Innovation Policy and the Economy: An Overview

The National Bureau of Economic Research (NBER) Innovation Policy and the Economy (IPE) grogroup seeks to provide an accessible forum to bring the work of leading academic researchers to an audience of policymakers and those interested in the interaction between public policy and innovation.  Our goals are:
 

• to provide an ongoing forum for the presentation of research on the impact of public policy on the innovative process;
• to stimulate such research by exposing potentially interested researchers to the issues that policymakers consider important;
• to increase the awareness of policymakers (and the public policy community more generally) concerning contemporary research in economics and the other social sciences that usefully informs the evaluation of current or prospective proposals relating to innovation policy.

The group sponsors a meeting in Cambridge each summer, greared to presenting the latest academic research on these topics, as well as a meeting in Washignton each April designed for policymakers.  The papers from each April session are published in an annual periodical by the MIT Press.  The first two volumes in this series are summarized below.

Summary of Volume 1:

The papers in the initial volume demonstrate the importance of issues related to innovation in current policy debates, the value of the insights that economic analysis can bring to these problems, and the breadth of interest of economics in innovation-related issues.  The papers were presented in the workshop held in April 2000.

The first two papers highlight the interaction between public policy and innovation in a specific but important sector -- the life sciences.  Motivated by the extraordinary rise in public expenditures on the life sciences over the past thirty years, Iain Cockburn and Rebecca Henderson initiate the volume by assessing the relationship between public investment in life sciences research and the rate and extent of innovation in the pharmaceutical industry.   Though the social benefits provided by basic research are notoriously difficult to quantify, economic research on the industry has identified several specific mechanisms through which public funding may spur innovation and the commercialization of new therapies.  For example, public funding of basic research in molecular biology has provided critical elements of the foundation of “rational” drug design -- a more efficient drug discovery technique whereby researchers investigating new compounds are guided by scientific evidence about the biochemical basis of disease.  Building both on case histories of specific drugs as well as evidence based on patenting and publication data from the industry, the paper assesses the returns to public expenditure in this sector.  While noting the limitations and assumptions associated with individual studies, their review suggests that prior econometric research “makes a quite convincing case for a high rate of return to public science.”

Michael Kremer's paper, on the other hand, examines an arena where private pharmaceutical companies do very little research: the development of vaccines for tropical diseases. He argues that the reluctance of pharmaceutical companies to undertake research to develop vaccines for diseases such as malaria, tuberculosis, and African stains of HIV is a consequence of the severe market failures. In particular, companies fear that were they to develop such products, governments would force prices down to a level that would not allow them to earn a satisfactory return. To address this problem, he proposes that public agencies commit to buying vaccines at a set price if a satisfactory vaccine can be developed by the private sector. Kremer suggests that such an initiative could address market failures far more effectively than alternative approaches, such as government subsidies for basic research in this area.

Carl Shapiro focuses on the increasingly important issue of the interaction between intellectual property protection and competition policy.  He notes that in several important technology areas, such as biotechnology, semiconductors and software, commercial innovation often requires use of numerous potentially overlapping or conflicting patent rights.  A particularly important circumstance in which such problems often arise is where standard setting is an essential part of the process by which new technologies are commercialized.  Parties can use a variety of contractual mechanisms to resolve these problems, including cross licensing, package licensing and patent pools.  Antitrust authorities, however, sometimes challenge such agreements, particularly when they occur between or among horizontal competitors.  Shapiro offers several suggestions as to how agreements that facilitate innovation can be distinguished from those with serious anti-competitive effects.

