology in the early stage <strong>and</strong> the increase of<br />
self‐limiting power in the later stage. However, an actual growth path of<br />
technological trajectory of a specific ec<strong>on</strong>omy may overwhelmingly be c<strong>on</strong>strained<br />
by social capability. The capability also endogenously changes as states of the<br />
ec<strong>on</strong>omy <strong>and</strong> technology evolve. The success of ec<strong>on</strong>omic growth due to diffusi<strong>on</strong> of<br />
advanced technology or the possibility of leapfrogging is mainly attributable to how<br />
the social capability evolves (i.e., which effects become more influential: growing<br />
resp<strong>on</strong>siveness to competiti<strong>on</strong> or growing obstacles to it <strong>on</strong> account of vested<br />
interests <strong>and</strong> established positi<strong>on</strong>s). Another observati<strong>on</strong> relates to policy inferences<br />
89<br />
337
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
<strong>on</strong> market structure, entrepreneurship, innovati<strong>on</strong> activity, industrial policy <strong>and</strong><br />
regulatory frameworks in promoting <strong>and</strong> hindering industry fr<strong>on</strong>tier races in a<br />
global industrial c<strong>on</strong>text. Does lagging behind <strong>on</strong>e’s closest technological rivals<br />
cause an industry to increase its innovative effort?<br />
On an industry level, am<strong>on</strong>g the key issues to be addressed is the apparent inability<br />
of technology oriented corporati<strong>on</strong>s to maintain leadership in fields that they<br />
pi<strong>on</strong>eered. There is a presumpti<strong>on</strong> that firms fail to remain competitive because of<br />
agency problems or other suboptimal managerial behaviour within these<br />
organizati<strong>on</strong>s. An alternative explanati<strong>on</strong> is that technologically trailing firms, in<br />
symmetric competitive situati<strong>on</strong>s, will devote greater effort to innovati<strong>on</strong>, so that a<br />
failure of technological leaders to maintain their positi<strong>on</strong> is an appropriate resp<strong>on</strong>se<br />
to the competitive envir<strong>on</strong>ment. In asymmetric situati<strong>on</strong>s, with entrants challenging<br />
incumbents, research does dem<strong>on</strong>strate that start‐up firms show a str<strong>on</strong>ger<br />
endeavour to close up to or leapfrog the competitors. Such issues highlight the<br />
dynamics of the race within the given market structure in any of the areas<br />
c<strong>on</strong>cerned.<br />
Catch‐up processes are taking place between leaders <strong>and</strong> followers within a group of<br />
industrialized countries in pursuit of higher levels of productivity <strong>and</strong> ec<strong>on</strong>omic<br />
growth. Supposing that the level of labour productivity were governed entirely by<br />
the level of technology embodied in capital stock, <strong>on</strong>e may c<strong>on</strong>sider that the<br />
differentials in productivities am<strong>on</strong>g countries are caused by the ‘technological age’<br />
of the stock used by a country relative to its ‘chr<strong>on</strong>ological age’. The technological<br />
age of capital is a period of technology at the time of investment plus years elapsing<br />
from that time. Since a leading country may be supposed to be furnished with the<br />
capital stock embodying, in each vintage, technology which was ‘at the very fr<strong>on</strong>tier’<br />
at the time of investment, ‘the technological age of the stock is, so to speak, the same<br />
as its chr<strong>on</strong>ological age’. While a leader is restricted in increasing its productivity by<br />
the advance of new technology, trailing countries ‘have the potential to make a<br />
larger leap’ as they are provided with the privilege of exploiting the backlog in<br />
additi<strong>on</strong> of the newly developed technology. Hence, followers being behind with a<br />
larger gap in technology will have a str<strong>on</strong>ger potential for growth in productivity.<br />
The potential, however, will be reduced as the catch‐up process goes <strong>on</strong> because the<br />
unexploited stock of technology becomes smaller <strong>and</strong> smaller. However, as new<br />
technologies arise <strong>and</strong> are rapidly adopted in a Schumpeterian process of ‘creative<br />
destructi<strong>on</strong>’, their network effects induce rapid accelerating <strong>and</strong> cumulative growth<br />
potentials being catalyzed through industry racing.<br />
In the absence of such a process, we can explain the tendency to c<strong>on</strong>vergence of<br />
productivity levels of follower countries. Historically, it fails to answer alleged<br />
puzzles of why a country, such as the United States, has preserved the st<strong>and</strong>ing of<br />
the technological leader for a l<strong>on</strong>g time since taking over leadership from Britain in<br />
90 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
338
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
around the end of the nineteenth century <strong>and</strong> why the shifts have taken place in the<br />
ranks of follower countries in their relative levels of productivity (i.e. technological<br />
gaps between them <strong>and</strong> the leader). Abramovitz (1986) poses some extensi<strong>on</strong>s <strong>and</strong><br />
qualificati<strong>on</strong>s <strong>on</strong> this simple catch‐up hypothesis in the attempt to explain these<br />
facts. Am<strong>on</strong>g other factors than technological backwardness, he lays stress <strong>on</strong> a<br />
country’s ‘social capability’ (i.e. years of educati<strong>on</strong> as a proxy of technical<br />
competence <strong>and</strong> its political, commercial, industrial, <strong>and</strong> financial instituti<strong>on</strong>s). The<br />
social capability of a country may become str<strong>on</strong>ger or weaker as technological gaps<br />
close <strong>and</strong> thus, he states, the actual catch‐up process does not lend itself to simple<br />
formulati<strong>on</strong>. This view has a comm<strong>on</strong> underst<strong>and</strong>ing to what another ec<strong>on</strong>omist,<br />
Ols<strong>on</strong> (1996), expresses to be public policies <strong>and</strong> instituti<strong>on</strong>s as his explanati<strong>on</strong> of the<br />
great differences in per capita income across countries, stating that any poorer<br />
countries that adopt relatively good ec<strong>on</strong>omic policies <strong>and</strong> instituti<strong>on</strong>s enjoy rapid<br />
catch‐up growth. The suggesti<strong>on</strong> should be taken seriously when we wish to<br />
underst<strong>and</strong> the technological catching‐up to American leadership by Japan, in<br />
particular, during the post‐war period <strong>and</strong> explore the possibility of a shift in<br />
st<strong>and</strong>ing between these two countries. This c<strong>on</strong>siderati<strong>on</strong> will directly bear <strong>on</strong> the<br />
future trend of the state of the art, which exerts a crucial influence <strong>on</strong> the<br />
development of the world ec<strong>on</strong>omy.<br />
These explanati<strong>on</strong>s notwithst<strong>and</strong>ing, we venture as a major factor for divergent<br />
growth processes the level of intensity of the racing process within the most<br />
