TIPS Pietrobelli IND POLICY Gaborone 9 Dec 2008.pdf

The Logic and the Method of IndustrialPolicy: Concepts and PracticesCarlo PietrobelliDirector of CREI, University of Rome 3, Italyemail: c.pietrobelli@uniroma3.iturl: www.pietrobelli.tkTIPS, Gaborone, 9 December 2008c.pietrobelli@uniroma3.itCREI, University Roma Tre1

‣ Final aim is always development promotion‣ Over time different paradigms prevailed in thinking how todo it…‣ … basic needs, capabilities, structural adjustment, “marketfriendly”environment, ….‣ … often only “cosmetic” changes and little consequences onactual policy strategies‣ Now “Industrial Policy” and “Private Sector Development(PSD)”! Following a simple logic: Private sector è growthè poverty reduction‣ Interest confirmed by several policy papers. E.g. EuropeanUnion Strategy for Africa (October 2005)c.pietrobelli@uniroma3.itCREI, University Roma Tre2

Aim of this seminar• Review/Discuss the theoretical foundations ofindustrial policy in the economics literature• Present a consistent and encompassing approach• Overview selected practices of industrial policy andPSD policies, to disclose and discuss their intrinsiclogic – with empirical evidence and cases of best,good and bad practices –• Draw implications for governments (and donors)c.pietrobelli@uniroma3.itCREI, University Roma Tre3

Definitions of Industrial Policy• Traditionally (market failures approach): identify and remedydistortions that prevent competitive market functioning, e.g. taxesand subsidies.• Heavier and widespread interventions (in theory) fell in disreputein many countries, due to inefficiencies and waste. Yet, oftenremained in place.• Now a revival of respectability…: “….Policies that may promotethe overall productivity of an economy. It has been proposed tocall them productive development policies (PDPs). The finalobjective is to raise growth….(IADB and Rodriguez Clare, 2008).• Rodrik and Harvard group: “A policy to speed up the process ofstructural change towards higher productivity activities, ….. And toassist firms in their search and discovery processes for new linesof comparative advantage” (Hausmann, Rodrik, Sabel, 2008)• Micro­foundationsof industrial policy (Lall and others)c.pietrobelli@uniroma3.itCREI, University Roma Tre4

The economic rationale for industrial policies1. Market failures vs. government failures2. The microeconomic foundations. TechnologicalCapabilities and market failures in technology andlearning3. When the “system” fails4. Coordination failures5. Who should identify and address these failures?6. New Industrial Policy: Search Networks and SystemIntegrators7. Examples of current practicesc.pietrobelli@uniroma3.itCREI, University Roma Tre5

1. Market failures vs. Government failures• Market is the optimal way of allocating scarce resources• Public policies to offset pre­existing market failures;Skeptical stand towards government intervention.• Typical market failures: externalities, market power, informationproblems (asymmetries, incompleteness), public goods, …. E.g.frequent for the market for R&D and knowledge creation.• Yet existence of market failures does not by itself establish thecase for intervention.– Design of interventions require information and skills,– Effective implementation requires autonomy, skills, impartiality.– Thus Governments may “fail”, and the related costs maymore than offset the benefits of intervention.• Ideological assumption that Governments are bound to failmore frequently in developing countries, and that Governmentsnever learn.c.pietrobelli@uniroma3.itCREI, University Roma Tre6

2. The microeconomic foundations. TechnologicalCapabilities and market failures in technology and learning• Neoclassical case against interventions based on a particularconceptualization of technology at the enterprise level ;• An alternative is the “technological capability (TC)”approach to understand innovation and learning in developingcountries. TCs are the skills­ technical and managerial – firmsneed to utilise technology efficiently and improve upon it.• This draws on the evolutionary approach, and on concepts suchas tacitness of technology, and technological efforts;• Thus, technological change is not exogenous butcomplementary to production. No essential difference betweenabsorbing, adapting and improving technologies and“breakthrough” innovation.• A learning process is always needed, and needs to be learnt.• The macroeconomic environment and the policy supportframework strongly affect the development of TCs.c.pietrobelli@uniroma3.itCREI, University Roma Tre7

1. …. TCs may explain differences in productionperformance (efficiency, productivity)“Production Function” (often a blackbox, the “firm”)ProductionFactors (L,K,Skills,organization, …)Outputs (producedwith different levels ofefficiency andperformance, ….):WHY?capabilities needed to carry out functions such as• (pre­)investment,• production (product­process­industrial engineering),• linkage with markets and technology, ….c.pietrobelli@uniroma3.itCREI, University Roma Tre8

