UC Berkeley

This dissertation explores the relationship between vertical integration and market structure formation. It does so by combining the empirical industrial organization literature on vertical integration with that on market entry. The first empirical essay, Chapter 2, explores the motives for vertical integration in the US generic pharmaceutical industry. The industry is made up of numerous drug markets that open up to competition among generic manufacturers at different points in time. Each market consists of an upstream segment that manufactures active pharmaceutical ingredients, and a downstream segment that processes the active pharmaceutical ingredients into finished formulations and supplies them to final consumers. The econometric analysis shows that vertical integration in the generic drug industry is characterized by bandwagon behavior. While bandwagon effects have been widely discussed in the vertical integration literature, this study is one of the first to present empirical evidence on its existence. The analysis also indicates that vertical integration is partly driven by the need for a particular form of relationship-specific non-contractible investment -- the early development of active pharmaceutical ingredients by upstream units. The relationship-specificity of such investments is greater in markets where generic firms try to enter by challenging the patents held by originator pharmaceutical companies. I find that in such markets, individual firms have a higher propensity to vertically integrate.

The second empirical essay, presented as Chapter 3, introduces an econometric model of a vertical entry game. The model is used to estimate rival effects -- the effect of rival entry on the post-entry profits of individual firms. These estimates allow us to make inferences about the competitive effects of vertical integration. Application of the model to the generic pharmaceutical industry yields the following result: vertical integration has significant efficiency effects that benefit unintegrated downstream firms. This implies that vertical integration is likely to be procompetitive from a static point of view. The parameter estimates are used to simulate the impact of a hypothetical policy that bans vertically integrated entry. The results indicate that such a ban tends to reduce the equilibrium number of downstream entrants. This suggests that the effect of vertical integration on market structure formation is also procompetitive.