prevalent value‐added industries with cross secti<strong>on</strong>al spillovers. These are the<br />
communicati<strong>on</strong>s <strong>and</strong> informati<strong>on</strong> industries, which have been shaped <strong>and</strong> led by<br />
leading American firms <strong>and</strong> where the rewards benefited their industries <strong>and</strong><br />
country. Though European <strong>and</strong> Japanese companies were part of the race they were<br />
left behind in core markets reaping lesser benefits. The IT investment relative to<br />
GDP, for example, used to be <strong>on</strong>ly less than half in countries such as Japan,<br />
Germany <strong>and</strong> France compared to the US. This does not bode well for a rapid catchup<br />
in those countries. Steering or guiding the process of racing through the pursuit<br />
of industrial policies aiming to increase competitive advantage of respective<br />
industries, as having been practised in Japan, would stimulate catch‐up races but<br />
appears to be less effective in promoting fr<strong>on</strong>tier racing. Another profound reas<strong>on</strong><br />
lies in the phenomen<strong>on</strong> of network externalities affecting IT industries. That is,<br />
racing ahead of rivals in respective industries may create external ec<strong>on</strong>omies to the<br />
effect that such ec<strong>on</strong>omies within dominant industries tend to improve their<br />
internati<strong>on</strong>al market positi<strong>on</strong> <strong>and</strong> therefore pull ahead in competitiveness vis‐a‐vis<br />
their (trading) partners.<br />
The point is that racing behaviour in leading high growth network industries by<br />
generating fr<strong>on</strong>tier positi<strong>on</strong>s create critical cluster <strong>and</strong> network externalities<br />
91<br />
339
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
pipelining through other sectors of the ec<strong>on</strong>omy <strong>and</strong> creating competitive<br />
advantages elsewhere, as supported by the increasing returns debate (Arthur, 1996).<br />
In this sense we can speak of positive externalities endogenizing growth of these<br />
ec<strong>on</strong>omies <strong>and</strong> c<strong>on</strong>tributing to competitive advantage.<br />
All these characteristics lay the foundati<strong>on</strong>s of the ‘Network Ec<strong>on</strong>omy’. The latter is<br />
formed through an ever emerging <strong>and</strong> interacting set of increasing returns<br />
industries, it is about high‐intensity, technology driven racing, dynamic<br />
entrepreneurship, focussed risk‐taking through (free) venture capital markets<br />
endogenized by societal <strong>and</strong> instituti<strong>on</strong>al support.<br />
Racing behaviour <strong>on</strong> technological positi<strong>on</strong>s am<strong>on</strong>g firms in high technology<br />
industries, as exemplified by the globally operating telecommunicati<strong>on</strong>s, <strong>and</strong><br />
computer industries, produce spillover benefits in terms of increasing returns <strong>and</strong><br />
widespread productivity gains. Due to relentless competiti<strong>on</strong> am<strong>on</strong>g technological<br />
leaders the network effects lead to significant advantages in the value added to this<br />
industry, c<strong>on</strong>tributing to faster growth of GDP, <strong>and</strong> through a flexible labour<br />
market, also to employment growth. This c<strong>on</strong>stitutes a new paradigm in ec<strong>on</strong>omic<br />
thinking through network ec<strong>on</strong>omies <strong>and</strong> is a major gauge to compare the wealth<br />
creating power of the US ec<strong>on</strong>omy against the European <strong>and</strong> advanced Asian<br />
ec<strong>on</strong>omies.<br />
REFERENCES<br />
Abramovitz, M. (1986). Catching up, forging ahead, <strong>and</strong> falling behind, Journal of<br />
Ec<strong>on</strong>omic History, 66, 385‐406.<br />
Abramovitz, M. & David, P. (1995). C<strong>on</strong>vergence <strong>and</strong> Deferred Catchup’, Chapter 1<br />
in Growth <strong>and</strong> <strong>Development</strong>: The Ec<strong>on</strong>omics of the 21 st Century, Ralph L<strong>and</strong>au,<br />
Timothy Taylor, <strong>and</strong> Gavin Wright, eds., Stanford CA: Stanford University Press.<br />
Albach, H. (1994). Informati<strong>on</strong>, Zeit und Wettbewerb (Informati<strong>on</strong>, Time <strong>and</strong><br />
Competiti<strong>on</strong>), J. H. v<strong>on</strong> Thuenen Lecture, in Jahrestagung des Vereins fuer<br />
Socialpolitik, Muenster 1993, Berlin: Duncker und Humblot, 113‐154.<br />
Albach, H. (1997). Global competiti<strong>on</strong> am<strong>on</strong>g the few, The Ehrenrooth Lectures,<br />
Swedish School of Ec<strong>on</strong>omics <strong>and</strong> Business Administrati<strong>on</strong>, Res. Reports 40,<br />
Helsingfors.<br />
Amsden, A. H. (1989). Asia’s Next Giant, New York: Oxford University Press.<br />
Arthur,W. B. (1996). Increasing returns <strong>and</strong> the new world of business, Harvard<br />
Business Review, July–August, 100‐109.<br />
Arthur, W.B. (2009). The Nature of <strong>Technology</strong>, New York: Free Press.<br />
92 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
340
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
Brezis, E., Krugman, P. <strong>and</strong> Tsidd<strong>on</strong>, D. (1991). Leapfrogging: a theory of cycles in<br />
nati<strong>on</strong>al technological leadership, Nati<strong>on</strong>al Bureau of Ec<strong>on</strong>omic Research (NBER),<br />
Working Paper # 3886.<br />
Burkett, P. & Hart‐L<strong>and</strong>sberg, M. (2003). A Critique of “Catch‐Up“ Theories of<br />
<strong>Development</strong>, Journal of C<strong>on</strong>temporary Asia 33(2), 147‐171.<br />
C<strong>on</strong>itzer, V. & S<strong>and</strong>holm, T. (2002). Complexity of Mechanism Design, Computer<br />
Science Dept., Carnegie Mell<strong>on</strong> Univ., Pittburg,Pa., Disc.Paper, May 28, 1‐8.<br />
Easterly, W. & Levine, R. (2002). It’s not Factor Accumulati<strong>on</strong>: Stylized Facts <strong>and</strong><br />
Growth Models, Central Bank of Chile , Working Papers No. 164.<br />
Easterly, W. (2013). The Tyranny of Experts, Ec<strong>on</strong>omists, Dictators, <strong>and</strong> the<br />
Forgotten Rights of the Poor, New York: Basic Books.<br />
Fagerberg, J. & Godinho, M. M. (2004). Innovati<strong>on</strong> <strong>and</strong> Catching‐Up’, Chap. 20 <strong>on</strong><br />
H<strong>and</strong>book of Innovati<strong>on</strong>, Oxford: Blackwell.<br />
Feigenbaum, J., Schapiro, M. & Shenker, S. (2007). Distributed Algorithmic<br />
Mechanism Design, In: Nisan et al, Algorithmic Game Theory, Cambridge:<br />
Cambridge Univ. Press, 363‐384.<br />
Gaidar, Y. (2004). Modern Ec<strong>on</strong>omic Growth in Russia <strong>and</strong> Catch‐Up <strong>Development</strong>,<br />
Herald of Europe, 1‐11.<br />
Gaidar, Y. (2012), Russia, A L<strong>on</strong>g View. Cambridge, MA: MIT Press.<br />
Gerschenkr<strong>on</strong>, A. (1962). Ec<strong>on</strong>omic Backwardness in Historical Perspective,<br />
Cambridge, MA: Belknap Press.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H. W. (2006a). Global Technological Races, Japan <strong>and</strong> the World Ec<strong>on</strong>omy<br />
18, 181‐193.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H. W. (2006b). Innovati<strong>on</strong>, <strong>Technology</strong> <strong>and</strong> Hypercompetiti<strong>on</strong>, L<strong>on</strong>d<strong>on</strong>:<br />
