David B. Wilkins, faculty director of the Center on the Legal Profession, recently sat down with Leemore Dafny, Bruce V. Rauner professor of business administration at Harvard Business School, for a discussion around consolidation in the health care market and what lessons those in the market for legal services might draw from their experience.
David B. Wilkins: Professor Dafny, thank you for making the time to talk to us here at The Practice. One of the things that’s so striking about the legal profession is that it is one of the most decentralized, fragmented, multibillion dollar industries in the world. There’s no producer that even has 1 percent of the market. If you look around, there were a lot of industries that, at one time, might have looked like law in this regard, but they’ve all undergone tremendous consolidation. One industry that law is always often compared to is medicine. And, in doing some research, we found that medicine and health care have been undergoing a consolidation and integration boom—which brought us to you! We’re interested in your understanding of why this is happening in medicine and health care, and what might be some of the lessons we could learn for the legal profession.
There are a handful of studies that find cost reductions through consolidation under certain limited circumstances, meaning that the story is not, if it ever was, so much about cost reductions.Leemore Dafny
So, let me start from the beginning, because, as your research has shown, there’s been tremendous consolidation in the health care industry, particularly recently. Why do you think this is happening now? What are the factors, both external and internal, that are facilitating this change?
Leemore Dafny: Thank you for giving me a chance to think about the parallels between law and health care around these issues. You asked why there’s consolidation now in health care. First, it’s not occurring just now. It’s been going on for quite some time, motivated at the beginning by cost control efforts, specifically the managed care organizations in the 1980s and Medicare’s transition away from cost-plus reimbursement—bill us what it costs, plus a small margin. Medicare’s pushback caused the delivery system to look inward and say, “How are we going to deal with costs?” And that led to the idea that consolidation (that is, bigger enterprises) would be more efficient. At the same time, pressure from the managed-care organizations that were starting to also shift away from cost-plus reimbursement and negotiate lower prices meant that the bigger you are, the stronger your bargaining hand. So, on the one hand, you have the public insurers. They’re a smaller share of your revenue, but nevertheless you want to be more efficient, so you don’t lose money serving these patients. On the other hand, the private players represent a bigger share of your revenue, and you want to be bigger so you can bargain more effectively with them. Consolidation seemed to solve both.
Now, 20 years on, I have not seen almost any research on minimum efficient scale for hospitals, that is, the smallest size at which they hit the lowest operating costs. There are a handful of studies that find cost reductions through consolidation under certain limited circumstances, meaning that the story is not, if it ever was, so much about cost reductions. That doesn’t mean mergers can’t reduce costs, but the most consistent message we see from research is that consolidation tends to lead to higher private prices. It doesn’t do anything for your public prices, because those are administered in the context of Medicare and Medicaid, but the private sector prices have gone up from around 10 percent more than what the public sector was paying in the late 1990s to more than double now. It’s huge. That gap has just grown and grown. And that wouldn’t be possible but for barriers to entry and market power on the part of the consolidated entities.
More recently, there has been a desire by payers to shift risk to health care groups and say, “Instead of paying a fixed amount per service, which still seems to be costing us a lot, let’s try to pay you closer to a fixed amount per human per month. We’ll project what it should be, and if you come in under, you’ll get savings, and if you come in over, you’ll have to bear some of the costs.” That has encouraged organizations to communicate and coordinate, and potentially to merge. If providers merge, they might manage care better and improve quality, and either way they will benefit from increased bargaining leverage.
Wilkins: Hearing that, I think there are more parallels, even than you might think, to what’s going on in law! One of the things we’ve been charting is the shift from cost-plus pricing to value-based pricing, which is being driven, in this case, not by insurance companies but by sophisticated clients, particularly in the corporate sector. The follow up to this consolidation is, of course, who are the winners and losers. Indeed, oftentimes it’s advertised as something that’ll be good for everybody: It’ll lower prices, it’ll increase efficiency, and it’ll increase the quality of services. But you’re already suggesting that perhaps prices were not lowered as much as they might be. I wonder with respect to some of the internal constituencies, say between doctors and hospital groups and big payers, how do you see the winners and losers shaking out?
The patients are losing, because there is no evidence that consolidation benefits them in any way.
