US Industrial Policies, R&D, And The WTO’s Definition Of Non-Actionable Subsidies

By Fred Block, Department of Sociology, UC Davis

The WTO framework, including the TRIPS agreement, has often been criticized for narrowing the permissible space for governments to pursue industrial policy. This is particularly true at the level of values; the current global trading system is based on free market ideas that celebrate the creativity of the private sector and denigrate the capacities of the public sector. However, commentators have also recognized that there is considerable “wiggle room” in the current trade regime that allows more scope for government policies than is often assumed.

But surprisingly, one of the most important areas of flexibility has received remarkably little attention. While the WTO rules are quite strict in treating most government subsidies to private firms as illegitimate and as justification for retaliatory action by trading partners, a very substantial exception is made for research and development subsidies. Trading partners have no grounds to object to these “non-actionable subsidies.”

The Uruguay Round Agreement on Subsidies and Countervailing Measures reads as follows:

8.2 Notwithstanding the provisions of Parts III and V, the following subsidies shall be non-actionable:
(a) assistance for research activities conducted by firms or by higher education or research establishments on a contract basis with firms if: the assistance covers not more than 75 per cent of the costs of industrial research or 50 per cent of the costs of pre-competitive development activity; and provided that such assistance is limited exclusively to:
(i) costs of personnel (researchers, technicians and other supporting staff employed exclusively in the research activity);
(ii) costs of instruments, equipment, land and buildings used exclusively and permanently (except when disposed of on a commercial basis) for the research activity;
(iii) costs of consultancy and equivalent services used exclusively for the research activity, including bought in research, technical knowledge, patents, etc.;
(iv) additional overhead costs incurred directly as a result of the research activity;
(v) other running costs (such as those of materials, supplies and the like), incurred directly as a result of the research activity.

This exception leaves considerable room for active government industrial policies designed to encourage private firms to innovate in both the development of new products and new production processes. By allowing subsidies to cover up to 75 percent of the cost of industrial research, governments can have a major influence on technology development by firms. Moreover, since “pre-competitive development research” is defined as everything that occurs before a firm produces a commercial prototype, states can play a major role in helping firms to transform new technologies into actual products.

Given the importance of US input into the Uruguay Round, it is hardly a coincidence that R&D subsidies to private firms have been a major element in the US government’s technology policies since the 1980s. In line with earlier successes with the computer industry and genetic engineering, funding agencies encourage scientists and engineers working in universities and government laboratories to create new firms to commercialize their breakthroughs. And it is generally seen as desirable that resources and ideas financed with federal research grants are then used by these startups. For example, the founders of Google began as graduate students – one of whom had funding from the National Science Foundation. The algorithm developed under that grant became the key piece of intellectual property for their newly created corporation.

Moreover, since the early 1980s, the US has had a program that focuses on supporting these startups. The Small Business Innovation Research Program (SBIR) gives out more than $2 billion per year to firms with fewer than 500 employees. It provides about three years of support – without strings – to firms that are making progress in commercializing a new technology. Private sector venture capital firms usually defer funding these small firms until they have gone through the SBIR process to move their idea closer to a commercial product.

It is not at all unusual that through SBIR and other federal programs, government might actually cover all of the initial costs for the startup firm for its first three or four years or even longer. Moreover, the Central Intelligence Agency and several other branches of government have their own venture capital firms that invest in promising startups. But once a firm does have a commercial prototype, it will usually be able to raise private sector capital in order to finance enhanced productive capacity. So the idea is that by the time the firm is ready to export, the government subsidy of development costs should be within the WTO rules.

But another feature of the US innovation landscape also makes it very difficult for potential competitors to prove that US R&D subsidies exceed the WTO threshold. It is part of the way the system is organized that government provides considerable in-kind support to private firms through publicly funded research facilities. The network of federal laboratories, mostly created during the Cold War, plays a central role here. Technologists on the federal payroll use the state of the art resources of these labs to help thousands of firms to overcome specific technological barriers. Sometimes private firms pay fees for research services provided by the federal laboratories, but it is unlikely that the charges are anywhere near what it would cost the firm to assemble comparable equipment and talent. Moreover, the details of these agreements between firms and federal laboratories – including the amount of money that changes hands – are not publicly disclosed.