Routledge.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H. W. & Goosen, M. (2012), Strategies of Ec<strong>on</strong>omic Growth <strong>and</strong> Catch‐Up,<br />
New York: Nova Science.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H. W. (2013). Quality of Service <strong>on</strong> Queueing Networks for the Internet,<br />
iBusiness 5, 1‐12.<br />
Grossman, G. & Shapiro, C. (1987). Dynamic R&D Competiti<strong>on</strong>, Ec<strong>on</strong>omic Journal,<br />
97, 372‐387.<br />
Grove, A. S. (1999). Only the Paranoid Survive, New York: Doubleday.<br />
Harris, C. & Vickers, J. (1987). Racing with uncertainty, Review of Ec<strong>on</strong>omic Studies,<br />
54, 305‐321.<br />
93<br />
341
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
Hayek, F. A. (1945). The Use of Knowledge in Society, American Ec<strong>on</strong>omic Review<br />
35, 519‐530.<br />
Huang, Y. & Khanna, T. (2003). Can India overtake China?, Foreign Policy July‐Aug.<br />
Hurwicz & Reiter, S. (2006). Designing Ec<strong>on</strong>omic Mechanisms, Cambridge Univ.<br />
Press, Cambridge<br />
Juma, C. & Clark, N. (2002). <strong>Technology</strong> Catch‐Up: Opportunities <strong>and</strong> Challenges for<br />
Developing Countries, Supra Occ. Papers, Research Centre for the Social Sciences,<br />
Univ. of Edinburgh, UK.<br />
Kessler, E. H. & Bierly, P. E. (2002). Is faster really better? An empirical test of the<br />
implicati<strong>on</strong>s of innovati<strong>on</strong> speed, IEEE Transacti<strong>on</strong>s <strong>on</strong> Engineering Management,<br />
49, 2‐12.<br />
Kamien, M. I. & Schwarz, N. L. (1982). Market Structure <strong>and</strong> Innovati<strong>on</strong>, Cambridge:<br />
Cambridge University Press.<br />
Krugman, P. (1991). Increasing Returns <strong>and</strong> Ec<strong>on</strong>omic Geography, Journal of<br />
Political Ec<strong>on</strong>omy, 99(3), 483‐499.<br />
Krugman, P. (1994). The Myth of Asia’s Miracle, Foreign Affairs 73 ,63‐78.<br />
Lee, K. (2013). Schumpeterian Analysis of<br />
Cambridge University Press.<br />
Ec<strong>on</strong>omic Catchup. Cambridge:<br />
Lerner, J. (1997). An empirical explorati<strong>on</strong> of a technology race, R<strong>and</strong> Journal of<br />
Ec<strong>on</strong>omics, 28(2), 228‐224.<br />
Lin, J. Y. (2013). Against the C<strong>on</strong>sensus, Cambridge: Cambridge University Press.<br />
Myers<strong>on</strong>, R. B. (2008). Perspectives of Mechanism Design in Ec<strong>on</strong>omic Theory,<br />
American Ec<strong>on</strong>omic Review, 98(3), 586‐603.<br />
Nisan, N. & R<strong>on</strong>en, A. (2001). Algorithmic Mechanism Design, Games <strong>and</strong> Ec<strong>on</strong>omic<br />
Behavior 35, 166‐196.<br />
Nisan, N., Roughgarden, T., Tardos, E. & Vazirani, V. V. eds. (2007). Algorithmic<br />
Game Theory, Cambridge: Cambridge University Press.<br />
Olsen, M. (1996). Big Bills left <strong>on</strong> the Sidewalk: Why some nati<strong>on</strong>s are rich <strong>and</strong> others<br />
poor, Journal of Ec<strong>on</strong>omic Perspectives 10, 3‐24.<br />
Owen, R. & Ulph, D. (1994). Racing in two dimensi<strong>on</strong>s, Journal of Evoluti<strong>on</strong>ary<br />
Ec<strong>on</strong>omics, 4, 185‐206.<br />
Rodrik,D. (2007). One Ec<strong>on</strong>omics, Many Recipes: Globalizati<strong>on</strong>, Instituti<strong>on</strong>s <strong>and</strong><br />
Ec<strong>on</strong>omic Growth, Princet<strong>on</strong>: Princet<strong>on</strong> University Press.<br />
94 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
342
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
Sachs, J. D. & Warner, A. M. (1996). Achieving Rapid Growth in the Transiti<strong>on</strong><br />
Ec<strong>on</strong>omies of Central Europe, Harvard Institute for Internati<strong>on</strong>al <strong>Development</strong>,<br />
Discussi<strong>on</strong> Paper 544, July.<br />
Scherer, F. M. (1992). Internati<strong>on</strong>al High‐<strong>Technology</strong> Competiti<strong>on</strong>, Cambridge:<br />
Harvard University Press.<br />
Scherer, F. M. (1999). New Perspectives <strong>on</strong> Ec<strong>on</strong>omic Growth <strong>and</strong> Technological<br />
Innovati<strong>on</strong>, Washingt<strong>on</strong> D. C.: Brookings Instituti<strong>on</strong> Press.<br />
Schumpeter, J. (1947). Capitalism, Socialism <strong>and</strong> Democracy, New York: Harper,<br />
Rev. Editi<strong>on</strong>.<br />
Wan, H. Y. (2004). Ec<strong>on</strong>omic <strong>Development</strong> in a Globalized Envir<strong>on</strong>ment, East Asian<br />
Evidence, Dordrecht, Kluwer.<br />
95<br />
343
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
344
3.4 ESCAPING FROM DEVELOPMENT TRAPS<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
ESCAPING FROM DEVELOPMENT<br />
TRAPS: INDUSTRIALIZATION AND<br />
RACING FROM THE BOTTOM<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g> 1* , Celia Umali 2<br />
1<br />
STRATEC Munich, Germany<br />
2<br />
Faculty of Ec<strong>on</strong>omics, Nagasaki University, Nagasaki, Japan<br />
*umari@nagasaki‐u.ac.jp<br />
Abstract<br />
For development ec<strong>on</strong>omies, escaping costly development traps an industrializati<strong>on</strong> policy is likely to<br />
be gradual rather than take a “big push” form <strong>and</strong> becoming more balanced over time. Under certain<br />
c<strong>on</strong>diti<strong>on</strong>s the optimal industrializati<strong>on</strong> policy should be more unbalanced the weaker are the<br />
sectorial linkages, the str<strong>on</strong>ger are increasing returns, entrepreneurial resources, <strong>and</strong> the smaller are<br />
the domestic market size <strong>and</strong> the lesser the degree of dynamic competiti<strong>on</strong>. We show how to make<br />
tradeoffs at different levels of development <strong>and</strong> from the perspective of the industrializati<strong>on</strong> debate in<br />
a historical c<strong>on</strong>text of modern development policies.<br />
Key words<br />
<strong>Development</strong> traps; Tradeoffs; Industrializati<strong>on</strong>.<br />
INTRODUCTION<br />
The cumulative literature <strong>on</strong> industrializati<strong>on</strong> has formalized the l<strong>on</strong>g‐st<strong>and</strong>ing idea<br />
that development traps are the result of a failure of ec<strong>on</strong>omic organizati<strong>on</strong> rather<br />
than a lack of resources or other technological c<strong>on</strong>straints. The so‐called “big push”<br />
models of industrializati<strong>on</strong> have shown how, in the presence of increasing returns,<br />
there can exist preferable states to advance the ec<strong>on</strong>omic states of countries in<br />
c<strong>on</strong>test with other countries. Such a view not <strong>on</strong>ly provides an explanati<strong>on</strong> for the<br />
co‐existence of industrialized <strong>and</strong> less industrialized ec<strong>on</strong>omies, but also a rati<strong>on</strong>ale<br />
for government interventi<strong>on</strong> to coordinate investment in a “big‐push” toward<br />
industrializati<strong>on</strong>. Moreover, unlike competing theories, these models emphasize the<br />
temporary nature of any policy. Thus, industrializati<strong>on</strong> policy involves facilitating an<br />
adjustment from <strong>on</strong>e equilibrium to another rather than any change in the nature of<br />
