The Incidental Economist

Historically, I’ve been  a very quantitative and multivariate thinker. That explains my undergraduate and graduate interest in physics and engineering, my fascination with the game of Go, and my professional focus on econometric techniques for causal inference in health care. But lately I’ve been paying a lot more attention to qualitative and descriptive evidence in the health services research literature.

It’s not that I suddenly think qualitative and descriptive analyses answer questions any more definitively than do the complex multivariate approaches I normally consider. It’s that qualitative and descriptive analyses are incredibly helpful in revealing potential relationships, suggesting hypotheses, and relatively quickly and easily revealing the possible broad nature of and context for an issue.

Moreover, less econometrically or statistically complex analyses are more easily communicated to thinkers outside the field, like policymakers, journalists, and beekeepers. That’s important, and I have no problem using results of simpler analytical techniques for policy advocacy so long as the qualitative conclusions they suggest are consistent with more rigorous, but complex and less broadly accessible, approaches.

But be on guard! How can non-practitioners tell when qualitative or descriptive analyses are “telling the truth” and when they’re just wrong? Oh boy, is that a thorny issue. For now I”ll just say, roughly speaking, they can’t. They need the guidance of experts. (Perhaps I’ll come back to this issue another time.)

With that as backdrop, here are two very good passages from just one of many qualitative/descriptive papers published by researchers with the Center for Studying Health System Change. The paper, by Devers et al., appeared in a 2003 issue of Health Services Research. First, on market power:

Market power is defined as the degree of control or influence an organization has over another organization (Scott 1987; Emerson 1962). Control or influence is shaped by the willingness and ability of one organization to sanction (i.e., punish or reward) another organization that it interacts with to attain key goals, such as survival, growth, or increased margins. The origin of market power is the dependency one organization has on the resources controlled by another….

This sociological definition highlights why and how an organization exercises market power, as well as the outcome (i.e., increased control or influence over another organization in a key area). As such, this definition of market power is broader than those used in economics, which focus primarily on the ability of an organization to influence price. For example, Carlton and Perloff (1994) define market power as the ability of a firm to charge a price above that which would prevail under perfect competition, usually taken to be marginal cost.

Next up, some commentary that relates market power (or “negotiating leverage,” which the authors take to be a synonymous moniker) to arrangements that may be encouraged by an accountable care organization (ACO) payment model.

Another key change [between 1996 and 2001] was the level of hospitals’ vertical integration with physicians (e.g., physician practice acquisition, formation of intermediary organizations such as physician–hospital organizations). Greater hospital-physician alignment strengthened hospitals’ negotiating leverage and weakened plans’ options. Many hospitals had implemented a range of physician-integration strategies, becoming a critical gateway for plans to physicians in the market. In many of the contract disputes noted above, plans were negotiating with hospital-physician organizations for physician professional services as well.

That pretty much sums up my concerns about ACOs, and those of others in my field. I think the qualitative ideas suggested–that vertical integration encouraged by ACOs may increase provider market power and consumer prices–are consistent with rigorous quantitative studies of hospital and provider market power. (I’ve cited about a bazillion of such studies in prior posts; or, you can just trust me.) So, I have no problem making some policy suggestions based on these concerns. Here goes:

First, let me be clear, I’m not saying ACOs are bad. I’m saying I have concerns about the implications for private-sector costs of such organizations, which will be encouraged under new Medicare and Medicaid, ACO-like payment schemes. What this means is that policymakers and regulators need to tread carefully in this area, think things through, and come up with public-side cost control approaches that don’t exacerbate private-side problems. (I suggested one idea, though without substantial detail. If someone were to pay me to study it–you know, rigorously, quantitatively–I’d have more to say. Meanwhile, back to thinking about ideas I’m actually paid for …)

I bet you think I’m going to discuss the housing market. Nope. There’s another kind of foreclosure. It’s also a market phenomenon that occurs when the actions of one firm prevent another from remaining in or entering a market.