1. Many Possible Categorizations of TCs (Lall)FUNCTIONSDegree ofCOMPLEXITYInvestmentProductionLinkages (withTechnology­Markets)BASICSIMPLE, ROUTINE(Experience based)INTERMEDIATEADAPTIVE,DUPLICATIVE(based on search,experimentation, inter­firmcooperation)ADVANCEDINNOVATIVE, RISKY(Based on research, purposiveeffort, advanced forms ofcollaboration)c.pietrobelli@uniroma3.itCREI, University Roma Tre9

3. When the “system” fails• Building on the idea of “… national, regional, local “systems” ofinnovation that influence the development, diffusion and useof innovation” (Freeman, Nelson, Lundvall, Edquist …);• Innovation is not sequential but the result of interactionsamong many actors within a system;• Institutions shape the actions and incentives of firms throughlaws, technical standards, public funding, social rules, …• Innovation is context­specific and interactive;• The flows of knowledge within the system are crucial (access tocomplementary knowledge is prerequisite for firms to innovate);• Differences between developed and developing countries…..Often focus on different organizations;• Policies need to address the failures of the system.c.pietrobelli@uniroma3.itCREI, University Roma Tre10

4. Coordination Failures• Production and investment decisions in upstream and downstreamsegments of industry are often interdependent;• Firms’ productivity also depends on the actions of other firms andorganizations influencing infrastructures, intermediate goods andpublic goods provisions, regulations,….• When market failures emerge in these markets for (intermediate)goods and services, the economy could be trapped in a lowinvestment equilibrium;• Rosenstein­Rodan’s notion of underdevelopment traps: no sectorwould be profitable industrializing alone.E.g. Build an airport with no hotels, train personnel in fashion design with no firms….• Failure to coordinate individuals’ actions leads to an equilibrium that isworse for everyone than an alternative equilibrium where manysectors were industrializing simultaneously• Coordination failures frequently occur at the local (cluster) level, but itis right in clusters that such failures may be addressed.c.pietrobelli@uniroma3.itCREI, University Roma Tre11

5. Who should address/identify these failures?Old and difficult question• Not anymore an alternative between two extremes:vs. Market, or Government vs. Private SectorState• In fact, greater reliance on markets needs a more proactive rolefor the government (not vice versa)• The issue is: How to exploit the best capacities of both?• Build a clever, dynamic and pragmatic partnership – inducingsmart and effective forms of collaboration.• Policies to promote industry and the private sector should beseen as a process of economic self­discovery in a broadsense, with an interactive process of strategic cooperationbetween the public and private sectors to elicit information onbusiness opportunities and constraints, and also generate policyinitiatives in response.• Partnerships at the national as well as at the local levelc.pietrobelli@uniroma3.itCREI, University Roma Tre12

6.”New Industrial Policy”. Search Networks andSystem IntegratorsSomehow along similar lines the approach proposed by Rodrik• Aims at “…. Solving economic development problems withoutpicking winners”.• Focus is on– bridging private­public public institutions (Fundación Chile)– Business networks linking global and local (i.e. search networks)like innovation clusters & value chains, diasporas)– Best performers in public and private sectors.• Target to generate missing connections – without allowingrent­seeking – and start a process of discovery have lowenough cost to be profitable.• System integrationSystem integration: move from good programs to goodsystems (Tekes, Corfo, …)c.pietrobelli@uniroma3.itCREI, University Roma Tre13

“Best”, “Good”, “Bad” PRACTICESAnalysis of several approaches followed in specificprojects (not a detailed evaluation)1. Enterprise clusters as a tool for industrial policy2. Domestic and international inter­firm linkages3. Scientific and Technical support organisations4. Business Development Services5. Entrepreneurship PromotionSome selected ideas on these approachesc.pietrobelli@uniroma3.itCREI, University Roma Tre14

1. Enterprise clusters as a toolDrawing from earlier research (Pietrobelli and Rabellotti for HUP2007 and others….)• Economic performance may be improved in clusters due tocollective efficiency (external economies and joint actions)• Clusters may help remedy market and coordination failures• Policies need to be context­specificand – in several regards– cluster­ and sector­specific. Take into account:ülocal specificities and cluster’s collective efficiency,ümode of governance of value chain(s),üLocal sectoral and cluster (filière) specialization• No general recipes are valid everywhere, regardless localhistory, idiosyncrasies and peculiarities.• Policies need to evolve over timec.pietrobelli@uniroma3.itCREI, University Roma Tre15