Shane Greenstein’s paper asks the question: why did the commercialization of the Internet go so well?   This question is intriguing and also important, because most historical examples of technology transfer from the public to the private sector have involved confusion and delay.  In contrast, the technologies that make up the Internet diffused rapidly and pervasively shortly after the National Science Foundation relaxed the regulations restricting private use in 1992.  Drawing upon his own research on Internet Service Providers (“ISPs”), Greenstein identifies four drivers of this unusually quick diffusion process:  (a) the absence of significant technical or commercial hurdles; (b) the economic and technical malleability of the Internet; (c) the potential for customization on a number of key dimensions and (d) the fortuitous coincidence that the Internet was commercialized at the same time that the World Wide Web was developed.  The diffusion of the Internet yields two policy insights.  First, the Internet experience may allow for better identification of the conditions under which technology may be transferred successfully from the public to the private sector.  Second, the Internet has had a ubiquitous impact on the telecommunications industry in particular, perhaps necessitating a rethinking of the regulatory institutions underlying this sector of the economy.

The last two papers in the volume examine the different facets of the relationship between academic institutions and innovation. The first of these, by David Mowery and Arvids Ziedonis, examines a direct impact: the commercialization of academic discoveries by the private sector. Drawing on in-depth studies of three leading research universities--Columbia, Stanford, and the University of California--the authors examine the impact of the policy reforms of the early 1980s on technology transfer activities. Disentangling the impact of this policy change from the contemporaneous shifts in federal funding of research and patent policy is challenging. But the results suggest that, at least at these three schools, the reforms of federal technology transfer policy served as a significant boost to commercial activities.  This suggests that the Bayh-Dole Act of 1980 is helping to achieve its stated goal of increasing the transfer of commercially useful technology from universities to the private sector.

Although commercialization of university technology is important, most scholars agree that the primary contribution of academia to commercial innovation is its training of scientists and engineers.  In the last paper of this volume, Paul Romer notes that, in principle, government policy to foster innovation could act on both the demand and supply sides of the R&D investment process.  He argues that, historically, policy has focused more on the demand side, using instruments such as R&D tax credits or other subsidies.  If the supply of R&D resources—primarily technically trained people—adjust only very slowly to changes in policy, then such policies will raise the wages of scientists and engineers without increasing research much.  Romer then analyzes how the structure and incentives of post-secondary and graduate education in the U.S. affect the supply of technically trained people to the commercial sector.  He concludes that undergraduate education discourages students from majoring in science or engineering, and that post-graduate programs are structured to produce Ph.D.s oriented towards teaching rather than commercial research.  He then considers several government policies that might counter these tendencies, thereby increasing the supply of scientists and engineers and ultimately the rate of commercial innovation and economic growth.

Summary of Volume 2:

The first two papers seek to define appropriate government policies in high-technology industries.  David S. Evans and Richard Schmalensee analyze the implications for antitrust analysis of the dynamic “winner-take-all” competition characterizing many markets where technological change is rapid and drastic innovation frequent.  As emphasized by Joseph Schumpeter over 50 years ago, the nature of competition in such markets is that successful innovators achieve only temporary market power.  Indeed, it is the acquisition of such temporary market power that provides the incentive for innovative investment.  Successful firms are subject to loss of their market position to subsequent innovators, however, so the use of temporary market power may not be a cause for policy concern.

Evans and Schmalensee show how the possibility that current market leaders may be displaced in the next round of innovation can lead to new conclusions about the appropriate scope of antitrust analysis.  .  For example, conventional tests for predatory behavior and tying can lead to misleading conclusions.  Practices labeled as monopolistic under traditional analysis are an inherent part of the competitive process in Schumpeterian industries.  Evans and Schmalensee conclude that appropriate competition policy in Schumpeterian industries requires the development of new models and analytical tools that address when consumers will benefit (or not) from antitrust policy intervention.

The second paper considers intellectual property policy.  For many decades, economic theorists assumed that the relationship between intellectual property rights and innovation was straightforward: stronger property rights would generate more innovations.  In the past decade, the literature on “sequential innovation” has raised important questions about this reasoning.  In many cases, generations of innovations are linked to one another: one discovery builds on another, which in turn relies on earlier work.  When innovation is a cumulative process, too broad an award to one innovator may discourage innovation by followers.

In this paper, two of the leading contributors to this literature, Nancy Gallini and Suzanne Scotchmer, discuss these recent works.  They review the recent writing on this topic, and highlight the implications of the “sequential innovation” models.  They also explore the question of whether radically redesigning the nature of the awards given to innovators might boost innovation.