Foreclosure can be brought about by some of types of market predation I described yesterday. In particular, ”vertical behavior” can lead to foreclosure. Vertical behavior is any action taken by a firm to more tightly control its supply or product chain. When a baker buys a wheat distributor, he has vertically integrated. He now owns a source of one of his inputs. An exclusive contract between a movie production company and a cable television network is a “vertical restraint.” It “restrains” that network from offering rivals’ content.

In their Handbook of Health Economics chapter on antitrust and competition in health care markets, Gaynor and Vogt describe the relationship between vertical behavior and foreclosure in health care:

The other major form of vertical behavior that has been of concern in health care antitrust have been vertical restraints that tend to reduce competition in the of the markets involved. These include vertical integration, exclusive dealing, and most-favored-nations contracts. A commonly used term for this effect is “foreclosure.” The reason for concern is obvious. Consider a situation with a health insurance duopoly and a hospital monopoly. If one of the insurers integrates with the hospital or engages it in an exclusive contract, it will have the ability to foreclose the other insurer from the market, thereby gaining monopoly power. … Since vertical restraints both involve potential anti-competitive effects and efficiencies, antitrust cases involving (non-price) vertical restraints are judged on a rule of reason basis. This makes economic analysis of effects on competition and efficiencies essential in such cases. …

The courts for the most part have found … insufficient evidence of anti-competitive effects. As indicated previously, the vast bulk of exclusive dealing cases or vertical integration cases have been rejected by the courts impacts

Not surprisingly, the authors note that the economics literature has found that vertical integration and restraints have been found to be efficiency enhancing, lowering firm costs. Hence, one might expect insurers to enter into exclusive or long-term contracts with providers.

Gal-Or (1996) considers the … [exclusive contracting] problem … with differentiated insurers. With differentiated insurers foreclosure can occur in equilibrium. In this case, a provider who agrees to an exclusive deal with an insurer will likely accept a lower payment rate in return for a larger volume of patients. If both insurers sign exclusive deals with different providers, this benefits insurers by reducing the outside options of the providers and thus reducing their payment rates. Encinosa (1996) considers exclusive deals between HMOs and physician groups. There is an incumbent HMO which has a cost advantage over a rival, but must invest in order to serve the entire market. When the incumbent HMO is risk averse, it may engage in an exclusive deal with the single provider. This will result in foreclosure and is socially inefficient. At present, however, exclusive contracts per se appear to be relatively rare between insurers and health care providers. Long term services contracts are common, and may confer a degree of exclusivity on an insurer who is a large buyer.

It is worth noting that the Gaynor and Vogt chapter from which I quote about a decade old. I do not know if exclusive contracting between hospitals and insurers has become more common.

(See also: Anticompetitive exclusion: Raising rivals’ costs to achieve power over price, by Krattenmaker and Salop. There the authors describe exclusionary rights as the “bottleneck” or “essential facilities” problem when it locks up the entire supply the the “supply squeeze” or “quantitative foreclosure” when it locks up a portion of it. Also described is the “price squeeze” or “cartel ringmaster” when a dominant firm facilitates a discriminatory cartel among suppliers. Finally, there’s the “Frankenstein monster” (!!!) which comes about from an exclusive contract with one supplier that results in a residual market structure that facilitates collusion among the remaining ones faced by rivals.)

References

Gal-Or, Esther (1996), “Exclusionary equilibria in health care markets,” Journal of Economics and Management Strategy, 6(1): 5-43.

Encinosa, W.E. (1996), “Exclusive contracting in health care markets,” unpublished manuscript, University of Michigan.

Today in the Washington Post, Alec MacGillis has an article about the Carilion Clinic’s provider network in Roanoke, Virginia (h/t Ezra Klein’s Wonkbook).

Carilion owns the two hospitals in town and six others in the region, employs 550 doctors and has set off a bitter local debate: Is its dominance a new model for health care or a blatant attempt to corner the market?