the set of equilibria per se.<br />
5<br />
345
HANS W. GOTTINGER<br />
3.4 ESCAPING FROM DEVELOPMENT TRAPS<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, Celia Umali<br />
Escaping from <strong>Development</strong> Traps: Industrializati<strong>on</strong> <strong>and</strong> Racing from the Bottom<br />
While recent formalizati<strong>on</strong> makes clear the possible role for the government in<br />
coordinating ec<strong>on</strong>omic activity, little has been said about the form such policy<br />
should take. Is there a c<strong>on</strong>ceptual model to analyze the questi<strong>on</strong>: what precise form<br />
should the “big push” take? It should be part of mechanism design for ec<strong>on</strong>omic<br />
development (<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, 2014). It is argued that while many different<br />
industrializati<strong>on</strong> policies can be successful in generating escapes from development<br />
traps, the form of the policy that minimizes the costs of this transiti<strong>on</strong> depends <strong>on</strong><br />
the characteristics of the ec<strong>on</strong>omic situati<strong>on</strong> at h<strong>and</strong>. Factors such as the strength of<br />
the complementarities, externalities <strong>and</strong> increasing returns, am<strong>on</strong>g others, all play a<br />
role in influencing the nature of a “getting‐ahead” industrializati<strong>on</strong> policy. Such<br />
ideas were already present in the debates in development ec<strong>on</strong>omics in the 1940s<br />
<strong>and</strong> 1950s regarding the form of industrializati<strong>on</strong> policy. The models underlying<br />
these less formal debates inspired the recent more formal research but the policy<br />
elements of these have not been addressed, to date, in any substantive way.<br />
The paper proceeds as follows. We first give a brief history to recall different<br />
development strategies proliferating in the literature, in Secti<strong>on</strong> 2. Then we show<br />
how the increasing returns debate <strong>on</strong> industries impact structural change <strong>and</strong><br />
development paths, in Secti<strong>on</strong> 3. Secti<strong>on</strong> 4 gives the industrializati<strong>on</strong> policydevelopment<br />
c<strong>on</strong>text in an optimizati<strong>on</strong> framework. C<strong>on</strong>clusi<strong>on</strong>s follow in Secti<strong>on</strong> 5.<br />
A BRIEF HISTORY OF MAJOR DEVELOPMENT STRATEGIES<br />
Principal am<strong>on</strong>g the earlier policy debates was that surrounding the efficacy <strong>and</strong><br />
costs involved in the alternative strategies of “balanced” versus “unbalanced<br />
growth.” Rosenstein‐Rodan (1943, 1961) <strong>and</strong> Nurkse (1952, 1953) provided the<br />
rati<strong>on</strong>ale for the noti<strong>on</strong> that the adopti<strong>on</strong> of modern technologies must proceed<br />
across a wide range of industries more or less simultaneously. It was argued that the<br />
neglect of investment in a sector(s) could undermine any industrializati<strong>on</strong> strategy.<br />
Reacting to this policy prescripti<strong>on</strong> was the “unbalanced growth” school led by<br />
Hirschman (1958) <strong>and</strong> Streeten (1956). They saw the balanced strategy as far too<br />
costly. The advantages of multiple development may make interesting reading for<br />
ec<strong>on</strong>omists, but they are gloomy news indeed for underdeveloped countries. The<br />
initial resources for simultaneous developments <strong>on</strong> many fr<strong>on</strong>ts are generally<br />
lacking. By targeting many sectors, it was argued that scarce resources would be<br />
spread too thin‐ so thin, that industrializati<strong>on</strong> would be thwarted. It seemed more<br />
fruitful to target a small number of “leading sectors” (Rostow, 1960). Then those<br />
investments would “….call forth complementary investments in the next period with<br />
a will <strong>and</strong> logic of their own: they block out a part of the road that lies ahead <strong>and</strong><br />
virtually compel certain additi<strong>on</strong>al investment decisi<strong>on</strong>s” (Hirschman, 1958: 42).<br />
Thus, the existence of complementarity between investments (in particular those<br />
involving human capital) <strong>and</strong> increasing returns motivated an unbalanced approach<br />
(Easterly, 2002). Curiously, at the same time, “complementarity of industries<br />
6 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL. 3, ISSUE 1 ‐ MARCH , 2015, PP. 5‐13<br />
346
3.4 ESCAPING FROM DEVELOPMENT TRAPS<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
provides the most important set of arguments in favor of a large‐scale planned<br />
industrializati<strong>on</strong>”(Rosenstein‐Rodan, 1943: 205). Further, <strong>on</strong>e of the first to preview<br />
the c<strong>on</strong>necti<strong>on</strong> between Big Push, Poverty Traps <strong>and</strong> Takeoffs was the essay by W.<br />
Easterly (2005) who integrated historical sources with present day modern<br />
development strategies. Both sides appeared to have agreed that a “big push” was<br />
warranted, but they disagreed as to its compositi<strong>on</strong>. Our purpose here is to use the<br />
guidelines provided by the more recent formalizati<strong>on</strong> of the “big push” theory of<br />
industrializati<strong>on</strong> to clarify the earlier debate of the appropriate degree of focus for<br />
industrializati<strong>on</strong> policy. After all, the more recent literature has stressed the roles of<br />
complementarities <strong>and</strong> increasing returns that both schools saw lying at the heart of<br />
their policy prescripti<strong>on</strong>s.<br />
The seminal article formalizing the “big push” theory of industrializati<strong>on</strong> is that of<br />
Murphy et al, (1989). In their model, firms choose between c<strong>on</strong>stant returns <strong>and</strong> an<br />
increasing returns technology based <strong>on</strong> their expectati<strong>on</strong>s of dem<strong>and</strong>. However,<br />
these choices spill over into aggregate dem<strong>and</strong> creating a strategic interacti<strong>on</strong> am<strong>on</strong>g<br />
sectors in their technology adopti<strong>on</strong> decisi<strong>on</strong>s. Thus, under certain c<strong>on</strong>diti<strong>on</strong>s, there<br />
exist two equilibria: with all firms choosing the c<strong>on</strong>stant returns or all choosing the<br />
increasing returns technology. Clearly, in the latter equilibrium, all households are<br />
better off.<br />
While the Murphy et al, (1989) model shows how increasing returns (<strong>and</strong> a wage<br />
effect) aggregate to strategic complementarity am<strong>on</strong>g sectors, it does not lend itself<br />
readily to the debate c<strong>on</strong>cerning the degree of balance in industrializati<strong>on</strong> policy.<br />
First, the static c<strong>on</strong>tent leaves open the questi<strong>on</strong> of whether the interventi<strong>on</strong> should<br />
take the form of anything more than indicative planning. Sec<strong>on</strong>d, the most<br />
comm<strong>on</strong>ly discussed policy instrument in the industrializati<strong>on</strong> debate is the<br />
subsidizati<strong>on</strong> of investments. However, in the Murphy et al, (1989) example, use of<br />
this instrument biases <strong>on</strong>e toward a more unbalanced policy. To see this, observe that<br />
