The effect of market size on pharmaceutical innovation: matching estimates to policy questions

The elasticity of pharmaceutical innovation with respect to market size is a key input for evaluating pricing reforms, yet published estimates vary widely.

We examine how much of this variation reflects measurement choices-such as which drugs are counted in innovation outcomes and which countries are included in market size measures-rather than differences in research design.

Using a comprehensive dataset of global pharmaceutical sales, we estimate period-class Poisson regressions following the specification of Dubois et al. (2015), systematically varying the sample of drugs and the measure of market size.

Under our preferred specification, estimated elasticities range from 0.21 to 0.30, with estimates as high as 0.47 in other specifications. Restricting to drugs launched in the United States or other large, high-income markets yields significantly higher elasticities than the global sample, but United States and global revenue measures produce similar estimates within each sample, contrary to a common conjecture in the literature.

Matching the right elasticity to the right policy question is essential for accurate projections; analysts should select estimates from studies whose measurement choices align with the specific policy and outcomes under evaluation.