Carilion says it represents an ideal envisioned by the nation’s new health-care law: a network that increases efficiency by bringing more doctors and hospitals onto one team, integrating care from the doctor’s office to the operating room. The name for such networks, which the new law strongly promotes with pilot programs, is accountable care organizations, or ACOs — providers joining together to be “accountable” for the total care of patients, with incentives from insurers to keep people healthy and costs down. …

But skeptics apply a more old-fashioned term to networks like Carilion: monopolies, which they say will make health care even more expensive.

The article goes on to speculate about implications of provider integration for costs, quality, and prices. Turns out Cory Capps, economist and published expert on the implications of hospital consolidation, recently shared a working paper with me that pertains to these issues, at least with respect to hospital consolidation. He has authorized me to use its content in blog post.

Let’s begin with some history. Capps takes us back to 1997.

By the standards outlined by the DOJ and FTC in the Horizontal Merger Guidelines, most MSAs were already highly concentrated [HHI > 1,800] by 1997, when the simple average [hospital] HHI within a MSA was over 4,000. By 2006, the average HHI rose an additional 299 points. Weighting MSAs by admissions, the average 1997 HHI was still over 2,000 and rose by 253 points by 2006.

(For the uninitiated: link to the definition of “MSA“; link to the definition of “HHI“; for anything else, Google it.)

The figure below (click to enlarge) illustrates hospital consolidations over time (number of mergers and acquisitions or M&A, red line, right-hand-side scale) and yearly change in hospital inpatient spending (blue line, left-hand-side scale).

The figure shows that the hospital industry experienced a wave of mergers in the 1990s and the rate of inpatient expenditures increased thereafter. A common explanation for these phenomena is managed care. As managed care peaked in the 1990s, hospital price competition intensified, pushing hospital costs and prices downward. Hospitals responded by merging to achieve economies of scale and scope while increasing bargaining power. After the managed care backlash, price competition was lower because of looser networks and increased hospital market concentration following the period of mergers.

Vogt and Town conducted a survey of literature on the relationship between hospital consolidation and concentration and costs,quality, and pricing. Capps summarizes the findings in each of these dimensions.

Costs. “Overall, Town and Vogt’s conclusion from their survey of the cost literature as follows: ‘[t]he balance of the evidence indicates that hospital consolidation produces some cost savings and that these cost savings can be significant when hospitals consolidate their services more fully.’”

Quality. “The majority of studies to date, however, conclude that hospital mergers and acquisitions have either no effect or a modest negative effect on quality, with the former finding being the more common.”

Pricing. “There is substantial evidence that hospitals compete within a fairly narrow geographic area, often smaller than a city or an MSA. Mergers within such a narrow area can lead to substantial price increases. Increases are most likely if the consolidation combines hospitals that, from the perspectives of insurers assembling provider networks, are close substitutes.”

If hospital mergers have such a significant effect on prices, why didn’t antitrust regulators do something? Well, they tried and failed, until they gave up. Capps:

From 1993 through 1998, the FTC and DOJ lost six consecutive hospital merger challenges; in 2001, the State of California lost a seventh. In the decade after the last of these losses, 1998 to 2008, neither the FTC nor DOJ challenged a prospective hospital merger in court. Over the 15 years spanning 1993–2008, antitrust policy likely had little restraining effect on hospital mergers over this period.

Capps estimates the consequence of hospital concentration in areas where the population is large enough to support multiple hospitals (i.e. a less concentrated market) is to increase national health expenditures by $10-$12 billion annually, which is only 0.4-0.5% of national health expenditures.

In an e-mail Capps explained to me that he was deliberately conservative in his estimate, both in its calculation and in its presentation. There are reasons to think the increase in health care costs due to hospital concentration is higher and more significant:

• Medicare and Medicaid prices are set administratively so are not directly influenced by market power. Only the 43% of total spending that is private spending could be affected by market power. Relative to that smaller base of health spending the increase in hospital market power raises prices by 1.2%, not 0.5%.
• If you focus further on private hospital spending, the price effect of market power goes up to 5%.
• In order to be conservative, he drew a pretty tight line for defining MSAs that could support a less concentrated hospital industry. He thinks that it is reasonable to expand the definition and, based on that, he argues his figures may be as much as 70% too low. That could push the price effect to 7% of private hospital spending.
• Consolidation for physician services arguably had an equivalent effect, doubling the impact of provider market power.
• Also, it’s worth noting that the distribution these costs is likely not uniform but focussed on a subset of Americans in highly concentrated markets.

