Simulating Hospital Merger Simulations*

Abstract

We perform a Monte Carlo experiment to assess the performance of three hospital merger simulation methods. Our analysis proceeds as follows: (i) specify a theoretical model of hospital markets and use it to generate “true” price effects for many simulated mergers; (ii) for each simulated merger, generate data of the kind commonly available in real-world merger analysis and apply the simulation methods to those data; and (iii) compare the predictions of the simulation methods to the true price effects. All three simulation methods perform reasonably well. We also develop a method for predicting price effects that extends Garmon [2017].