Are Hospital Acquisitions of Physician Practices Anticompetitive?

This paper empirically analyzes the effects of mergers between complementary firms on competition and pricing. As these non-horizontal mergers have become more common, there is increasing interest in evaluating both potential efficiencies such as eliminating double marginalization and potential anticompetitive effects such as foreclosure and recapture. The mergers we study – hospital acquisitions of physician practices – have reshaped the $1 trillion US physician industry, nearly doubling the share of physicians working for hospitals between 2008 and 2016. We combine novel data and machine learning algorithms to identify a large number of integration events, spanning a wide range of markets with different competitive circumstances. We merge the integration events with claims data from a large national insurer to study their effects on prices. Focusing on childbirths, the most ubiquitous admission among the privately insured, we find that, on average, these mergers led to price increases for hospitals and physicians of 3.3% and 15.1%, respectively, with no discernible effects on quality measures. Using demand estimation to characterize substitution patterns for both physicians and hospitals, we construct tests that demonstrate price increases are larger among transactions with greater scope for foreclosure and recapture. Our estimates suggest that the costs of these mergers of hospitals and physicians have been substantial, and our mechanism tests offer guidance in predicting where the anticompetitive effects of non-horizontal mergers are likely to be strongest.