By Ryan Abbott
In Beyond the Patents-Prizes Debate (forthcoming in the Texas Law Review), Daniel Hemel and Lisa Larrimore Ouellette articulate a new theoretical framework for thinking about R&D funding mechanisms. They note that patents, prizes, government grants, and tax credits for research all play an important role in facilitating innovation, although academic literature has disproportionately focused on the role of patents and prizes—this is the first detailed analysis of how tax credits compare as an incentive mechanism.
The Orphan Drug Act (ODA) illustrates how patents, grants, and tax credits can all be used to achieve similar goals. The ODA was passed in 1982 to incentivize R&D on rare diseases. It incorporates several different types of incentives, including direct government grants for research, a patent-like market exclusivity period for approved drugs, and tax credits for half of the cost of clinical research. The ODA has been widely acknowledged as successful. From 1973–1983, U.S. pharmaceutical companies only brought 10 orphan drugs to market. In the twenty-five years since the Act’s passage, 326 new drugs were approved for rare diseases. The ODA was also the model I proposed in an article last year for a new system to boost R&D for evidence-based complementary and alternative medicine (CAM).
Given that all these mechanisms can accomplish the same goal, possibly even with the same costs, and no single mechanism is always more efficient than the others, a framework is needed for determining which incentive is most appropriate for a given goal. The authors do this by providing a structure for balancing the benefits and detriments of each incentive on the basis of who should determine the reward size, when the reward should be provided, and who should pay.
For example, patents and credits generally rely on market decision makers, as opposed to prizes and grants, which rely on government priority setting. This suggests that patents and credits may be more efficient when the government has less information than private parties. The authors also explain that while patents and prizes provide strong incentives for success, inventors may be more responsive to the earlier and more certain rewards provided by tax credits and grants due to capital market frictions and risk aversion.
In terms of who pays, patents are unique among these incentives in that the cost of patents is borne primarily by consumers of the patented goods, while prizes, grants, and credits are usually funded by the general public. Which is preferable will depend on one’s distributive preferences, but the authors argue that “user pays” mechanisms such as patents might be more appropriate for incentivizing luxury goods, and less appropriate for incentivizing goods of great social value with poor market demand such as vaccines for malaria. In addition, while the cost of prizes, grants, and credits are entirely shouldered by government, because of the TRIPS agreement, patents end up subsidizing U.S. research by spreading the cost of domestic innovation to other countries.
Patents are also different in that they don’t require “public” or upfront financing, and their ultimate cost to consumers and government is difficult to calculate. Thus patents have a political advantage as a result of being “off-budget,” despite the fact that in the long run they may end up having the highest overall costs to the system as a whole.
This is where the article makes its most innovative proposal: a “shadow sales tax” that accounts for the cost paid by users of patented goods and services. The authors argue that taking this cost into account in policy making would allow more accurate comparison between incentives. This could be accomplished if the Congressional Budget Office (CBO) “scored” changes to the patent law to estimate the “shadow” tax on consumers. Until then, private research organizations could fill the CBO’s role.
Of course, as the authors note, it would be challenging to estimate the true cost of patents to consumers and the government in advance. However, they argue imperfect information is preferable to the absence of information. On this point I tend to agree, as empirical data is imperfect in evidence-based decision making. One can only make decisions based on the best data available.