Dafny: The losers would be typically the unconsolidated practices that remain smaller independents, whether they’re dialysis facilities, physical therapists, or independent physicians who bear a lot of risk and are reaping much fewer rewards. They’re making the choice to be independent—but it’s a less lucrative choice than it used to be. It’s very challenging too. It’s a lot of work dealing with reimbursement hassles that have multiplied. Larger entities can handle frequent changes more easily because they have scale and expertise.
Also, and in my view, more importantly, the patients are losing, because there is no evidence that consolidation benefits them in any way. That doesn’t mean that certain consolidations have not or do not. But, on average, the research is not very encouraging, whether it’s horizontal or vertical—and it is more damning—thus far, given the larger body of research we have on it—for horizontal integration.
Now, I will say that the horizontal integrations that have been studied are often those that could cause damage, as opposed to something like the average four-clinician practice joining another four-clinician practice. That type of consolidation—depending on the market size and specialty—typically doesn’t get studied.
Acquisition of physician practices by hospitals, which is a form of vertical integration, has been well-studied, though. The studies find these acquisitions tend to drive up prices. If you’ve ever seen a little sign in the doctor’s office that says, “We are a hospital-based facility,” or, “You will see a separate charge because we joined such and such hospital”—overnight your charges go up, because hospitals have these fees they can tack on, and the payers pay them whether they’re private or public payers.
You might say, “Well, wait a second. They just bought a practice. They changed nothing except for the sign and some back-office work.” Or, “They bought a surgical center, popped a sign on it, and now it’s part of the hospital. And why do the payers just say, ‘Oh, okay, I’ll pay you more?’” It’s because there’s market power that enables these delivery systems to do that, whether they are buying a facility or another hospital. They just roll the new facilities onto their contracts and say, “Now we charge more for their services.”
Wilkins: This is partly what I read in the popular press and gleaned from some of your work, which is what’s generating the regulatory pushback. I wonder about the role of regulation. Again, just to give you a little context, like medicine, law used to be an incredibly regulated business, which really gave a kind of de facto monopoly to lawyers to do legal services. The definition of legal services was incredibly broad and, therefore, there was very little competition, and there was a lot of other anti-competitive regulation. Over the years, much of that regulation has been either challenged as anti-competitive in court cases or rolled back, mostly outside of the United States, by governments who are worried that it’s increasing the cost of services. So, on the one hand, there’s been a deregulatory move, which for example, has loosened, particularly in the UK, restrictions against outside ownership of law firms, or legal organizations by “nonlawyer” investment, and that has opened the way for multi-disciplinary partnerships. And, on the other hand, this has brought in a number of new players, including technology companies and the Big Four, into the marketplace.
In medicine, it sounds like that kind of deregulation also happened, but now you’re seeing a press towards reregulation around consolidation. I wonder how you think about how these regulatory pressures are shaping the consolidation that is happening.
The real action has not been on the labor market side—it has been trying to challenge anti-competitive mergers.
Dafny: I think of the regulation vis-a-vis the organizations and the big systems as a little bit separate from regulations vis-a-vis the labor markets. Perhaps because they don’t intersect as closely as they do in other sectors.
On the labor side, there is a range of practices that courts or regulators have identified as anticompetitive, particularly in the dental sector, around what dental hygienists or “nondentists” can do, for example, with teeth whitening. There are also state-sanctioned regulations that protect physicians from competition from other health professionals, such as advanced practice registered nurses. The FTC has worked hard to advocate for elimination of such “scope of practice” restrictions unless they are warranted for reasons of patient safety. Otherwise, such regulations limit competition among physicians.
But, in my view, the real action has not been on the labor market side—it has been trying to challenge anti-competitive mergers. Less than 3 percent of hospital mergers even go to second request at the FTC, i.e., are thoroughly investigated due to a potential concern. And vertical combinations are even less likely to be investigated. But the volume of deals is so high that the agency is incredibly busy with these transactions—a reason they could use more funding.
Wilkins: I want to pick up on that, since I know you’ve done some fantastic work thinking about the vertical integration of supply chains, particularly around skilled nursing care facilities into larger hospital care networks. We’re very interested in that phase, because what we’re seeing is, for example, law firms have their own captive legal process outsourcing organizations, or they are starting technology companies and platforms. The Big Four have also moved quite aggressively into the law business. And the question is, should they think about vertically integrating? Should they think about alliances and networks? Should they think about partnerships?