it is the role of the government to facilitate a move to the industrializing equilibrium.<br />
This means that the government must subsidize a sufficient amount of investment to<br />
make it profitable for all sectors to adopt the modern technology.<br />
Given the binary choice set, there then exists some minimum critical mass of sectors<br />
that must be targeted to achieve a successful transiti<strong>on</strong>. A greater range of successful<br />
industrializati<strong>on</strong> policies might be more plausible, however, if firms had the choice<br />
of a wider variety of technology to choose from (<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, 2006; <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g> &<br />
Goosen, 2012). One might suppose that targeting a large number of sectors to<br />
modernize a little <strong>and</strong> targeting a small number of sectors for more radical<br />
modernizati<strong>on</strong> might both generate a big push. Thus, to c<strong>on</strong>sider the balanced<br />
approach properly, a greater technological choice space is required.<br />
7<br />
347
HANS W. GOTTINGER<br />
3.4 ESCAPING FROM DEVELOPMENT TRAPS<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, Celia Umali<br />
Escaping from <strong>Development</strong> Traps: Industrializati<strong>on</strong> <strong>and</strong> Racing from the Bottom<br />
INCREASING RETURNS, STRUCTURAL CHANGE AND<br />
DEVELOPMENT PATHS<br />
What would be the choice variables available to the government provided it would<br />
be able to pick up what is likely to be increasing returns industries in the future?<br />
First, in each period, the government can choose the set of firms that it targets for<br />
structural change. Sec<strong>on</strong>d, for each targeted firm, the government can choose a<br />
target level for ‘increasing returns industry’ modernizati<strong>on</strong> in the period. Al<strong>on</strong>g this<br />
vein, the government could choose to target the same number of firms in each period<br />
but induce those firms to modernize gradually over time. Or in c<strong>on</strong>trast, the<br />
government chooses a single level of modernizati<strong>on</strong> to occur across all firms <strong>and</strong> all<br />
periods. It then targets a mass of firms each period for entry <strong>and</strong> modernizati<strong>on</strong>. This<br />
means that industrializati<strong>on</strong> policy is solely characterized by the critical mass of<br />
sectors targeted, <strong>and</strong> the target level of modernizati<strong>on</strong>. The level of modernizati<strong>on</strong><br />
could be sequentially exp<strong>and</strong>ed by infrastructural upgrading across the board to<br />
benefit all major sectors as suggested by the Chinese ec<strong>on</strong>omist Justin Yifu Lin<br />
(2013).<br />
Given a parameterized development path, the most significant parameter represents<br />
the strength of increasing returns in the technology adopted by industrial sectors,<br />
which generates a rati<strong>on</strong>ale for “big push” interventi<strong>on</strong>. A “big push” can be<br />
activated if the ec<strong>on</strong>omy is stuck in a “development trap” from which an escape<br />
could be made through sufficient coordinati<strong>on</strong> of decisi<strong>on</strong>s by input producers. For a<br />
developing ec<strong>on</strong>omy in its early phase a “poverty trap” is a special case of a<br />
“development trap” defined by Barro <strong>and</strong> Sala‐i‐Martin (1995: 49) as a stable steadystate<br />
with low levels of per capita output <strong>and</strong> capital stock. This is a trap because, if<br />
agents attempt to break out of it, the ec<strong>on</strong>omy has a tendency to return to the lowlevel<br />
steady‐state. Only by a very large change in their behavior, can the ec<strong>on</strong>omy<br />
break out of the poverty trap <strong>and</strong> move to the high‐income steady state. To evaluate<br />
the ec<strong>on</strong>omic characteristics, i.e., the strengths of complementarities <strong>and</strong> increasing<br />
returns, would affect the government’s policy choices <strong>and</strong> industrial policies (Gans,<br />
1994).<br />
Big Push theories of industrializati<strong>on</strong> could lead to ‘development traps’ if<br />
sequential industrializati<strong>on</strong> would add more diminishing returns than increasing<br />
returns industries which could be a result of government’s coordinati<strong>on</strong> failure.<br />
This would point to deficiencies in instituti<strong>on</strong>al quality as outlined by North (1990)<br />
impacting ec<strong>on</strong>omic performance. They could give explanati<strong>on</strong>s for decade l<strong>on</strong>g<br />
lackluster performance of Latin American ec<strong>on</strong>omies (Fukuyama, 2008). When a<br />
development trap is purely the result of coordinati<strong>on</strong> failure, to escape from the trap,<br />
would technically require the government to synchr<strong>on</strong>ize the expectati<strong>on</strong>s of<br />
individual agents (entrepreneurs) with targeting investment in industrializati<strong>on</strong><br />
activities. If a government were to announce that firms should modernize to a<br />
8 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL. 3, ISSUE 1 ‐ MARCH , 2015, PP. 5‐13<br />
348
3.4 ESCAPING FROM DEVELOPMENT TRAPS<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
certain degree, even if this were believed perfectly by individuals <strong>and</strong> firms, each<br />
firm might still have an incentive to wait before investing. In that case, the optimistic<br />
expectati<strong>on</strong>s by the government would not be realized <strong>and</strong> the policy would be<br />
ineffective. Irreversibility <strong>and</strong> the time lag of producti<strong>on</strong> mean that history rather<br />
than expectati<strong>on</strong>s matter (Krugman, 1991). The previous level of industrializati<strong>on</strong><br />
determines what path the ec<strong>on</strong>omy will take in the future. This is why it is difficult<br />
to characterize the industrializing paths of the ec<strong>on</strong>omy. There is ec<strong>on</strong>ometric<br />
evidence that a c<strong>on</strong>tributing factor toward the emergence of development traps is the<br />
lack of surpassing some threshold of technological integrati<strong>on</strong> in the industrializing<br />
(manufacturing) sector (Ortiz et al, 2009).<br />
INDUSTRIALIZATION POLICIES AND DEVELOPMENT<br />
In the c<strong>on</strong>text of a big push development strategy the government faces a tradeoff<br />
between the number of sectors it targets <strong>and</strong> the degree to which it wishes them to<br />
modernize, that is, it chooses the critical mass of sectors that must be targeted at any<br />
point in time in order to generate an escape from a development trap <strong>and</strong> to achieve<br />
increasing returns. Let’s take a simple case where the industrializati<strong>on</strong> policy takes<br />
the form of a “big bang”, that is interventi<strong>on</strong> occurs for <strong>on</strong>e period <strong>on</strong>ly granting that<br />
the resources exist in that period to allow for such a policy. This means that the<br />
industrializati<strong>on</strong> policy is solely characterized by the critical mass of sectors targeted<br />
s * <strong>and</strong> the target level of modernizati<strong>on</strong> ƒ.<br />
