From an interesting review paper in the Journal of Economic Perspectives by Craig Garthwaite:
Although research and development investments clearly respond to market opportunities, the exact benefits created by these incremental investments remain unclear. This is largely because the exact marginal products developed in response to these incentives are elusive to identify. Certainly, the amount of R&D spending or a simple count of the number of resulting new products is an incomplete metric of social welfare benefits.
One effort to understand the social benefit of incremental products has focused on the scientific novelty of these marginal drugs. Dranove, Garthwaite, and Hermosilla (2022) examine the scientific novelty of the targets of clinical trials launched because of the increase in market size resulting from the creation of Medicare Part D (that is, the social insurer for prescription drugs for the elderly). They find the products involved in these additional trials were not novel in terms of their scientific approach, but they did represent some novel combinations of existing scientific approaches—which could provide new benefits to the market. For example, treatment effects or adverse events may vary across patients among even those products with quite similar scientific approaches (Jena et al. 2009).
In contrast, Krieger, Li, and Papanikolaou (2022) exploit the fact that creating Medicare Part D caused different immediate impacts on the balance sheets of firms with products already on the market. They find that firms enjoying a larger cash shock from Part D increased their investments in products that are more scientifically novel compared to firms not enjoying similarly large cash infusions. Importantly, this effect is not driven by the potential returns of the new products but instead by the cash infusions reducing frictions in the financing market for new innovations.
The paper is interesting throughout and provides a great overview of how economic markets impact drug development and social welfare. You can read the full paper here.
