Hospital industry consolidation – Still more to come?

By Leemore Dafny
The New England Journal of Medicine, December 11, 2013

The Affordable Care Act (ACA) has unleashed a merger frenzy, with hospitals scrambling to shore up their market positions, improve operational efficiency, and create organizations capable of managing population health. The figures are impressive: 105 deals were reported in 2012 alone, up from 50 to 60 annually in the pre-ACA, pre-recession years of 2005–2007. This activity could have lasting repercussions for consumers; the last hospital-merger wave (in the 1990s) led to substantial price increases with little or no countervailing benefits. …[A] compelling argument can be made for putting the brakes on consolidation. Indeed, unless new public and private initiatives are developed to discourage consolidation and to support enforcement of antitrust law, most of these deals will proceed unchallenged. [endnotes omitted]


[I]t would behoove health care analysts and policymakers who are concerned about consolidation to give enforcers more tools for doing their jobs and to develop other avenues for slowing the march toward conglomeration.


A full discussion of possible initiatives is beyond the scope of this article, but three ideas are worth mentioning. First, angel investors and venture capitalists could help create innovative health care – provider organizations that deliver clinically integrated, evidence-based care at the lowest possible cost without reducing competition. New funders would consider different organizational ideas and bring strategic and operational skills to their ventures, and they might be better positioned than local health care systems or physician groups to accept the associated risks. Second, Medicare could experiment with reimbursement schemes that provide incentives to newly forming accountable care organizations to pursue organizational structures that do not involve joint ownership of all assets. Joint ventures and contractual relationships would be easier to unwind than mergers, if that proved necessary.

Last, but certainly not least, we could urge private and public insurers to make detailed claims data readily available to public agencies and private researchers….. These data would enable researchers and enforcers to assess how the latest types of consolidations affect both costs and quality.…


By Kip Sullivan, JD

The first seven paragraphs of this essay present an important and rarely made argument, namely, consolidation within the hospital industry is accelerating once again, it is a serious problem, and antitrust law is too weak to stop it. The author correctly implies that the sections in the ACA authorizing ACOs are a significant cause of the latest epidemic of merger fever. But in the final two paragraphs the author prescribes utterly futile solutions. She recommends:

•  that outside investors sponsor new ACOs to compete with ACOs being created by hospitals;

•  that CMS encourage ACOs to form by contract rather than by merger so that in the remote event the Department of Justice were to win an antitrust suit the ACO would be easier to break apart; and

•  that scholars and antitrust authorities do more research on ACOs.

Merger madness throughout the health care system will not stop until the incentive to merge is removed. That incentive is the conventional wisdom that the solution to our health care crisis is to shift insurance risk onto hospitals and doctors. The ACO fad is merely the latest expression of that groupthink.  Encouraging venture capitalists and CMS to sponsor more ACOs is not the solution.

Within the health care sector of our economy, the most powerful incentive to build empires is the widespread belief, nourished enthusiastically by the insurance industry, much of academia, and both political parties, that our health care crisis is caused primarily by excessive volume of medical services induced by the fee-for-service payment method. If you buy this diagnosis, then you are compelled to endorse a “solution” that neutralizes the FFS incentive (salary) or turns the fee-for service incentive upside down. For reasons I won’t get into here, putting all doctors on salary has never been seriously considered. Instead, those who bought the overuse diagnosis adopted a solution that encourages providers to order fewer services. Since the early 1970s when this conventional wisdom took root, the antidote to FFS has been to force clinics and hospitals to share insurance risk with insurers and to subject providers to utilization review by insurers in the event that risk-sharing fails to cut volume enough.

The earliest method of shifting insurance risk was called “capitation.” Capitation was later joined by “bonus sharing/withholding,” “pay-for-performance” (where “performance” was judged by cost to the insurance company), and lately “shared savings.” It all amounted to the same thing – making providers share the risk with insurance companies that the premium revenue taken in by insurance companies would not be sufficient to cover the cost of the medical services provided to the insurance companies’ policyholders. (To make the function of these schemes clear, it would have been helpful if the powers-that-be had called “capitation” and other forms of risk shifting “premium-splitting,” but that would, of course, have called attention to capitation’s close kinship with “fee-splitting,” and we wouldn’t want that.)

