The Economics of Drug Shortages

Abstract

In the first chapter of this dissertation, I study the impact of the entry of nonprofit pharmaceutical companies on drug shortages in generic pharmaceutical markets. I employ a staggered difference-in-differences (DiD) regression design for estimation, exploiting the unannounced entry of Civica into markets for multiple injectable drug products. Examining the effect of Civica on drug shortages is important because it is the first large nonprofit manufacturer in the pharmaceutical market in the U.S., and its impact on supply disruption has not yet been studied. I find a 4% reduction in drug shortage incidence after Civica Rx entry. This result is robust to including all reasons for shortages except for manufacturer discontinuations. In the second chapter, I investigate whether the incidence of national drug shortages across all drug classes varies by healthcare provider, patient, and drug characteristics using linear regression models (OLS). Prior literature on patient and provider characteristics and drug shortages focused only on the use of oncology medications. My findings highlight which racial, ethnic, and socioeconomic groups of patients endure the greatest burden of drug shortages. Also, I identify the characteristics of healthcare providers able to secure better access to drugs that are in short supply. I find that the severity of drug shortages is lower among physicians who see patients with higher average Hierarchical Condition Category (HCC) risk scores. The evidence suggests that physicians who see more Black patients are more affected by shortages, while physicians who treat Hispanic patients face fewer shortages. Additionally, female doctors have better access to drugs in short supply compared to male doctors. While data limitations prevent controlling for potential hospital and health system level confounders, this chapter presents some of the first findings on these important issues.