Competitive Industrial Development in the Age of Information

Richard J. Braudo, Jeffrey Macintosh, Richard J. Braudo, Jeffrey Macintosh

• Publisher: Routledge
• Year: 2005
• ISBN:
  9781134683659
• Edition: 1
• Topic: Commerce
• Subtopic: Commerce Général

This book examines how transnational corporations, small to medium enterprises and governments have emerged as the principal players in industrial development. This valuable work examines this trend, with particular reference to the role of the tax policy in technology development, the financing of technology-sector SMEs, the role of government policy and the relationship between competition and co-operation.

Part I COMPETITION AS COOPERATION: Policy perspective

1 HIGH-TECHNOLOGY INDUSTRIAL POLICY IN CANADA: The future of an illusion or an illusion of the future

Robert Howse

INTRODUCTION¹

This chapter is intended as a critical analysis of a claim frequently made nowadays in discussions about the role of government in economic policy. Perhaps the clearest articulation of the claim that I have found in the literature is the following statement by David Crane, an economics journalist and prominent advocate of activist industrial policy:

If technology, in the form of innovation, ideas and knowledge, is a key factor in economic growth then it follows that governments can do a great deal to facilitate growth…. In addition [to policies that support education, training and basic research] governments have a vital role to play in facilitating pre-competitive research and development with industry and in assisting the actual commercialization of new goods and services. The risks are so high in the development of new technologies today that companies often cannot undertake these efforts on their own; similarly the spillover effects of new technology can be significant so that companies making the expenditures on technological development may not capture all or a large part of the benefits…. The importance of risk sharing between industry and government is well understood in Japan, the United States and Europe but much less so in Canada.²

Like most analysts of industrial policy (IP) who are influenced by the neoclassical economic paradigm, I have generally expressed deep scepticism about this kind of claim.³ However, I recently came across an article by one of the most thoughtful neoclassical economists of our time, Paul Krugman, in which he seemed to repudiate much of his earlier work on industrial policy, claiming that he and other neoclassical economists had significantly underestimated the significance of external economies, or externalities, and the potential, at least in theory, for industrial policy to internalize these externalities.⁴ In particular, Krugman has been impressed by the evidence presented by Michael Porter, among others, about the importance of geographical clusters in economic growth or development. This evidence demonstrates the extent to which individual firms do not stand or fall by their own efforts but depend on the mutually reinforcing effects of each other’s success.⁵ Nevertheless, Krugman concludes his paper with the caveat that, while stating a theoretical justification for industrial policy may be relatively easy, the ‘hard issues’ are those of choosing the right instruments and preventing new industrial policies from being hobbled by the same mistakes that beset the old ones.

I wish to begin more or less at the point at which Krugman ends. Accepting Krugman’s claim that state-of-the-art economic research supports the view that external economies are significant, what kinds of government policies are likely to be successful in internalizing them? Given that governments probably do not have the resources to internalize all external economies, and given that these external economies are relatively pervasive, how do we choose which to internalize? And, not of least importance, are the indicated policies likely to achieve their intended results in the Canadian context, given the historical experience of the political economy of industrial policy in this country?

EXTERNALITIES, PUBLIC GOODS AND FREE-RIDERS

Perhaps the most theoretically coherent rationale for government support of industries is that of positive externalities. Particularly in the high technology sector, it is argued, the innovative activity of one firm is likely to have significant spillover effects for other firms or other sectors. Advocates of IP argue that government action is necessary to internalize positive externalities and to ensure a socially optimal level of investment in firms that produce such externalities. They draw on an impressive and increasing literature on the positive linkages between firms and industries in successful regional or national economies.⁶

The IP literature often makes the error of identifying the existence of positive externalities as itself a demonstration of market failure. However, the more precise issue is whether significant externalities exist that cannot be internalized by a properly functioning market. When firms confer benefits on other firms, the value of these benefits will normally be internalized when the benefits can be proprietized and a market price charged to the firms that receive them. The market fails only where the firms that receive the benefits are able to free-ride.