Bottom line, hospital consolidation matters, increasing health care costs by tens of billions per year and, in general, not delivering higher quality. It’s not the only factor in health care prices, and it may not be the most important one. But it is arguably pushing hospital prices for private payers up by at least 5% nationally.

Will wider integration of hospitals and non-hospital based physician groups, such as that achieved by Carilion, improve quality without increasing prices? It remains to be seen.

Product differentiation is the term for variations in characteristics of products within the same market. Without differentiation, products are commodities. Corn is corn, oil is oil, and so forth. Most consumer products aren’t like that. Corn flakes are not the same as raisin bran, but they compete with one another in the breakfast cereal market.

The concept is fuzzy. It’s a continuum really. When do the differences between two products cause them to be in different markets altogether? It comes down to the degree of substitutability. Post’s version of corn flakes is highly substitutable from Kellogg’s (presuming they both make such a thing). But it is slightly less substitutable for General Mill’s Cheerios, or so the thinking goes. (Really, product characteristics are multidimensional so it’s possible General Mill’s Cheerios are actually closer to Post’s corn flakes in some sense.)

But they’re all breakfast cereals. They are designed to fill the same consumer desire to eat something crunchy with milk. One could argue whether bagels, waffles, and eggs are in the same market or not. They are substitutes, but maybe they’re so different that the degree of substitutability is weak enough that they shouldn’t be lumped together with breakfast cereals. What about oatmeal? (See, complicated!) We can say for sure that iPods don’t belong. You can’t eat them for breakfast (or shouldn’t).

So, it gets complicated. That’s a challenge for economics theory and econometrics. In a world of perfectly substitutable commodities, theory is relatively easy. It is also the world we tend to imagination when we contemplate markets. That’s where some of our intuition can go awry.

Take hospitals, for example. If they’re perfect substitutes then the market power of any one of them is a function of the number of competitors and consumer demand for hospital services. There’s no heterogeneity in market power. One hospital is just like another. Reasoning is easy.

But that’s not how the market works. Hospitals distinguish themselves from one another (differentiate) in many ways, as do other products. Location is a simple distinguishing feature. Specialization in a particular type of care is another. The list is long.

Such differentiation gives rise to the possibility of “star” hospitals, those that have brand power. They are the ones that are hard for insurers to exclude from their networks. They command high prices. In fact, they have monopoly pricing power and can charge markups over marginal costs.

So, that’s reality. And it’s relatively harder to model it, though many techniques exist to do so. They tend not to fit well on the back of an envelope. They’re more complicated than a commodities market.

(All of the above applies to the health insurance market too. Products differ and are not perfect substitutes.)

For further reading in the health economics literature on hospital differentiation, below are a few references that point to others. I grabbed these from Abraham, Gaynor, and Vogt (2007). I know there are many more and am not claiming these represent the full complement of approaches to the issue. It’s a place to start.

Further Reading on Hospital Product Differentiation

Abraham, J.; Gaynor, M. and Vogt, W., 2007, Entry and Competition in Local Hospital Markets, The Journal of Industrial Economics, Vol. LV (2), pp. 265-288.

Capps, C.; Dranove, D. and Satterthwaite, M., 2003, Competition and Market Power in Option Demand Markets, Rand Journal of Economics, Vol. 34 (4), pp. 737–763.

Gaynor, M. and Vogt, W. B., 2000, ‘Antitrust and Competition in Health Care Markets,’ in Culyer, A. J. and Newhouse, J. P. editors, Handbook of Health Economics, Vol. 1B, chapter 27, (Elsevier Science B.V., Amsterdam). pp. 1405–1487.

