Brown Bag Seminar: Estimating R&R Costs in the Pharmaceutical Innovation Industry

This paper studies how pharmaceutical firms make dynamic investment decisions along the drug development pipeline. We develop a structural model in which firms sequentially decide whether to continue clinical development after each phase, facing uncertain success probabilities, stochastic development durations, and forward-looking expectations about future revenues. The empirical analysis combines two uniquely complementary data sources. Using global project-level data from Pharmaprojects, we track the universe of drug candidates as they move through clinical development over time, observing phase transitions and durations for each molecule. We merge these data with IQVIA sales information to construct lifetime revenue measures for marketed drugs. This novel linkage allows us to estimate a dynamic model that jointly captures development risk, timing, and commercial outcomes. From the model, we recover therapeutic class and phase-specific development costs. This framework enables a wide range of policy counterfactuals. In particular, the model can quantify how policies affecting development timelines, such as accelerated approval pathways, changes in expected market revenues, or phase-targeted R&D subsidies, shift firms’ investment decisions and ultimately shape innovation output. This is a joint work with Pierre Dubois from Touloise School of Economics.