The federal government also routinely funds the creation of research centres on university campuses that are intended to help a particular industry meet a specific technological challenge. Once established, some of these centres derive part of their funding from dues paid by industry, but here again, the benefits probably exceed the costs.

Two examples suggest the importance of this in-kind support. Starting in the 1980s, the Advanced Research Projects Agency of the US Department of Defense funded a laboratory in Southern California that would fabricate any chip design sent in by a legitimate researcher. This was a way of overcoming the bottleneck created by the high cost of fabricating chips; it opened up the industry to new ideas from graduate students and others outside of the dominant firms. More recently, the National Nanotechnology Initiative has built a series of expensive nano labs most of which are on university campuses. Private firms are encouraged to carry out their nano experiments in these facilities rather than building their own laboratories.

In addition to direct support for startups and in-kind subsidies, another tool that federal agencies use are R&D cost-sharing agreements with established firms. After lengthy consultation to identify a particular technological hurdle, the two parties agree to split the R&D costs over a specified time period to meet a certain technological objective.

The Department of Energy used this method in the early 1990s to get General Electric and Westinghouse to develop a new generation of high-efficiency, low-emission power plants that burn natural gas at extremely high temperatures. In kind support came as well through the funding of a university-based research centre dedicated to high temperature turbines. The project was successful and these power plants have been sold abroad. The Department of Commerce sponsored a similar initiative in the late 1990s with the Big Three automakers to upgrade the machining of parts for the industry.

As these examples illustrate, a robust strategy of industrial upgrading can be organized through these kinds of subsidies, In fact, a number of countries have very explicitly copied the SBIR program and China has been using all of these tools as part of its development strategy. Chile is another example of a country that has successfully used government-funded research to facilitate the successful upgrading of such industries as salmon farming and wine production.

But industrial policy through state R&D subsidies is a problematic development path for those less developed countries that have only a rudimentary science and technology infrastructure. When government budgets are extremely tight and basic human needs have not been met, it would be irresponsible for governments to devote resources to R&D subsidies that are inherently risky.

In short, use of this particular loophole in the global trading system is limited to countries that are already above a certain development threshold. Since countries below that threshold are also prohibited from using some of the strategies that helped rich countries develop in the past, such as export subsidies and reverse engineering of foreign products, the fairness of this regime is very much open to question.

But it is not just the fairness that needs to be questioned; the theoretical foundations of the current trading system are also problematic. The vision that animates the current trading system is the idea of the global economy as a level playing field where firms do not gain advantage by virtue of actions of their home governments. But how realistic is that vision if some of the world’s richest nations routinely use R&D subsidies to maintain their firms’ technological advantage in global trade?

The reality is that states and national economies are inevitably so intertwined that the level playing field vision for firms is neither feasible nor desirable. Just as they did in the past, nations will continue to use government as an instrument to improve the standard of living in a given territory. Trying to keep this from happening – in the name of some abstract free trade ideal – is like King Canute trying to hold back the tide.

The solution seems obvious. If richer nations are free to use R&D subsidies, then developing nations should be provided with compensatory concessions in the WTO framework such as more generous protections for infant industries, the opportunity to use export subsidies for limited time periods, and exemptions from some of the current TRIPS obligations. Moreover, as science and technology become ever more important to economic success, the richer nations have an obligation to help finance a science and technology infrastructure in developing nations that helps those countries meet their particular development challenges.

[Author’s Update: This language on non-actionable subsidies was agreed to provisionally for five years and its official renewal has been tied up in the Doha Round negotiations. However, in the ten years since the provision has lapsed, it does not appear that any research and development subsidies have been challenged in the WTO process except perhaps those involved in the longstanding US-EU dispute about government support for commercial airline production. In short, other nations continue to respect the strong US desire to maintain a “safe harbor” for R&D subsidies. The point remains that the least developed nations should be granted substantial concessions in exchange for tolerating subsidies that disproportionately benefit the most developed nations.]

Fred Block is co-editor with Matthew R. Keller of State of Innovation: The U.S. Government’s Role in Technology Development (Paradigm Publishers, 2011). He is a research professor of sociology at the University of California at Davis.