Gaynor, M. and Vogt, W. B., 2003, Competition Among Hospitals, Rand Journal of Economics, Vol. 34 (4), pp. 764–85.

A 2009 American Hospital Association paper expansively titled “The Case for Reinvigorating Antitrust Enforcement for Health Plan Mergers and Anticompetitive Conduct to Protect Consumers and Providers and Support Meaningful Reform“ (whew!) includes the following quote from an unnamed U.S. Health CEO about it’s 1996 merger with Aetna:

[We did the deal] to get the mass we needed, the power to negotiate with the physicians, hospitals, the drug companies and force down their charges.

Nothing surprising about this. Just thought I’d give an insurer CEO equal time, having already posted quotes from hospital CEOs about market power.

If the deal did what this CEO says it was intended to do–reduce provider prices–what happened to premium levels relative to what they would have been without the deal? The answer to this question is relevant to how we should think about increasing (or decreasing) market concentration in the insurance industry. It’s not always bad (or good) for consumers.

Katherine Ho’s paper titled “Insurer-Provider Networks in the Medical Care Market” (American Economic Review, 2009; ungated version available) models the bargaining process between insurers and hospitals. It’s fascinating, but far too mathy for a post. (If I can distill it down to something simpler I’ll write more about it later.)

The paper also includes this revealing passage:

The executive director of one hospital system described a potential outcome in such markets [in which managed care is strong and hospitals compete for contracts]: “There are examples where there were too many hospitals in an area and the plans played them off against each other to the point where the price paid was no more than marginal cost.”

The more interesting situation arises when hospitals tailor their characteristics in order to capture positive profits. Interviewees noted that the negotiations could be very different in these markets. A hospital director said the following: “In market X [where hospitals are very strong], the prices [the best hospitals] charge are based on their very high patient satisfaction results and their strong reputation. They can get high prices from any plan in the market and they don’t need them all.” The CEO of a small hospital in a different market had a similar story: “Large [hospitals] in this market can dictate whatever prices they want. The bigger names can demand the higher prices.”

Market structure matters. If we’re to rely on the market for health care the prices (premiums) we pay are a function of the balance of market power between insurers and providers. That’s not the only factor relevant to prices, but it is one driver of geographic and temporal variation in premium levels and changes.

I’ve wondered what the role of non-profit hospitals will be once nearly all Americans have health insurance. The problem of uncompensated care will largely go away. Will the non-profits retain their tax-preferred status? One way they might not is via merger with a for-profit hospital. Jenny Gold of Kaiser Health News reports that some hospitals are considering doing just that.

[H]ospital mergers and acquisitions tend to go in boom-and-bust cycles. “My guess is that this is entering a period of expansion again,” says Gerard Anderson, director of the Johns Hopkins Center for Hospital Finance and Management.

One big reason is that the new health overhaul law will eventually extend coverage to an additional 32 million people, reducing the financial burden of the DMC and other hospitals that treat a lot of uninsured patients. “Health reform gets rid of a big chunk of the uncompensated care problem,” making urban hospitals more attractive acquisition targets, says Jack Wheeler, a professor of health management and policy at the University of Michigan.

Hospitals are merging while the political focus is on increasing competition in the insurance market. If market power swings (further) in favor of hospitals, there’s really very little reason to be optimistic about a market-based solution to reducing health care costs on the private side. The only element of the new health reform law that might help is the Cadillac tax. That doesn’t kick in until 2018 and won’t have a substantial effect for at least a decade after that. I don’t think the American public is going to be satisfied with the rate of premium growth over the next two decades.

I’ll just throw in the relevant figure again (below). Moving in the direction of “A” is probably more costly than in the direction of “C” because most health care premium dollars (~85%) collected by insurers flow to providers. (In math speak: the slope between “A” and “B” is larger in magnitude than between “B” and “C”.) Thus, all other things being equal, increasing market concentration among providers leads to a higher absolute markup in prices (premiums).

