Matt Schmitt’s research shows hospital prices rise after acquisitions, even where local competition is unchanged
With health care costs rising and merger activity between hospital systems strong, antitrust regulators frequently scrutinize deals that would reduce the number of competing hospitals within a geographic area. The Federal Trade Commission often requires these buyers in turn to sell specific hospitals where the combined company would dominate the markets, as it did in Community Health Systems’ $7.6 billion acquisition of Health Management Associates in 2014. The FTC noted that a complete merger would give Community Health Systems a “near monopoly” position for its hospitals in certain parts of Alabama and South Carolina, forcing Community to sell facilities in those locations to gain approval for the deal.
In theory, these forced divestitures keep price competition healthy even as the overall number of players shrinks. But new research by UCLA Anderson’s Matt Schmitt suggests that hospital system mergers may lead to higher prices regardless, even in places where local competition is unchanged. The study is forthcoming in the American Economic Journal: Economic Policy.
Schmitt’s findings suggest that companies are less competitive when they face each other in multiple markets. Economists have long suspected that multiple market competition dampens normal price competition, but the theory has proven difficult to test.
Cutting U.S. health care costs — a goal considered critical for sustaining good care and a healthy economy — will be particularly challenging if, as Schmitt’s work implies, hospital costs rise as hospital systems grow. Owing to rapid industry consolidation in recent years, hospital chains that compete simultaneously in numerous markets are growing more common. Continued hospital mergers may produce higher health care costs, even when there is no geographic overlap in the two companies’ properties for the FTC to protest, the findings suggest.
Medical expenditures in the U.S. totaled $3.2 trillion in 2015, with costs for hospital care making up about a third of the total, according to the Centers for Medicare and Medicaid Services. In 2016, the U.S. health care system continued to rank as the most inefficient in the developed world, costing 17.1 percent of GDP, according to World Bank rankings of expenditures by country. By comparison, Europe’s high spenders, Switzerland and France, paid 11.7 percent and 11.5 percent of their own GDP, respectively.
Meanwhile, some 1,260 U.S. hospitals changed hands in 561 mergers in 2010 through 2015, with many of the deals pushing more facilities into multi-hospital systems, according to the American Hospital Association. Merger activity in 2016 was 55 percent higher than in 2010, and buying is expected to continue at a strong pace for the foreseeable future, according to merger attorneys Kaufman, Hall and Associates.
Taking an unusual approach to data from about 3,000 hospitals, Schmitt isolates the effects of multiple-market competition in ways that traditional studies do not. Like other researchers in the field, he studied mergers in which the acquirer operated in competition with a particular company in one location, and then, via acquisition, became its competitor in another market. But rather than compare prices in both of those markets, Schmitt focuses on changes in the market where the two companies compete that is most removed — sometimes hundreds of miles away — from the acquisition activity.
Consider a hypothetical example of how the mergers were structured, and Schmitt’s approach to studying them. ABC Health Care owns hospitals in Denver and Atlanta. XYZ owns them only in Denver. When XYZ acquired a third hospital chain whose assets include properties in Atlanta, Schmitt looked at price trends in Denver. He then compared the Denver trends to prices in a control group consisting of uninvolved hospitals.
Because the findings do not rely on data from recently acquired properties, price changes are less likely to be affected by management changes and other merger activity unrelated to competition, the study states. They are more likely to be caused by the additional contact ABC has with XYZ overall.
Schmitt does not directly investigate how increased contact between corporations might lead to less price competition, but he relates some ideas from other research. Rivals may be less willing to undercut prices in one location for fear that the competitor might retaliate in another region. Or competitors that meet in multiple parts of their businesses may find it easier to tacitly collude on pricing.