The dissertation focuses on two distinctive issues in pharmaceutical advertising. One on the matching choices between advertisers and advertising agencies, and the other on the effect of paid-link advertising on consumer search for online pharmacies. The goal of this dissertation is to empirically uncover the underlying economic mechanisms. Moreover, the analysis of matching problem provides new insights on the formation of vertical relationships between clients and professional service agencies and has implications for professional service market consolidations. And the examination of consumer searches for pharmaceuticals online sheds lights on consumers' concerns over quality and affordability of prescription drugs and draws attention on advertising regulation.

In the first two chapters, I focus on two essential features of the market for professional services. One is the necessary mutual agreement in forming relationships, and the other is that a client perceives conflict when hiring the same service agency as his product market competitor. To incorporate these two features, I construct and estimate a two-sided matching model and allow agents' choices to depend on conflict. The results show that conflict does indeed reduce match surplus, and the reduction is greater for a pair of agents who have matched with each other in the previous period. Also, preserving previously formed matches yields much higher surplus than forming new matches. Based on these estimates, I conduct a counterfactual exercise to illustrate the effect of conflict on allocation of matches and another counterfactual exercise to illustrate the effect of a merger between advertising agencies on market equilibrium.

In the third chapter, coauthored with Matthew Chesnes and Ginger Jin, we examine how government's sudden ban of foreign online pharmacies from paid search on Google and other search engines changes consumer searches for the banned websites. Using click-through data from comScore, we find that non-NABP-certified pharmacies receive fewer clicks after the ban, and this effect is heterogenous. In particular, pharmacies not certified by the NABP but certified by other sources, referred to as tier-B sites, experience a reduction in total clicks, and some of their lost paid clicks are replaced by organic clicks. These results have implications for the change in consumer search cost and health concern.