Essays on Industrial Organization and Innovation

Abstract

In my dissertation, I explore policy-relevant topics in the pharmaceutical industry and in the intellectual property system. In Chapter 1, I begin by studying prescription drug demand, and argue that standard cross-sectional demand estimates can lead to misleading substitution patterns and policy implications. The motivating empirical fact behind my research is that patients on a chronic drug the previous year will keep taking the same drug 80-90\% of the time, with most of the remaining patients taking no drug rather than another substitutable drug. I show through quasi-experiments that history-dependence plays a significant role in generating the data, and that most of the continuing patients would pick another drug if their current drug were not available. I conclude by constructing an empirical model of demand around these findings, which can provide more accurate counterfactuals for analyzing firm behavior and policy questions surrounding pricing, formulary design, and the value of drug innovation. In Chapter 2, I argue that existing models of prescription drug pricing miss out on key market features, including the importance of pricing intermediaries, known as pharmacy benefit managers (PBMs), and the history-dependent nature of demand found in Chapter 1. In the chapter, I look to understand the impact of PBMs on pricing, drug company profits, and overall spending. First, I contribute to the literature by constructing average negotiated prices from financial filings, which reflects the actual amount paid to drug companies per drug. These prices are quite different to the list prices commonly used in the literature. Then, I build a model to explain the negotiated prices, which incorporates the PBMs role in negotiating prices using their insurance design power. Incorporating PBMs is crucial for explaining several features of the price data, and my model suggests that PBMs reduce drug company profits by 25\%, but capture up to 40\% of the savings. I conclude by applying the model to evaluate vertical integration of PBMs with drug companies, as well as the pros and cons of potential government-led price negotiation. Vertical integration leads PBMs to favor drugs owned by their parent company in formulary decisions, and the government's effectiveness in negotiating prices would depend on its willingness to exclude drugs, which may be politically challenging in the US. In Chapter 3, I turn to studying the nature of patents, a key component of the innovation system. More specifically, my co-author and I find a strong link between how patent documents are crafted and its subsequent value and usage. To do this, we exploit variation generated by patent examiners at the US Patent Office, and show that the activities of Patent Assertion Entities (PAEs), commonly referred to as patent trolls, and the probability of litigation are quite sensitive to the language in the final patent document. These results show the pitfalls of the influential ``rational ignorance'' view of the patent office, which argues that there is little at stake in the patent examination process.