The plan sets the exclusivity period for pharmaceuticals to begin when the drug is approved for sale. That makes patent terms uniform across inventions and removes the disincentive to invest in drugs with long clinical trials.

MAKING PATENT TERMS MORE UNIFORM FOR PHARMACEUTICALS Our third proposal addresses the fact that, even though the patent system is meant to provide uniform patent terms across inventions, it fails to do so in practice. Perhaps the clearest evidence of this failure comes from pharmaceuticals, so we focus our reform proposal on that sector. In the pharmaceutical sector, failure to provide uniform patent terms may cause underinvestment in candidate drug compounds that require long clinical trials. In the United States and most other countries, patent terms generally last 20 years after an application is filed. But the effective patent term (i.e., the patent term that remains once a patented product is on the market) can be much shorter than 20 years when there is a long lag between patent application filing and first sale of the patented product to consumers. Drug patents are filed prior to the start of clinical trials, but effective patent terms do not start until after the trials have been completed and the drug has been approved by the U.S. Food and Drug Administration (FDA) for sale to U.S. consumers (Lietzan and Lybecker 2020; Roin 2014). Patents thus paradoxically provide the weakest incentive for inventions that take the longest time to develop, whereas simple theoretical models would suggest that society might prefer the opposite. Budish, Roin, and Williams (2015) discuss two examples that—while not definitive evidence of this distortion— illustrate how this problem could manifest itself in practice. Consider two clinical studies investigating treatments for prostate cancer, both of which were published in the New England Journal of Medicine in 2011. The first study (de Bono et al. 2011) analyzed a treatment for metastatic prostate cancer patients whose cancer had spread beyond the initial tumor and who had relatively poor survival prospects (in their data, a five-year survival rate of around 20 percent). The study needed to measure patient survival for just over a year in order to document a statistically significant improvement in survival in the (randomized) treatment group, and the total trial length was around three years. The second study (Jones et al. 2011) analyzed a treatment for localized prostate cancer patients whose cancer had not yet spread beyond the prostate and who had better survival prospects (in their data, a five-year survival rate of around 80 percent). Because these patients were in better health, this study needed to measure patient survival for a full nine years, and the total trial length was 18 years. In practice, drugs need more than one clinical trial to receive FDA approval, but taking the simple case where drugs get a 20-year patent term and patents are filed prior to the start of a single clinical trial, the first drug would receive 17 (20 minus 3) years of effective patent life, whereas the second drug would receive only 2 (20 minus 18) years. Perhaps because of the much shorter effective patent term provided in the second case, the first (metastatic) clinical trial was funded by a small biotech company, whereas the second (localized) clinical trial was publicly funded. This potential concern has long been recognized by policymakers, and the 1984 Hatch–Waxman Act included a provision that granted some qualifying firms a partial extension of their patent life to compensate for the time drugs spent in clinical trials. Specifically, Hatch–Waxman awards qualifying firms one half-year of additional patent life for each year spent in clinical trials, up to a maximum of five additional years, but not to exceed a total of 14 years after FDA approval. In practice, however, the structure of the current Hatch–Waxman provision appears flawed in the sense of not fully correcting for this distortion. Lietzan and Lybecker (2020) document that, for all 642 drugs that had their patent term restored between September 28, 1984 and April 1, 2017, the average effective patent life was less than 12 years, with drugs that required longer clinical trials receiving shorter effective terms. Congress could address this flaw by reforming Hatch–Waxman, for example by starting the exclusivity period (of some length; as we discuss below, potentially shorter than the current 20-year patent term) for pharmaceuticals at the time the drug is approved by the FDA for sale to patients. Simultaneously, patent and antitrust law should be vigorously enforced to limit opportunistic extensions of effective exclusivity periods that do not contribute to social value, such as so-called evergreening practices (Feldman 2018).