By the way, this figure isn’t just based on my own imagining. I’ve shown it to many health economists and all agree it has the right shape. Also, I’ve roughed out some theory that reproduces it algebraically. I’m confident in saying that the health care market really has the qualitative characteristics the figure suggests.

Uwe Reinhardt puts his finger on the issue I’ve been puzzling about for over a year, and blogged about many times. How can increased competition among insurers decrease health insurance premiums when the vast majority of premium dollars flow to medical providers? On the Health Affairs blog, he writes,

The bulk of the medical benefits procured by an insurer for residents in a given market area are produced by providers within that market area. In general, both private and public insurers have only limited, if any, control over the volume of the medical benefits that local clinical decision makers ask insurers to purchase for the insured. Furthermore, the larger the number of insurance companies active in a local market, the smaller any insurer’s market share will be — other things being equal — and the less leverage any insurer will have in bargaining with area providers over the prices of health care. …

The current nouvelle vague – so-called Accountable Care Organizations (ACOs) – will only further encourage that concentration. I find it hard to believe that, in the face of this trend, fragmenting the buy side of health care even more would serve the goal of cost containment.

Ideally, in my view, the market for health insurance would be oligopolistic, which means that only a few insurers — each with some market clout vis à vis providers — would compete for enrollees in a local market. What the ideal number would be is an interesting question on which economists can have a lively debate.

So what am I missing here? Why do so many otherwise sensible people believe that fragmenting the buy side of the health care market even more than it already is will help contain the rising cost of health care? I would argue just the opposite.

I invite readers and fellow bloggers to enlighten me.

Reinhardt should not hold his breath. I can think of no sensible argument that–holding all else constant–starts with a reduction in insurers’ market power and ends with a decrease in medical costs and health care premiums. However, if one is willing to add in other elements of reform, I can think of some possibilities that include increased insurer competition. For example, if insurers are permitted to collude to set all-payer rates to medical providers then they can maintain a high degree of leverage over providers while competing vigorously for policyholders. Or, the provider market could be commensurately diluted so that the relative provider-insurer balance of power was held constant.

Do you think for a second either of those things will happen (all-payer rates setting or the break up of dominant hospitals)? I don’t either. If insurers can’t constrain medical costs in the private health care market we’re likely to get, what will? This question has not yet been answered or even sufficiently debated.

In a comment, steve makes a good point, “[T]he rumor is that my hospital is going to buy up practices in anticipation of an ACO model.” This is consistent with what I blogged several days ago. Providers and insurers are consolidating now in anticipation of Medicare payment system reform that will reward integrated delivery systems.

This is interesting because we don’t yet know for sure what the new payment system reform will be. Whether it really rewards consolidation or not, consolidation is what we’re getting.

In a comment yesterday, steve wrote,

Has anyone looked at the issue of market power in the context of rural vs urban providers? Most of the cases I have seen cited seem to identify large urban facilities, often university affiliated, as market powers. I would think that isolated, rural hospitals would have the same negotiating power, yet my experience would indicate that it seldom works that way.

It just so happens that today I was reading “Can Hospitals And Physicians Shift The Effects Of Cuts In Medicare Reimbursement To Private Payers?” by Paul Ginsberg in which he wrote,

The potential for cost shifting is also likely to vary geographically. Hospitals probably have more market power in smaller communities because concentration tends to be higher. Providers in small communities also might face more effective pressures to keep rates as low as possible. Both combined suggest that cost shifting has the potential to be more extensive. Indeed, many in contact with the insurance industry have noted sharp increases in rural hospital payment rates to private insurers in response to Medicare payment rate reductions resulting from the 1997 BBA. Rural hospitals, which tend to be natural monopolies, could have held rates way below their potential, so that when Medicare payment rates fell in relation to their costs, they could then raise rates to private insurers.

By the way, Ginsberg’s paper does a very nice job of laying out the most basic variants of cost shifting theory. There are more nuanced and complex theories out there that he doesn’t describe (not a critique, just a fact).