Suppose naturally that individual transiti<strong>on</strong> costs are n<strong>on</strong>‐decreasing in ƒ, the<br />
optimal critical mass in terms of ƒ can be described by the path<br />
s * (ƒ¸) (ƒ + 1) ((1‐)(‐1)/ L)<br />
varieties of the industrial ec<strong>on</strong>omy.<br />
<br />
<br />
sI ) + sI with sI as the basic input<br />
Substituting this into the objective functi<strong>on</strong> with cost c (ƒ,1; sI, ), the ‘big bang’<br />
industrializati<strong>on</strong> policy problem becomes<br />
min (ƒ + 1) ((1‐)(‐1)/ L) sI ) c (ƒ,1; sI, )<br />
where use is made of the symmetry of the cost functi<strong>on</strong>s <strong>and</strong> the fact that s sI<br />
firms are targeted. could represent any given exogenous parameter, i.e. or<br />
L a given parameter linked to L <br />
In designing an optimal industrializati<strong>on</strong> policy it shows that a cost minimizing<br />
policy in the industry transiti<strong>on</strong> entails setting certain development model<br />
(exogeneous) parameters such as labour productivity improvements (), upstream<br />
firms discount future earnings () the fixed size of the labour force (L ), the number<br />
of basic industrial sector varieties (sI) , the product linkages between intermediate<br />
input producers (), <strong>and</strong> the use of the intermediate input composite (), the latter<br />
two showing a certain degree of interacti<strong>on</strong> referring to as the returns to<br />
9<br />
349
HANS W. GOTTINGER<br />
3.4 ESCAPING FROM DEVELOPMENT TRAPS<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, Celia Umali<br />
Escaping from <strong>Development</strong> Traps: Industrializati<strong>on</strong> <strong>and</strong> Racing from the Bottom<br />
specializati<strong>on</strong> (Romer, 1986). Discussing these parameters qualitatively in terms of<br />
comparative statics would indicate industrial change. Raising any of these<br />
parameters , , L , <strong>and</strong> sI increases the marginal returns to upstream firms in both<br />
their entry <strong>and</strong> modernizati<strong>on</strong> decisi<strong>on</strong>s.<br />
Raising means that sunk costs are translated into labor improvements more<br />
effectively. Similarly, since the costs of modernizati<strong>on</strong> <strong>and</strong> entry are carried today<br />
<strong>and</strong> most of the returns occur in the future, the more likely they are to undertake<br />
those acti<strong>on</strong>s. A large market, a higher L , also raises the marginal return to entry<br />
<strong>and</strong> modernizati<strong>on</strong>. Finally, more industrial varieties mean that the past level of<br />
industrializati<strong>on</strong> is greater, thereby, reducing the marginal costs of inducing firms to<br />
adopt more modern technologies. Given this, the resp<strong>on</strong>siveness of firms to<br />
inducements by the government is enhanced when any of these parameters is raised.<br />
Therefore, the higher are these parameters, the fewer firms need to be targeted to<br />
facilitate an escape (from a development trap). Of these parameters has probably<br />
received the most discussi<strong>on</strong>. In many ways, this parameter represents the strength<br />
of increasing returns in the technology adopted by upstream producers. This is<br />
because higher levels of imply that, when they choose to modernize, upstream<br />
firms will choose technologies involving greater sunk (or fixed) costs. Therefore,<br />
while <strong>on</strong>e requires some degree of increasing returns or ec<strong>on</strong>omies of scale in<br />
producti<strong>on</strong> to generate a rati<strong>on</strong>ale for a “big bang” interventi<strong>on</strong>, the str<strong>on</strong>ger are<br />
those increasing returns to support a more unbalanced industrializati<strong>on</strong> policy. This<br />
relates back to arguments made <strong>on</strong> balanced vs. unbalanced growth. Of the three<br />
other parameters, <strong>on</strong>ly the discount rate seems to have been given a potential role<br />
in the past debate <strong>on</strong> industrializati<strong>on</strong> policy. Matsuyama (1992) interprets the<br />
discount rate as measure of effectiveness of entrepreneurship in coordinating<br />
investment, with a low discount rate indicating existence of greater entrepreneurial<br />
resources. If so, then the above result seems to imply that with a relative scarcity of<br />
entrepreneurial talent a more balanced approach should be followed.<br />
The comparative statics results for <strong>and</strong> require more restricti<strong>on</strong>s because each of<br />
these has two effects. On the <strong>on</strong>e h<strong>and</strong>, lowering <strong>and</strong> increasing raises the<br />
strength of strategic complementarities am<strong>on</strong>g upstream sectors. This tends to favor<br />
a more balanced growth approach. On the other h<strong>and</strong>, <strong>and</strong> each affect the<br />
marginal returns to entry <strong>and</strong> modernizati<strong>on</strong> of firms. The sec<strong>on</strong>d effect reinforces<br />
the first <strong>and</strong> leads to more balanced strategy that is, lowering <strong>and</strong> lifting <br />
increase the marginal returns to entry <strong>and</strong> modernizati<strong>on</strong>. A lower also implies<br />
str<strong>on</strong>ger technical complementarities. This effect is sometimes referred to as the<br />
returns to specializati<strong>on</strong> (Romer, 1987). The c<strong>on</strong>sequence is that a lower raises the<br />
marginal returns to employing greater variety of inputs in producti<strong>on</strong>. The higher is<br />
the weaker linkages am<strong>on</strong>g intermediate input sectors. C<strong>on</strong>versely, str<strong>on</strong>ger<br />
linkages between sectors raise the marginal return to targeting an additi<strong>on</strong>al sector<br />
for change supporting the arguments of the balanced growth strategy.<br />
10 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL. 3, ISSUE 1 ‐ MARCH , 2015, PP. 5‐13<br />
350
3.4 ESCAPING FROM DEVELOPMENT TRAPS<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
Looking at , it is a measure of the appropriability of the returns from supply as an<br />
additi<strong>on</strong>al intermediate input. As Romer (1994) discusses, the larger is , the greater<br />
is the surplus gained by intermediate input producers from the employment of their<br />
product in final goods producti<strong>on</strong>. Therefore, producers of inputs targeted in an<br />
industrializati<strong>on</strong> policy are more likely to react positively (in terms of adopting<br />
better technology) when the appropriable returns from the introducti<strong>on</strong> of their<br />
variety is larger. This effect would tend to favor a more unbalanced approach as <br />
increases.<br />
Summarizing, we have outlined the role of several parameters in influencing the<br />
kind <strong>and</strong> degree of balance in industrializati<strong>on</strong> policy. Factors addressed in the<br />
earlier literature such as strength of linkages, increasing returns <strong>and</strong> entrepreneurial<br />
resources all influence the compositi<strong>on</strong> of the ‘big push’. By c<strong>on</strong>sidering a ‘big bang’<br />
policy, some results are possible. For instance, str<strong>on</strong>g increasing returns in<br />
c<strong>on</strong>juncti<strong>on</strong> with weak sector linkages tend to favor a more unbalanced approach in<br />