In the third paper, Manuel Trajtenberg considers more direct interventions: when government does not just set the “rules of the game,” but directly funds research and development.  He uses the science and technology policy of Israel as a case study to explore several core issues related to such policies.

For several decades, Israel has had a systematic policy designed to foster commercial R&D.  This policy was founded on technological “neutrality,” i.e. the government did not attempt to choose what technologies to support, but rather provided a relatively generous subsidy to any firm proposing a commercial R&D project that met certain basic qualification criteria.  Trajtenberg describes the experience with this program, and the evidence regarding its impact on overall R&D and innovation.  He also discusses the tensions that have arisen in recent years as the government’s desire to impose a budget constraint have come into conflict with the principle of neutrality.  He concludes that Israeli policy appears to have been generally quite successful in being flexible and responsive to changing circumstances.  At the same time, the Israeli high-tech sector has enjoyed a remarkable boom, though it is unclear how large a role the public sector policies played in this success story.

This paper shows that ideological aversion to government support of commercial technology and ideological commitment to solving market failures through government technology programs are both equally unhelpful.  Needed are a careful analysis of what works and what doesn’t and a quantification of the economic effects of different policies.  Trajtenberg shows that in Israel, at least, flexible, technology-neutral support of commercial R&D appears to foster high-tech growth, though we are far from being able to determine whether the social returns to this growth justify the investments made.

The last two papers in the volume address the character and policy consequences of the “New Economy.”  Timothy Bresnahan offers a synthetic framework for evaluating the consequences of information technology for economy-wide productivity growth.  Motivated by recent theoretical and empirical advances, Bresnahan argues that traditional conceptualizations of the value arising from innovation are flawed.  First, the value arising from information technology is sometimes due to network effects, and so consumer welfare increases with the extent of usage across the population.  Second, in an even larger number of contexts, the benefits from information technology are only realized if there is “co-invention” – complementary innovation by users of information technology.  Bresnahan argues that these two factors account for a very important portion of the total returns to IT, with implications for long-term growth and economic policy.  When co-invention is important, the economy-wide benefits from information technology accumulate slowly and unevenly.  As a result, the rate of productivity growth at a given point in time critically depends on prior innovation levels and on the ability of co-inventors to extract the value from these prior inventions.  While most areas of innovation policy – from intellectual property policy to tax credits to antitrust – focus almost exclusively on providing appropriate incentives to initial innovators, Bresnahan’s analysis suggests that equal attention must be paid to the incentives and resources available to co-inventors.  For example, it may be appropriate to substantially broaden the definition of activities that are covered under the R&D tax credit in order to induce a higher rate of investment in the downstream activities so critical to realizing the benefits from information technology advances.

The final paper builds on these themes, assessing whether the microeconomic evidence “adds up.”  Brad DeLong links the diffusion of telecommunications and information technology to the most striking changes in the economy as a whole over the last decade.  Recognizing that macroeconomic volatility is still an ever-present threat, DeLong evaluates whether the circumstances associated with distinct drivers of macroeconomic fluctuation can themselves be tied to technological change.

He makes three key arguments.  First, DeLong argues that the five-year-old resurgence in the productivity growth rate can indeed be tied to technological advance and is likely to be sustained.  Recent empirical studies highlight a key reason for optimism: information technology now accounts for a large enough fraction of total capital services to impact aggregate productivity growth.  Second, this boost in productivity growth may have helped to both lower the natural rate of unemployment and increase the sensitivity of inventories to changes in aggregate demand.  While the evidence is not conclusive, higher productivity growth and lower search costs may have therefore provided policymakers with increased flexibility in monetary and fiscal policy.  Finally, on a more speculative note, recent patterns of technological change may have increased the volatility of financial markets.  Consequently, though the “new economy” seems to have offered increased macroeconomic stability up to this point, policymakers must address themselves to the new policy challenges resulting from these structural shifts in the economy.