Dafny: I’ll get to my research in a moment, but I think that it would be useful to give some context. One key motive for vertical integration, from the economist’s perspective, is that it can enable two parties to get to the optimal level of relationship-specific investments. If I’m a pharmaceutical company, and I only have a blockbuster every couple of years, I could just outsource and hire a marketing firm when I am ready with a drug instead of having that function in-house. Why vertically integrate and do that internally? Well, if I want them to build relationships with doctors, to be promoting my medications and learning about why they’re so much better than everything else out there, they can’t worry that next year they’re going to be selling somebody else’s drug. If they have a relationship-specific investment—that is, they have long-term employment with the pharmaco—they will want to invest in their knowledge of the company and the product and how they convey that value. So, there are reasons it might make sense to be vertically integrated.
Another reason you might vertically integrate is if you need somebody to know all of your clients, and they need time to build relationships. You also don’t want them to just take what they know somewhere else. And vertical integration can also avoid what’s called the “hold-up problem.” Take coal plants and coal mines. You create a plant to burn the coal with the sulfur content from a specific mine. And then the mine says, “Oh, you know what? We’re going to raise our prices.” That changes the economics of the whole thing. By vertically integrating, the power plant doesn’t have to worry about the relationship-specific investment it made in equipment and can avoid the hold-up problem that would cut profits for owners.
I would have expected, in health care, to see greater effort by payers to reward the effective providers and exclude those who are ineffective, but there’s fairly limited evidence of either of those things really going on.
How does this play out in health care? To give a concrete example, in practice, 20 percent of Medicare patients are discharged from the hospital to a skilled nursing facility (SNF)—it is a very common destination. There is an argument to be made that if the hospital owns the facility—if the facility were vertically integrated—they may be able to do a better job in terms of both quality of patient care and costs. On the other hand, the fear is that by vertically integrating, the hospital will foreclose rival nursing facilities. The research project I did looks at how a change in Medicare reimbursement for SNFs affected vertically integrated SNFs as compared to independent SNFs. The bottom line is that vertically integrated facilities self-refer more profitable patients. Equally important, we did not find evidence that a vertically integrated skilled nursing facility generated better outcomes for the patients, or that it led to lower total Medicare spending. In a separate analysis, we found that vertical integration is associated with greater exit by independents. So, at the end of the day, if you don’t have better outcomes for patients or payers and you do have a reduction in competition, that seems like a bad deal for patients.
Wilkins: And that really brings the last piece that I wanted to ask you about because, at the end of the day, this is all supposed to be about producing quality outcomes—whether for patients in health care or for clients in the law. One of the challenges in law is that we had no way of measuring quality outcomes. We only had an input-based model, often based on who did it—if you went to Harvard Law School, if you clerked on the Supreme Court, if you worked for a fancy law firm. Or, we had the crudest measures of outcomes. Did you win or did you lose? But we know that bad lawyers win and good lawyers lose.
I just wonder how you think about, one, the relationship between consolidation and integration and quality and, two, how you go about trying to evaluate or measure it.
Dafny: What we are producing are largely credence goods or experience goods. We have to trust that, when the doctor does some surgery, that that actually happened in the way that it was supposed to. And we may not know if it was properly done for a long time—or ever. Even today, measurement of actual outcomes, while adjusting for the risks of the patient, is pretty limited, except for some very high-volume procedures.
I would have expected, in health care, to see greater effort by payers to reward and exclude—reward the effective providers and exclude those who are ineffective. But there’s, in my view, fairly limited evidence of either of those things really going on.
What insurers have preferred to do is—and this is partly the power of physicians in preventing payers from interfering with the medical decision—is prior authorization and utilization review. Payers screen for appropriateness. That may improve quality, but I haven’t seen any studies supporting that. At the end of the day, you can’t get what you aren’t measuring. I would say that there a lot of efforts. But while these efforts are not in their infancy, they have a lot of room to grow.
Leemore Dafny is the Bruce V. Rauner Professor of Business Administration at Harvard Business School, where she teaches courses in health care strategy and codirects the PhD program in business economics.
David B. Wilkins is the Lester Kissel Professor of Law at Harvard Law School, vice dean for Global Initiatives on the Legal Profession, and faculty director of the Center on the Legal Profession.