order to minimize costs.<br />
CONCLUSIONS<br />
A major problem of the industrializati<strong>on</strong> debate is the timing of the industrializati<strong>on</strong><br />
policy <strong>and</strong> its degree of focus is complex <strong>and</strong> dependent <strong>on</strong> the characteristics of the<br />
case specific ec<strong>on</strong>omy. A ‘big push’ perspective <strong>on</strong> industrializati<strong>on</strong> does not imply<br />
that transiti<strong>on</strong> can be a simple matter of coordinating expectati<strong>on</strong>s via some kind of<br />
indicative planning. Nor does it mean that policy must be balanced <strong>and</strong> take a ‘big<br />
bang’ form in order to be successful. A wide variety of industrializati<strong>on</strong> policies can<br />
generate a ‘big push’ <strong>and</strong> the choice between them is therefore a matter of costs.<br />
In a dynamic model, however, this wide variety of industrializati<strong>on</strong> policies makes a<br />
characterizati<strong>on</strong> of the optimal policy quite difficult. To take advantage of full<br />
marginal modernizati<strong>on</strong> <strong>and</strong> entry costs, a gradual policy is always optimal.<br />
Moreover, in a policy of gradual entry, the number of sectors targeted in each period<br />
is rising over time. However, pairwise interacti<strong>on</strong>s between choice variables <strong>and</strong><br />
exogenous parameters tend to be qualitatively ambiguous in a dynamic setting. For<br />
instance, str<strong>on</strong>g increasing returns accompanied by weak sector linkages tend to<br />
favor a more unbalanced approach in order to minimize costs. The former effect<br />
favors the arguments of the balanced growth school, while the latter was part of the<br />
intuiti<strong>on</strong> of the unbalanced growth school.<br />
11<br />
351
HANS W. GOTTINGER<br />
3.4 ESCAPING FROM DEVELOPMENT TRAPS<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, Celia Umali<br />
Escaping from <strong>Development</strong> Traps: Industrializati<strong>on</strong> <strong>and</strong> Racing from the Bottom<br />
REFERENCES<br />
Barro, R. A. & Sala‐i‐Martin, X. (1995). Ec<strong>on</strong>omic Growth, Cambridge, Ma.: MIT<br />
Press.<br />
Easterly, W. (2002). The Elusive Quest for Growth, Cambridge, Ma.: MIT Press.<br />
Easterly, W. (2005). Reliving the ‘50s’: the Big Push, Poverty Traps <strong>and</strong> Takeoffs in<br />
Ec<strong>on</strong>omic <strong>Development</strong>, Center for Global <strong>Development</strong>, Washingt<strong>on</strong>, D.C., WP No.<br />
65, Aug. 2005, 1‐37.<br />
Fukuyama, F. ed. (2008). Falling Behind, Explaining the <strong>Development</strong> Gap between<br />
Latin America <strong>and</strong> the United States, Oxford: Oxford University Press.<br />
Gans, J. (1994). Industrializati<strong>on</strong> with a Menu of Technologies: Complementarities,<br />
Market Structure <strong>and</strong> Multiple Equilibria, Ec<strong>on</strong>omics Department, Stanford<br />
University.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H. (2006). Innovati<strong>on</strong>, <strong>Technology</strong> <strong>and</strong> Hypercompetiti<strong>on</strong>, L<strong>on</strong>d<strong>on</strong>:<br />
Routledge.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H. & Goosen, M. (2012). Strategies of Ec<strong>on</strong>omic Growth <strong>and</strong> Catch‐Up,<br />
New York: Nova Science.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H. (2014). Catch‐Up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic<br />
<strong>Development</strong>, Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business, 2(3), 67‐95.<br />
Hirschman, A. O. (1958). The Strategy of Ec<strong>on</strong>omic <strong>Development</strong>, New Haven: Yale<br />
University Press.<br />
Krugman, P. R. (1991). History versus Expectati<strong>on</strong>s, Quarterly Journal of Ec<strong>on</strong>omics<br />
106(2), 651‐667.<br />
Lin, J. Y. (2013). Against the C<strong>on</strong>sensus, Cambridge: Cambridge University Press.<br />
Matsuyama, K. (1992). The Market Size, Entrepreneurship, <strong>and</strong> the Big Push,<br />
Journal of the Japanese <strong>and</strong> Internati<strong>on</strong>al Ec<strong>on</strong>omies, 6, 347‐364.<br />
Murphy, K. M., Shleifer, A. & Vishny, R. W. (1989). Industrializati<strong>on</strong> <strong>and</strong> the Big<br />
Push, Journal of Political Ec<strong>on</strong>omy, 97, 1003‐1102.<br />
North, D. C. (1990). Instituti<strong>on</strong>s, Instituti<strong>on</strong>al Change <strong>and</strong> Ec<strong>on</strong>omic Performance,<br />
Cambridge: Cambridge University Press.<br />
Nurkse, R. (1952). Some Internati<strong>on</strong>al Aspects of the Problem of Ec<strong>on</strong>omic<br />
<strong>Development</strong>, American Ec<strong>on</strong>omic Review, 42, 571‐583.<br />
Nurkse, R. (1953). Problems of Capital Formati<strong>on</strong> in Underdeveloped Countries,<br />
Oxford: Basil Blackwell.<br />
Ortiz, C. H., Castro, J. A. & Badillo, E. R. (2009). Industrializati<strong>on</strong> of Growth:<br />
Threshold Effects of <strong>Technology</strong> Integrati<strong>on</strong> , Documentos de Trabajo,<br />
12 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL. 3, ISSUE 1 ‐ MARCH , 2015, PP. 5‐13<br />
352
3.4 ESCAPING FROM DEVELOPMENT TRAPS<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
Departamento de Ec<strong>on</strong>omia, A.A. 25360, Cali, Columbia.<br />
Romer, P. M. (1986). Increasing Returns <strong>and</strong> L<strong>on</strong>g‐Run Growth, Journal of Political<br />
Ec<strong>on</strong>omy, 94, 1002‐1037.<br />
Romer, P. M. (1987). Growth Based <strong>on</strong> Increasing Returns Due to Specializati<strong>on</strong>,<br />
American Ec<strong>on</strong>omic Review, 77(2), 56‐62.<br />
Romer, P. M. (1994). New Goods, Old Theory, <strong>and</strong> the Welfare Costs of Trade<br />
Restricti<strong>on</strong>s, Journal of <strong>Development</strong> Ec<strong>on</strong>omics, 43, 5‐38.<br />
Rosenstein‐Rodan, P. N. (1943). Problems of Industrializati<strong>on</strong> of Eastern <strong>and</strong><br />
Southeastern Europe, Ec<strong>on</strong>omic Journal, 53, 202‐211.<br />
Rosenstein‐Rodan, P. N. (1961). Notes <strong>on</strong> the Theory of the “Big Push”, In: Ec<strong>on</strong>omic<br />
<strong>Development</strong> for Latin America, H. S. Ellis <strong>and</strong> H. C. Wallich (Eds), New York: St.<br />
Martin’s.<br />
Rostow, W. (1960). The Stages of Ec<strong>on</strong>omic Growth, Cambridge: Cambridge<br />
University Press.<br />
Streeten, P. (1956). Unbalanced Growth, Oxford Ec<strong>on</strong>omic Papers, 167‐190.<br />
13<br />
353
HANS W. GOTTINGER<br />
3.4 ESCAPING FROM DEVELOPMENT TRAPS<br />
354
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
HANS W. GOTTINGER<br />
355
HANS W. GOTTINGER<br />
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
356
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
HANS W. GOTTINGER<br />
357
HANS W. GOTTINGER<br />
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
358
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
HANS W. GOTTINGER<br />
359
HANS W. GOTTINGER<br />
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
360
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
HANS W. GOTTINGER<br />
361
HANS W. GOTTINGER<br />
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
362
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
HANS W. GOTTINGER<br />
363