In the case of innovation, or R&D, intellectual property rights are, of course, the principal legal instrument that addresses the problem of free-riding. However, not all the benefits that flow from research and development can be proprietized through intellectual property rights but only those that fall within legal definitions of what can be patented, copyrighted, and so on. While one response to the continued existence of free-riding could be to expand intellectual property protection, the current approach to such protection, with its winner-take-all characteristics, may lead to ‘racing’; that is, two or more firms devoting resources to the developing of a single innovation in order to be first across the finish line. Moreover, as Itoh, Kiyono and Okuno-Fugiwara note, ‘the patenting system involves an inherent contradiction in that while it raises ex ante R&D incentives by extending property rights, it also renders the diffusion of the developmental fruits, socially desirable in an ex post sense, difficult’.⁷

In addition, significant externalities may exist with respect to non-firm specific training and know-how that is conferred on workers. Firms may try to internalize as much as possible the benefits that flow from education and training through long-term employment contracts or covenants that restrict the capacity for workers to work for competing firms, but contract law rules themselves may significantly constrain such strategies.

The theme of this conference—cooperation and industrial policy—suggests that possible solutions to these problems may be found in cooperative arrangements among firms. For instance, if a particular R&D project promises to confer benefits on several firms but insufficient benefits on any one firm to justify that firm undertaking the project, the solution may be a cooperative arrangement of sharing of the project’s costs among the firms which stand to benefit. Under some circumstances, the possibility of joint gains may justify not simply a joint venture but a merger.

Where the gains of cooperation, or the positive externalities at issue, are broadly spread across firms and sectors, it is rather easy to see why collective action problems might exist that would prevent firms themselves cooperating so as to share costs and benefits and internalize externalities. Indeed, neoclassical economists have long recognized the ‘public goods’ character of basic education and highly generic R&D as well as of some training.⁸ However, where significant linkages exist among a relatively small number of firms or sectors, it is less clear why activist government policies would be needed to achieve the desired cooperative outcome. It is often argued that some forms of inter-firm cooperation may need exemptions from anti-trust or competition laws, as was the case with the American SEMATECH venture to be discussed by another panel at this conference. However, beyond removing this kind of legal obstacle to cooperation, it is unclear why government would need to be active, for example, by providing subsidies or research facilities.

One justification sometimes offered for a greater government role is that of information failures. Individual firms are prone to make investment decisions in ignorance of how these decisions may affect other firms. Since developing proprietary research makes secrecy important, firms may not disclose to each other enough information about their planned or existing activities to make clear the possible gains from cooperation. It is arguable that a government agency staffed with highly qualified analysts, scientists and engineers may be able to discern linkages and opportunities for cooperation to which individual firms are blind. The agency would have access to a vast amount of information about R&D and new technology being developed throughout the economy. This kind of role has often been attributed to MITI, the Japanese Ministry of International Trade and Industry.⁹ According to Porter,

MITI conducts or commissions countless study groups, industry committees, and reports concerned with new technologies, trends in international competition, and future issues. These are conducted with the input of the best Japanese experts, academics, high-level industry representatives, and government officials. The reports are broadly disseminated and publicized and are widely covered in the press. The major function of such studies is to awaken firms to emerging trends and problems.¹⁰

Another bar to cooperation between firms that captures positive externalities is the transaction costs of cooperative activity. Even when a cooperative activity yields joint gains, each firm may be concerned that the other will reap the larger gains from the results of the cooperation. If the consequence is that the other firm will gain a relative competitive advantage it did not have before cooperation, then a concern about ‘relative gains’ may dissuade some firms from entering into cooperative arrangements that promise to yield ‘absolute gains’ to all the participants.

