Time to Take Another Look at Bayh-Dole
In a paper released recently, three fellows at the Brookings Institution identify promising policy ideas to encourage entrepreneurship and innovative growth in the technology industry. The recommendations draw on the expertise gathered from a workshop, organized by the Center for Technology Innovation at Brookings, as well as feedback from online crowd-sourcing. The following excerpt from the paper concerns technology commercialization.
One of the most important policy questions about innovation is how the nation can extract the maximum social benefit from its investments in research and development. The absolute level of R&D investments in the U.S. is by far the largest among all industrialized nations, exceeding $402 billion. Private industry remains the major funder of research and development, contributing to 62 percent of the national total. The public sector is the second major investor in R&D, contributing about 31 percent of the total. The commitment of government is even higher when looking only at research; in fact, 53 percent of the $40 billion invested in basic science is funded with tax-dollars.
This picture reveals a double objective for government. First, the government must create the environment where private and public investments in innovation are translated into productivity gains, jobs and economic activity. Second, public R&D investments must be made in the public interest so that they improve public health, strengthen national security and raise standards of living across all socioeconomic strata. In most circumstances, these two goals are one and the same. Still, they could sometimes be at odds and, when this happens, the challenge for government is to align them.
Commercialization and dissemination of publicly funded research are two domains where fostering innovation in the interest of the two government objectives—promoting economic activity and advancing the public interest—may conflict. The protection of intellectual property introduces the profit incentive to inventive activity, but this is a rather ineffective incentive if the innovator does not or cannot pursue profit. That is precisely the case of public R&D because the general expectation is that new knowledge created with taxpayers’ money may be made available to taxpayers at minimum cost. If new knowledge is inexpensively disseminated, profit-seeking enterprises will not seek its commercialization.
The expectation of a wide dissemination of public research has inspired a bill that is currently being considered in Congress. We should support the Federal Research Public Access Act that mandates public dissemination of federally funded research within six months of publication (for agencies with extramural funding exceeding $100 million). The bill proposes an exclusion of classified research, books from which authors receive a royalty and patentable discoveries. If some accommodation can be made to compensate for revenue lost by for-profit publishers of academic journals, we believe that this bill is consistent with the goal of pursuing widespread dissemination of the knowledge funded with public monies.
With respect to commercialization, the regime is governed by the Bayh-Dole Act of 1980 whereby the government allowed research contractors to take title to discoveries made with federal grants. This policy effectively converts a public good into a private one by assigning ownership. Bayh-Dole has made it simple for universities to patent (a few have even profited handsomely from licensing their patents) and in this way it contributed to streamlining the translation of research into new products. It is less clear, however, whether this act has always been effective in directing public research into the public interest.
An example of this ambiguity can be appreciated in the evolution of innovation and costs of healthcare. The revolutionary advances in biomedicine, largely derived from publicly funded science, have given new hope to patients suffering from medical conditions that not long ago eluded early diagnosis and effective treatment. But hope does not mean that more patients can access or afford the latest technologies of medical care or even afford older ones. In fact, the costs of healthcare and health insurance are rising much faster than the prices of general consumption excluding a large size of the population from healthcare.
Coordination among insurance companies and healthcare providers and asymmetries of information inherent to the provision of care are largely responsible for inflation in this sector. But innovation could play a significant balancing role. As a new generation of treatments enters the market, the older cohort should become less expensive. But there is no evidence that new discoveries are playing that role. Rather, it is quite possible that one unforeseen consequence of Bayh-Dole allows for modes of commercialization that have an inflationary effect on the whole healthcare system, not just new products.
This is plausible because established pharmaceutical companies can outbid smaller firms seeking to license a promising university patent. New companies will not rise to replace old established ones and any new technological platform in biomedicine will not produce the creative destruction that injects markets with dynamism. As a result, global pharmaceutical industry can set prices for final products above what would otherwise be competitive prices. Innovation strengthens the power of established companies and thereby innovation becomes complicit in keeping healthcare prices on the rise. It should be added that in a free market economy, firms are allowed to use their own resources to gain market power—in this case, research paid by the pharmaceutical companies themselves—but much of biomedical innovation is coming out of research done with public funds. If tax dollars fund an important part of biomedical innovation, it is not altogether unreasonable for the government to exercise some degree of control over pricing excesses. Yet, no such measures are currently in effect.
Looking into policy changes for the immediate future, it is important to build upon the successes of Bayh-Dole in fostering the translation of publicly funded research into economic activity. Nevertheless, to achieve the second objective of directing public research investments into the public interest, Congress should amend Bayh-Dole to promote the formation of competitive as opposed to monopolistic markets. A step in that direction is for the letter of the law to explicitly encourage the use of non-exclusive licenses. Without banning the use of exclusive licenses, an explicit government preference for non-exclusive licenses would increase the power of federal agencies to promote wider use of patents, regulate monopolistic practices with the products that result from them, and shift the weight-of-proof from the bureaucracy to the licensees to justify an exclusive patent.
Some changes could be made by executive order. Federal laboratories make a significant contribution to research-based innovation and could be directed to initiate “responsible licensing programs” modeled after similar programs tested at universities. These programs structure licensing contracts in creative manners, for instance, assigning exclusive licenses within specified jurisdictions, out of which non-exclusive licenses are allowed (e.g. an exclusive license could apply for OECD countries, while companies in developing countries can have it on non-exclusive basis). A licensing contract may also have a conditional exclusivity clause, where exclusivity expires if the licensee company does not commercialize the product in a manner consistent with wide dissemination of the patented product and certain agreed upon uses.
Agencies can also play a role in fostering competitive modes of innovation. Federal agencies may in various forms introduce incentives for the formation of patent pools, by increasing the scoring of a grant proposal that lays out a plan for patented discoveries to be part of an existing patent pool, or require all grantees of specified grant-purses to agree to patent discoveries and assign them to a patent pool. The joint ownership of a patent portfolio should address several concerns with the current regime. First, legal analysts have feared that patenting research tools may reduce the collective ability to innovate as scientists make regular use of tools developed by their peers. Second, licensing contracts must be negotiated in the best interest of all the pool’s partners, increasing the likelihood that patents will be used to create competitive markets of innovation rather than strengthening monopolies.
Schools need better metrics for measuring commercialization and technology transfer. Right now, there is too much focus on numbers of patents and startups without determining which products make it to the marketplace and what type of social and economic impact they have. We would be in a stronger position to address innovation roadblocks if we had more complete data on commercialization.
Darrell M. Vest is vice president and director of governance studies at the Center for Technology Innovation at the Brookings Institution; Alan A. Friedman is a fellow and research director; Walter D. Valdiva is a fellow at the Center.