Economists frequently use the two-stage model of hospital competition to assess the competitive effects of hospital mergers. This model predicts that transactions between hospitals whose products are substitutes will generate higher post-merger pricing incentives as the closeness of substitution between the parties increases. However, empirical studies of post-merger price changes do not corroborate the theoretical prediction consistently. A possible explanation for the disconnect between the theory and real-world evidence is complementarities between the merging hospitals, as explained in a recent article by Easterbrook, Gowrisankaran, Aguilar and Wu. Moreover, hospital complementarities may be especially important in cross-market merger analyses.
An intrinsic assumption of the two-stage model is that merging hospitals are substitutes. In this model, a merger can increase a hospital’s bargaining leverage in rate negotiations with health plans, because the substitutability between the hospitals causes each hospital to be less valuable when the other hospital is already in the network. Intuitively, patients have less need for the second hospital if they already have access to the first. From the perspective of a health plan, whose objective is to construct a network of hospitals that patients find desirable, the plan’s willingness to pay for the second hospital is lower when the first hospital already is in the network. If the plan faces a threat of losing both hospitals simultaneously, its willingness to pay for either one of the two hospitals will be higher. These dynamics underlie post-merger upward pricing incentives when hospitals are substitutes.
If the hospitals are complements, however, each hospital is more valuable if the other already is in the network. The classic heuristic that illustrates complementarities is the left-shoe/right-shoe combination. One shoe alone provides limited value to the wearer, and the pair of shoes is significantly more valuable than the sum of the values of each shoe alone. In the context of hospitals, complementarities may arise when each system offers a critical service that the other does not. For example, one hospital might specialize in women’s services while the other hospital offers all other general acute-care services excluding obstetrics. Employers purchasing health plan services may desire networks of hospitals such that all the needs of their insured members can be met somewhere in the network. Since each hospital fulfills a distinct medical need, a network with one limited-service hospital offers low value relative to a network that includes both that hospital and its complement. The addition of the second hospital generates significantly more incremental value when the first hospital already is in-network. Thus, the plan may be willing to pay a high price for either hospital when it knows it can get the other hospital into the network.
Complementarities also can apply in the context of cross-market mergers. In cross-market analyses, hospitals are considered not to compete for patients from the same areas. Because two hospitals are in separate markets, they are not substitutes for each other from the patients’ perspective, even if they offer identical services. Likewise, hospitals in separate markets would not be substitutes from the perspective of an employer purchasing a health plan network. Employers that wish to cover employees in two markets cannot fulfill their network needs by choosing one hospital or the other – they need both hospitals. Employers’ needs for both hospitals to serve their employees make the hospitals complements. Thus, health plans that are marketing to employers with employees in multiple markets need complementary hospitals to construct multi-market networks and to attract those customers.
These types of complementarities can affect hospitals’ negotiating leverage when competing for inclusion in employer networks and, as a result, the rates hospitals receive, since rates are set at this stage of the competitive process. When hospitals are network complements, each one separately is a critical piece of a health plan’s ability to market the network to employers. As such, each hospital can use its threat to leave the network as leverage to negotiate higher rates from the health plan. If complementary hospitals merge and contract as one, they no longer have separate abilities to threaten to leave the network, so their joint incremental value declines, and their negotiated rate declines as well.
Current horizontal merger screening methods do not explicitly account for complementarities, and it is unclear to what extent this limitation drives inaccurate model predictions of price effects. For instance, a recent study by Garmon considers twenty-eight consummated hospital mergers and compares the predictions from current merger screening methods to the actual post-merger price changes for each of these mergers. Garmon’s study finds statistically significant price increases in only nine out of the twenty-eight consummated transactions between hospitals that were substitutes for at least some patients. The other nineteen transactions had no statistically significant price effect or showed statistically significant decreases in price. These empirical findings suggest that more detailed hospital competition models, including those that could explicitly account for complementarities, comprise a productive avenue for further research.
Hospital mergers are common and face increasing scrutiny by regulators, Congress, and the public. Some empirical findings suggest that current screening methods have failed to predict post-merger price changes accurately in a large proportion of real-world cases, and researchers are in the process of improving economic tools to deliver more accurate predictions. In the meantime, a qualitative assessment of the presence of complementarities and their potential impact on post-merger prices could provide meaningful insights in both horizontal and cross-market transactions.