Aside from concerns about ‘relative gains’, concerns about cheating may also deter firms from entering cooperative arrangements. For instance, firms may worry that they will end up giving competitors access to trade secrets or proprietary research. The most advanced firms in an industry or in a technology will be concerned that other firms will free-ride within the arrangement and so will be less likely to provide their best scientists and engineers. Indeed, the evidence of attempts at inter-firm collaboration in high-tech R&D in Japan suggests that distrust between firms has been almost universally present initially, and that only after sustained efforts has effective collaboration been possible.¹¹ Government involvement in the formation of such cooperative arrangements could be one response to the transactions costs posed by relative gains concerns and by the possibility of cheating. Partial government funding of research could be viewed as compensation to firms in the syndicate for the risk that the ultimate research result will end up primarily benefiting the other firms. Moreover, government may be able to intervene more directly, for example, by punishing or by threatening to punish firms that cheat, by withholding other government benefits or threatening to exclude them from future cooperative endeavours. It is quite conceivable that this latter role of government as arbitrator and enforcer could substantially reduce the transaction costs of cooperation.

Third, it is often argued that capital market imperfections may lead to underinvestment in high-risk, high-cost R&D-based ventures, even if cooperation succeeds in the capturing of significant external economies. Such ventures may simply be too risky given the risk-averseness of most investors. It is also claimed that managers of pension funds or mutual funds are under pressure to maximize short-term returns and will find investment in ventures that are based on high short-term costs, with returns taking many years to be realized, unattractive. Finally, it is claimed that the sheer volume of capital needed to undertake some high-tech ventures is simply not available on the market on any terms.

As an investor, government supposedly is not under the same constraints as capital markets. Because it can raise a large quantity of capital from tax revenues, it can supposedly do an end run around the risk-averseness of private investors. However, I am sceptical of characterizing the unavailability, or scarcity, of capital for very high-risk ventures as a market failure or imperfection. To say that investors do not want to take the risks is simply to say that, based on investors’ revealed preferences about risk, the costs of capital are prohibitive and the venture is not efficient. (Again, if the claim is that social efficiencies are not captured in the projected return to potential investors we are back to the externalities-based arguments.)

An important qualification of this statement would be if one could show that investors’ perceptions of risks or risk/return trade-offs are affected by heuristic biases. Some evidence in recent psychology literature, for example, suggests that people systematically overestimate risks where they are required to make choices about risk under conditions of considerable uncertainty.¹² Another qualification might be that, because of the large amounts of capital at its disposal, government can diversify risks in ways that are difficult or impossible for private investors. However, given the possibility of diversification through mutual funds or portfolio techniques of investment this claim appears implausible.

Finally, in some situations government may have lower agency costs of capital than private investors. Depending on the culture of government-industry relations in a particular society, firms and their managers may be prepared to disclose more information to governments than to private investors. The Japanese government is said to have enormous leverage that it can exercise over the companies it invests in. It is able to reward and punish firms and their managers in ways that would be all but impossible for any private provider of capital. In high-tech ventures, where considerable amounts of capital must be locked in for long periods of time before returns are realized, these informal kinds of ‘voice’ may be extremely important in controlling agency costs. Needless to say, however, exercising this ‘voice’ effectively depends on the government having the institutional capacity to monitor closely the ventures it invests in. Students of Japanese industrial policy argue that, in MITI, the Japanese government has assembled just such an institutional capacity. And, indeed, empirical evidence suggests that, in the case of high-risk ventures in which the Japanese government has invested considerable resources, MITI tends to micro-manage, or at least micro-monitor the venture, aggressively demanding changes in strategy or personnel if results are not to its satisfaction.¹³

WHAT’S SO SPECIAL ABOUT HIGH TECH?

