Estimation and Identification of Merger Effects: An Application to Hospital Mergers

Existing empirical estimates of merger effects are compromised by the fact that merging and nonmerging entities differ in unobserved ways that independently affect outcomes of interest. To obtain an unbiased estimate of the effect of consummated mergers, I propose an approach that focuses on the response of rivals to mergers and accounts for the endogeneity of exposure to these mergers. I apply this approach to evaluate the impact of independent hospital mergers in the United States between 1989 and 1996. Using the physical colocation of rivals as an instrument for whether they merge, I find a sizeable, one-time increase in price following a rival's merger, with the greatest increase occurring among hospitals nearest the merging hospitals. These results are more consistent with predictions from structural models of the hospital industry than with prior observational estimates of the effects of hospital mergers.