HANS W. GOTTINGER<br />
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
364
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
HANS W. GOTTINGER<br />
365
HANS W. GOTTINGER<br />
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
366
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
HANS W. GOTTINGER<br />
367
HANS W. GOTTINGER<br />
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
368
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
HANS W. GOTTINGER<br />
369
HANS W. GOTTINGER<br />
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
370
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
HANS W. GOTTINGER<br />
371
HANS W. GOTTINGER<br />
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
372
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
HANS W. GOTTINGER<br />
373
HANS W. GOTTINGER<br />
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
374
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
HANS W. GOTTINGER<br />
375
HANS W. GOTTINGER<br />
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
376
3.6 OPTIMAL ECONOMIC GROWTH WHEN CO2 CONSTRAINTS ARE CRITICAL<br />
HANS W. GOTTINGER<br />
377
HANS W. GOTTINGER<br />
3.6 OPTIMAL ECONOMIC GROWTH WHEN CO2 CONSTRAINTS ARE CRITICAL<br />
378
3.6 OPTIMAL ECONOMIC GROWTH WHEN CO2 CONSTRAINTS ARE CRITICAL<br />
HANS W. GOTTINGER<br />
379
HANS W. GOTTINGER<br />
3.6 OPTIMAL ECONOMIC GROWTH WHEN CO2 CONSTRAINTS ARE CRITICAL<br />
380
3.6 OPTIMAL ECONOMIC GROWTH WHEN CO2 CONSTRAINTS ARE CRITICAL<br />
HANS W. GOTTINGER<br />
381
HANS W. GOTTINGER<br />
3.6 OPTIMAL ECONOMIC GROWTH WHEN CO2 CONSTRAINTS ARE CRITICAL<br />
382
3.6 OPTIMAL ECONOMIC GROWTH WHEN CO2 CONSTRAINTS ARE CRITICAL<br />
HANS W. GOTTINGER<br />
383
HANS W. GOTTINGER<br />
3.6 OPTIMAL ECONOMIC GROWTH WHEN CO2 CONSTRAINTS ARE CRITICAL<br />
384
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
385
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
386
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
387
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
388
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
389
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
390
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
391
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
392
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
393
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
394
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
395
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
396
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
397
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
398
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
399
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
400
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
401
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
402
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
403
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
404
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
405
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
406
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
407
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
408
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
409
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
410
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
411
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
412
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
<strong>Technology</strong> Diffusi<strong>on</strong> 119<br />
Mansfield, Edwin, "Technical Change <strong>and</strong> the Rate of Imitati<strong>on</strong>". Ec<strong>on</strong>ometrica, Val. 29,<br />
1961, pp. 741-766.<br />
-, Industrial Research <strong>and</strong> Technological Innovati<strong>on</strong>. New York 1968.<br />
Manski, Charles, Daniel McFadden, Structural Analysis of Discrete Data with Ec<strong>on</strong>ometric<br />
Applicati<strong>on</strong>s. Cambridge, Mass., 1981.<br />
McFadden, Daniel, "C<strong>on</strong>diti<strong>on</strong>al Logit Analysis of Qualitative Choice Behavior". In: Paul<br />
Zarembka (Ed.), Fr<strong>on</strong>tiers in Ec<strong>on</strong>ometrics. New York 1973, pp. 105-142.<br />
Metcalfe, J. Stanley, "Diffusi<strong>on</strong> of Innovati<strong>on</strong> in the Lancashire Textile". Manchester School<br />
of Ec<strong>on</strong>omics <strong>and</strong> Social Studies, Val. 10, 1970, pp. 35-50.<br />
Nabseth, Lars, George F. Ray, The Diffusi<strong>on</strong> of Industrial Processes: An Internati<strong>on</strong>al Study.<br />
Cambridge 1974.<br />
Nati<strong>on</strong>al Science Foundati<strong>on</strong>, <strong>Technology</strong> in Retrospect <strong>and</strong> Critical Events in Science<br />
(TRACES). Washingt<strong>on</strong>, D.C., 1968.<br />
Nels<strong>on</strong>, Richard (Ed.), Government <strong>and</strong> Technical Progress: A Cross lndustry Analysis. New<br />
York 1982.<br />
-, Sidney G. Winter, An Evoluti<strong>on</strong>ary Theory of Ec<strong>on</strong>omic Change. Cambridge, Mass., 1982.<br />
Oster, Shar<strong>on</strong>, ''The Diffusi<strong>on</strong> of Innovati<strong>on</strong> am<strong>on</strong>g Steel Firms". The Bell Journal of Ec<strong>on</strong>omics,<br />
Val. 13, 1982, pp. 45-56.<br />
Peterka, V., Macrodynamics of Technological Change: Market Penetrati<strong>on</strong> by New Technologies.<br />
1.1.A.S.A. RR-77-22. Laxenburg, Austria, November 1977.<br />
Reinganum, Jennifer F., "<strong>Technology</strong> Adopti<strong>on</strong> under Imperfect Informati<strong>on</strong>". The Bell Journal<br />
of Ec<strong>on</strong>omics, Val. 14, 1983, pp. 57-69.<br />
Rosenberg, Nathan, Inside the Black Box: <strong>Technology</strong> <strong>and</strong> Ec<strong>on</strong>omics. New York 1982.<br />
-, Ralph L<strong>and</strong>au (Eds.), The Positive Sum Strategy. Washingt<strong>on</strong>, D.C., 1986.<br />
Scherer, Frederic M., Innovati<strong>on</strong> <strong>and</strong> Growth: Schumpeterian Perspectives. Cambridge,<br />
t\1ass., 1984.<br />
Spinrad, Bernard 1., Market Substituti<strong>on</strong> Models <strong>and</strong> Ec<strong>on</strong>omic Parameter. 1.1.A.S.A.<br />
RR-80-28. Laxenburg, Austria, July 1980.<br />
Telser, Lester G., "A Theory of Innovati<strong>on</strong> <strong>and</strong> Its Effects". The Bell Journal of Ec<strong>on</strong>omics,<br />
Val. 13, 1982, pp. 60-92.<br />
* * *<br />
Zusammenfassung: ök<strong>on</strong>omische Wahl und Verbreitung der Technologie auf den<br />
Märkten für neue Produkte. - Dieser Aufsatz versucht modellhaft zu entwickeln und zu<br />
schätzen, wie sich Innovati<strong>on</strong>en unter den Produzenten im Falle v<strong>on</strong> Spitzentechnologien<br />
ausbreiten. Nach der Einführ'ung einer Innovati<strong>on</strong> soll es annahmegemäß Firmen geben, die<br />
die Innovati<strong>on</strong> übernehmen, und Firmen, die sie imitieren; außer diesen beiden Firmengruppen<br />
gibt es keine <strong>and</strong>eren Marktteilnehmer. Die grundlegende Annahme ist, daß es einen<br />
inneren Zusammenhang zwischen dem Muster der Diffusi<strong>on</strong> und der Dynamik des Marktes<br />
gibt. Es wird unterstellt, daß eine "repräsentantive" Firma in einem gegebenen Markt vor die<br />
Wahl gestellt ist, ob sie sich auf dem Markt betätigen will oder nicht bzw. ob sie aus ihm<br />
ausscheiden will oder nicht, und daß die Wahlmöglichkeiten v<strong>on</strong> wirtschaftlichen Parametern<br />
413
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
414