The Challenge of Low-Carbon Development - World Bank Internet ...

technologies could accelerate diffusion of those technologieswithin particular middle-income countries, possiblywith high leverage in reducing CO 2emissions. Recentlyapproved World Bank projects to rehabilitate Chinese andIndian coal generators use this rationale.Choices at the country level—whether financed by the WBGor not—would be illuminated by systemwide analyses of expansionoptions. Such assessments would consider efficiencyoptions, assign costs to domestic pollution, and exploredifferent shadow prices for CO 2. Such analyses would clearlyshow when there are no domestically affordable alternativesto coal power and would help to defuse controversy. Thisapproach is consistent with that of the SFDCC, but with anemphasis of the additionality of WBG support in effectingpoverty reduction or technology diffusion benefits.Decisions about coal should usesystemwide analyses that considerefficiency alternatives, local pollution costs,and shadow prices of CO 2.Technology Promotion and TransferGreat hopes are pinned on technology, a cornerstone ofboth the Bali Action Plan and the SFDCC. Developingcountries hope not only to acquire hardware—such as windturbines and solar panels—but also to gain the capability tomanufacture and innovate, sparking industrial growth.At the global level, new technologies are conventionally understoodto follow a path from laboratory research, throughpiloting and technical demonstration, to commercial demonstration,and finally widespread deployment and diffusion,with continual improvements and innovations alongthe way. With increasing cumulative production, firmslearn and costs decline, tracing a learning or experiencecurve. This reflects the solution of technical problems andthe advantage of economies of scale (box 5.1).At the global level, technology costs declinewith increasing production.There is debate about where to draw the line betweenpublic and private support and between coordination andcompetition. There is general agreement, however, that expensivebasic research, such as that underpinning nuclearfusion, must be government supported. Public sponsorshipof pilot or demonstration plants, with data providedto all in the industry, also makes sense as a public good.The existence of a declining cost curve suggests that thereare increasing returns to concentrating resources in a fewtechnologies—a “big push” could produce a competitiveproduct. However, many worry that public sector groupsare ill equipped to pick winners in this manner.There is debate about where in thetechnology development cycle to draw theline between public and private support.Similarly, there has been a vigorous debate about the roleof intellectual property rights (IPRs) in energy and climatetechnologies. What is the proper balance betweenrewarding innovators and accelerating access to new ideas?A growing literature on this topic notes that patents, oreven trade secrets, are only one facet of technology transferand typically represent only a small proportion of energytechnology costs. Possibly more important are transfer oftacit knowledge and learning by doing.Complementing the global technology development cycle isthe process through which technologies diffuse across andwithin nations. The WBG has been active in this technologytransfer process. It encompasses piloting, where globallyavailable technologies are tested against and adaptedto local conditions; demonstration, to convince producers,investors, and users of the technology’s reliability and costBox 5.1Technology Learning (or Experience) CurvesMany studies have shown that manufacturing costs decline with an industry’s cumulative production. The reasonsinclude debugging and refinement of processes and economies of scale.Learning rates are expressed as the percentage decline in unit costs with each doubling of cumulative industryproduction. According to a review by Neij (2008), learning rates in renewable energy range from 2.5 percent forgeothermal and 5 percent for biofuel to 15 percent for wind and 20 percent for solar photovoltaics.These statistical results are useful for summarizing experience, but they do not tell us how learning works. Costscan decline as a result of true learning as manufacturers tune their equipment and procedures, research and development,economies of scale, or increased competition among producers or component suppliers. The rate of costdecline is not predetermined, but can be influenced through these different channels.Source: Neij 2008.Special Topics | 65