I am coauthoring (with Cory Capps) a chapter on healthcare antitrust for the forthcoming International Handbook of Antitrust Economics. As we finish our first draft, we were searching for a good way to tie everything together. We both thought of concluding by discussing antitrust and ACOs. Cory and I believe that the underappreciated (and often excruciatingly boring) topic of antitrust is fully interwoven with the story of ACOs. And even if the Supreme Court strikes down the ACA (note to readers of my prior blog – I was just kidding), ACOs may endure. So this is as good a time as any to explain the connections between antitrust and ACOs.
I first recognized this connection twenty years ago, when my colleague Steve Shortell was touting the growth of integrated delivery systems. Steve even offered a universal health insurance proposal (which several states explored) built around competing IDSs. In Steve’s world, an IDS would consist of several hospitals and hundreds of physicians. I argued with Steve that economic theory provided little support for massive vertical integration (and theory is still not all that kind to the idea.) I granted Steve that if integration made theoretical sense, integration would be all well and good for Chicago, where there might be four or five competing IDS. But what about Milwaukee, Cleveland, or any number of other midsize metropolitan areas? They would do well to have two or three IDS. Indeed, even with a legislative mandate to form IDS, consolidation has left these and other midsize markets with just two or three health systems. Smaller metro areas might have only had one IDS.
The empirical evidence on IDS confirmed economic theory. IDS failed to generate consistent efficiencies or quality improvements and no one knows if things will be any better this time. But there is one aspect of integration about which theory and evidence agree. Market power is almost surely bad. Monopolists and oligopolists have less incentive to innovate, reduce costs, and constrain prices. They also tend to take rash steps to maintain their monopolies, signing exclusive contracts with insurers that foreclose rivals, and buying out physician practices to deny referrals to competitors. In the rush to form IDSs, Shortell would have left us with a free market system without competitive dynamics, where rent seeking replaced rent creation. I didn’t see the point.
Fast forward to today and the same issues prevail. Economic theory still suggests that independent providers can be more efficient, especially if they have integrated EMR (which may be one of the few arguments in favor of integration.) But there is no way to shake concerns about market power. Outside of a handful of large metropolitan areas, the concept of competing ACOs may be oxymoronic. The FTC and DOJ have said that they will closely examine any ACO in whose physician members have more than a 30 percent market share in a given specialty. (The horse may have left the barn on this one; many specialty groups have shares exceeding 30 percent.) That could have a chilling effect on ACO formation in smaller markets. But before we blame the FTC and DOJ, let’s remember that there is little research evidence to suggest that physicians have to control 30 percent of the market to realize efficiencies. More problematically, providers are engaging in a range of questionable activities in the name of ACO formation. Hospitals are acquiring physician practices in unprecedented numbers and demanding various exclusive arrangements from insurers. Providers claim they are taking these steps to realize integration efficiencies but there is little evidence to back them up. How simple it is for firms to invoke the mantra of efficiency when the ulterior motive is to exclude rivals.
Do not get me wrong; there is reason to hope that ACOs can correct the mistakes made by IDS. Economic theory is not completely one sided on this matter and integration could lead to efficiencies. But if we allow providers to circumvent antitrust laws in the name of ACOs, then we will give up the benefits of competitive dynamics and even encourage destructive rent seeking behavior, all in the name of a shaky theory with no empirical foundation. That is not an experiment worth taking.