Advocates of targeted high-tech industrial policies claim that the current range of policies and government resources devoted to supporting technological innovation is inadequate. Showing that externalities exist, as Krugman suggests, may not be difficult. However, showing that existing market institutions and government policies do not provide for sufficient internalization of these externalities is much more difficult. Besides, IP advocates are interested in convincing policy-makers, and the internalization of external economies as an end in itself is of little interest to most Canadians, although it undoubtedly excites professional economists. This leads most industrial policy advocates to sidestep the issue by adopting instead a comparative benchmark of how ‘we’ as a country are doing in comparison to others. Usually the claim is that we are falling behind some other country, whether Japan, Germany, or the United States, both in our policy responses to the high-tech challenge and in our overall economic performance. The citation of both claims together often substitutes for any causal analysis of how the second is the result of the first.

The crudest way of stating the claim that existing policies and institutions are inadequate is to compare Canadian spending on research and development or the technology-intensity of our exports to the ‘performance’ of other countries. This comparison almost invariably provides a grim picture of the Canadian situation. For instance, of the total export market shares of seven OECD countries in 1986, Canada made up only 3.5 per cent in the ‘High Intensity R&D’ category (as a point of comparison, the US share was 22 per cent). Canadian exports accounted for 5.5 per cent of the ‘Low Intensity R&D’ category with American exports accounting for 8.5 per cent.¹⁴ Using export receipts from the sale of technology itself, the Canadian performance is even more lacklustre—among the same group of OECD countries, Canada accounted for only 1.9 per cent of the total.¹⁵ Moreover, there is also a dramatic trend, beginning in the early 1970s, showing a decline in the use of domestic Canadian technology in Canadian manufacturing, with a drop of 27 per cent noted for the period from 1972 to 1984 alone.¹⁶ Total Canadian spending with respect to research and development (with private and government spending combined) has lagged behind that of several other industrialized countries, including the United States.¹⁷

To those Canadians dazzled by the ability to produce gadgets and contraptions, these figures lead to a sense of wounded national pride. However, they must be examined side by side with different figures that suggest Canada’s rate of economic growth and our export performance over much the same period have been relatively positive, including when compared to other countries whose industry is characterized by a much higher level of production of technology-intensive products. For example, during the 1975–1987 period, a time of greatly increased competition for export markets among industrialized countries and between NICs and industrialized countries, Canada was the only country besides Japan in the group of seven OECD countries to actually increase its share of world exports.¹⁸ Measured against members of the UN, a much larger group of countries, Canada’s share of total exports remained between 4 per cent and 5 per cent from 1950 to 1989, a period in which the US share declined from 18 per cent to 14 per cent and Japan’s share increased from 2 per cent to 8 per cent.¹⁹ In its overall levels of economic growth, during the 1980s and 1990s Canada has lagged behind other OECD countries at some times and outpaced them at others. If we use an even more general indicator of Canada’s economic performance, one which takes into account factors such as the quality of life that the Canadian economy is able to sustain, our comparative achievements appear even more impressive. Thus, in the United Nations Development Program Human Development Report, Canada has consistently ranked first or second among all countries in recent years.

That Canada’s comparative performance has been extremely good, however measured, while our dependence on domestic technology has declined and the technology-intensity of our exports remains low, justifies some initial scepticism as to the case for targeted high-tech industrial policies. This is not to say that, in absolute terms, Canada’s economic performance could not be improved. In particular, persistently high levels of unemployment in depressed regions of the country are a genuine problem. Yet it is in this last area that government policies have been most intensive but have largely failed in solving the problem. Another problem, also experienced by countries with much higher levels of government investment in high-tech industry than Canada, is chronic unemployment among unskilled or semi-skilled workers, especially young people.²⁰ Here, also, it is far from clear that a targeted high-tech industrial policy would result in the kind of economic growth that would put unskilled workers to work.²¹ Indeed, better internalization of technological externalities might actually accelerate the processes of technological change that result in capital’s replacement of labour, making the problem of unemployment among the unskilled even worse.

More fundamentally, the fact that Canada remains a successful exporter while its domestic production of high technology and technology-intensive products remains relatively low s...