The solution has been, in short, to make clinics and hospitals bear some or all of the insurance risk traditionally borne by insurance companies. The ACO fad takes this logic to its natural endpoint: providers are being induced to accept some insurance risk now and will be induced or forced to accept all insurance risk later on, in other words, become insurance companies. (If this sounds incredible, you haven’t been reading what leading ACO advocates such as Elliot Fisher and the Medicare Payment Advisory Commission are talking about. These advocates have been silent on what it is insurance companies are supposed to do in a world where clinics and hospitals bear all insurance risk.)

This logic – that overuse induced by fee-for-service is the problem and shifting insurance risk is the solution — in turn required that clinics and hospitals be herded into large networks organized by insurance companies, and that patients be required or given incentives to see only providers within these networks. Herding providers into large systems was essential because small hospitals and clinics cannot bear insurance risk. Forcing patients to stay within networks was essential because you can’t ask providers to share insurance risk for patients who aren’t contractually required to come to them in the event of illness.

Let us stop for a moment and review this cascade of undocumented assumptions:

•  Overuse, not underuse and not excessive price, is the cause of our health care crisis;

•  The FFS method of payment is the main cause of overuse;

•  Forcing providers to bear insurance risk and to allow insurance companies to second guess provider-patient decisions is the appropriate antidote to FFS;

•  Providers must become big to shoulder insurance risk;

•  Patients must be forced to accept limited choice of provider if providers are going to be expected to bear insurance risk.

Once we array the elements of the establishment’s syllogism this way, we can see how the overuse diagnosis (and the refusal to acknowledge the primary role of high prices and administrative waste which drives prices up) created a strong incentive among providers to merge. But there was one other consequence of this syllogism which turned out to be perhaps the most powerful stimulus – the effect this syllogism had on incentives to get big within the insurance industry.

Allowing insurance companies to create provider networks and limit patient choice gave the early-adopting insurers much more market power over providers. The first insurers to create their own provider networks and limit patient choice (they were called HMOs) used this market power to extract enormous discounts from providers, and to exert more influence over decisions by doctors and patients. (By the early 1990s, HMOs were forcing hospitals to give them rates at 30-40 percent below those of competing insurers that had not yet adopted HMO tactics.)

By the late 1980s this lopsided growth in market power in favor of insurance companies had induced a counter-reaction among providers, especially hospitals. They began to merge in order to create what John Kenneth Galbraith called “countervailing power” to offset the new power acquired by the burgeoning insurers. The insurance industry responded with more mergers of their own, which induced even more mergers among hospitals and clinics. And around the vicious cycle goes. The insurance industry’s effort to exclude many hospitals from networks available on ACA exchanges is just the latest skirmish in this never-ending and pointless battle.…

To sum up, by the mid-1990s the conventional wisdom that FFS-induced-overuse is the problem, and that shifting insurance risk to providers is the solution, had put enormous pressure on both insurers and providers to get big. This pressure has only been aggravated by the ACO provisions in the ACA, and to a lesser extent by the cuts in Medicare reimbursements to hospitals authorized by the ACA in order to finance enormous subsidies for the insurance industry.

Repealing those ACA provisions would be helpful, but far from sufficient. The urge to merge will not subside until policymakers stop demonizing FFS and lionizing risk-shifting to doctors and hospitals.

Abandoning the quest to stamp out FFS does not mean we must give up on eliminating the overuse that does exist. We could enact a single-payer system and, failing that, take other steps within the current system. Single-payer systems affect overuse not by micromanaging doctors and hospitals, but by

•  allocating capital equipment (such as MRIs) equitably and intelligently,

•  funding more research on which treatments are most effective and educating doctors and patients on the results,

•  controlling the relative price of services that appear to be overused (for example, lowering the fee for C-sections relative to the fee for normal deliveries), and

•  making some types of fraud more easily detectable.

Some of these solutions (such as more research and education) are available even under a multiple-payer system.

What we should not do is continue to experiment with shifting risk to doctors and hospitals.