On Evolution of Policies: the ChileanSalmon ClusterTremendous performance:Þ From 0 to 25% of world salmon farmingÞ Exports: 1985 US$ 1 mill., 2002 US$ 1,000 mill.Policies have evolved over time1978­85 “Initial learning”: regulation, technology transfer,investment in pre­competitive research1986­95 “Maturing”: physical infrastructure, exportpromotion and marketing, innovation and development ofsuppliers (cages, nets, food)1996­today: “Globalization”: productivity increase andtechnology transfer, environmental management,biotechnology (diseases and genetic handling)c.pietrobelli@uniroma3.itCREI, University Roma Tre

YearChilean Salmonand TroutProduction (.000tons)% Chile inWorldProductionEstimatedProd. In 10thRegión (.000tons)Exports(US$ millionFOB)1987 2 1.5 % 1 8.01990 26 7.9 % 20 140.01993 69 17.1 % 60 291.41996 165 24.4 % 150 525.01999 201 22.1 % 180 817.82001 405 32.4 % 350 964.3Fuente: APSTCH, SalmonChile, diversos años.c.pietrobelli@uniroma3.itCREI, University Roma Tre

Table 2.2 Stages in the Life Cycle of the Salmon Cluster and Evolution of PolicyFrameworkINITIAL LEARNING1978­85from 50 tons to 900 t.price: US$ 9­10/kgMATURING1986­1995from 1,350 t. to 143,000t.price: US$ 4­5/kgGLOBALISATION1996­2003from 150,000 t. to 400,000 t.price: US$ 2.8­4.5/kgMain objectiveSurvival of the fishIncrease volume ofproductionIncrease productivityMarketDestinationProduct: fresh Pacificcoho (élite market)Pacific coho, frozen,mainly to JapanFillets and portions to the USmarket diversification: Japan, US,niches in Europe and emergingmarketsCommercialization channelsDirect sale andcooperativesBrokers, associativechannel of nationalproduction ­SalmoexportWholesalers (supermarkets)Strategic alliances or integrationwith final retailersTechnologicalchallengeExperimental fishfarmingKnow­how infatteningBackward linkages (fishfarming)Quality certificationIncreased productionscaleNational eggs production, Forwardlinkages (process), Salmondevelopment cycle, systems ofautomated control of water, light,etc., vaccines and food,Sustainability of the whole systemc.pietrobelli@uniroma3.itCREI, University Roma Tre

INITIAL LEARNING1978­85from 50 tons to 900t.price: US$ 9­10/kgMATURING1986­1995from 1,350 t. to 143,000t.price: US$ 4­5/kgGLOBALISATION1996­2003from 150,000 t. to 400,000 t.price: US$ 2.8­4.5/kgPublic policies• regulation• transfer oftechnology• investment andpre­competitiveresearch• Physicalinfrastructure• promotion andmarketing (missions)• innovation andtechnologydevelopment ofsuppliers (cages,nets, food)• environmental handling• increase of productivity andtransfer of technology(technology missions)• biotechnology (diseasesand genetic handling)Type of companywithin theclusterSMEsSMEs withPresence of foreign groupsLarge enterprises (integration andconcentration), some TNCs.Type ofsupplyingcompanyFew and precarious.Companies seek selfsufficiencyMajor outsourcing, localcompanies gainprofessionalismSpecialized local SMEsLarge presence of highly specializedTNCsExternalitiesSocial capital –setting“Demonstration effect”Pioneering public andprivate effortsSource: Maggi in Pietrobelli and Rabellotti, 2007.Access to suppliersCritical mass achievedAssociative will betweenproducersDissemination of good practicesProductive system inserted in a globalproduction chain – commercializationc.pietrobelli@uniroma3.itCREI, University Roma Tre

2. Domestic and international inter­firm linkages• ….. Global Value Chains and local firms• Foreign Direct Investments, TransnationalCorporations and Spillovers on domestic firmsc.pietrobelli@uniroma3.itCREI, University Roma Tre20

3. Scientific Research and Technical SupportOrganizations• In the present context of globalization and rapidtechnological changes, countries need to create andupgrade new skills to manage technical change,• New role for technical support organizations in standards,metrology, quality, testing, R&D, productivity, SMEextension to complete and improve the “technologysystem”• Industrial policies in advanced countries are often trulyScience, Technology and Innovation (S&T&I) Policies• The practice of the European Union• Industria 2015 in Italy (and similar efforts in France andGermany)c.pietrobelli@uniroma3.itCREI, University Roma Tre21

4. Entrepreneurship PromotionDrawing from recent comparative studies of entrepreneurshippromotion programs (Kantis, 2005 IADB).• no single prescription for success: strategies are alwayscontext­dependent; significant differences in the strategic,geographic, and budgetary reach;• Knowledge of the initial conditions is essential;• common to combine generic (national and sectoral) and nichebasedstrategies (the latter oriented to local and socialgroups);• Sustainability crucially depends on involvement of the privatesector and of civil society;• The style of interventions must be itself “entrepreneurial”;• A flexible strategy requires efficient monitoring and evaluation.c.pietrobelli@uniroma3.itCREI, University Roma Tre22

5. Business Development Services (BDS) or real services, fortheir impact on structural features of company behaviour, and on theircompetitiveness‣ no easy recipes to copy, no ‘ideal’ best practice.‣ BDS centres have a role, but the market may do a lotwithout public subsidies.‣ Embeddedness in the local business environment iscrucial (deep involvement of private sector, sectoralspecialisation, location close to potential customers).‣ Managerial and technical skills in the Centres andcapabilities are key‣ The density of their presence matters.c.pietrobelli@uniroma3.itCREI, University Roma Tre23

5. Business Development Services (BDS)‣ A role in stimulating demand of new services from firms.This requires anticipating tacit, unexpressed needs ­incollaboration with firms­ and convincing them of theirrelevance for future competitiveness‣ An alternative: help BDS Centres act as “network networkfacilitators”;‣ But in poorer countries, a Centre is often bound tooperate on its own, and should first improve its skills andcapabilities;‣ Evaluation is difficult but necessary (quantify benefits,costs, impacts, and repeat evaluations regularly).c.pietrobelli@uniroma3.itCREI, University Roma Tre24

Concluding: Implications (1/3)‣ A process (dynamics not statics)‣ a mindset?: not a big­bang, once for all reform: trialand error with continuous efforts and learning‣ experimentation, gradual transformation through selfcorrectionand identification/removal of obstacles.‣ Public­private dialogue and active collaboration foreffective policies (e.g. Fundación Chile)‣ there is still “policy space” in the WTO regimec.pietrobelli@uniroma3.itCREI, University Roma Tre25

Concluding: Implications (2/3)‣ Starting point is always the firm­level.Analysis of firmlevelweaknesses – notably in innovation and learning– should drive policy­makers and donor agencies.‣address market and coordination failures, but also systemicimperfections” in markets, and notably in knowledge &technology, i.e. target components as well as theirrelationships.‣Target the coherence of the whole system of policysupport. Beyond good “projects” to good “systems systems” ofconsistent and integrated programmes and interventions(e.g. CORFO)‣ Improve capabilities for strategic policy design,formulation, implementation.c.pietrobelli@uniroma3.itCREI, University Roma Tre26

Concluding: Implications (3/3)What concrete practices to improve the effectiveness ofpolicy interventions, of their design and delivery?‣ Benchmarking with other experiences. This may also providefirms and organizations with the incentives to act andimprove.‣ Starting point is the firm­level.Analysis of firm­levelweaknesses – notably in innovation and learning –.‣ Experiment with indirect inducements instead of directinterventions, i.e. building appropriate institutions andincentive mechanisms‣ Open dialogue, transparency, accountability and constantevaluation in policy design and implementation (….minimizecorruption and private capturing of all benefits)c.pietrobelli@uniroma3.itCREI, University Roma Tre27

Thank you !!!!P rof. Carlo P ietrobelliDirector of CREI,University of Rome 3, Italyc.pietrobelli@uniroma3.itw w w .pietrobelli.tkc.pietrobelli@uniroma3.itCREI, University Roma Tre28

Questions for group work andpresentations by participantsIs the idea of cluster used in your country?Is the idea used for (industrial) policy purposes?What is the approach to cluster development followed inyour country (if any)?At what level do you work? Enterprises, networks ofenterprises, support institutions, policy level, ..?What are the motivations and objectives of your clusterdevelopment activities? For example: SME development,poverty alleviation, new firms creation, export, etc.c.pietrobelli@uniroma3.itCREI, University Roma Tre29