• The medical arms race

    This is a TIE-U post associated with Jonathan Kolstad’s The Economics of Health Care and Policy (Penn’s HCMG 903-001, Spring 2012). For other posts in this series, see the course intro.

    Standard economics theory teaches us that competition increases “social welfare.” But that theory is premised on a market with certain attributes that are not always present. (Strictly speaking, they’re never present, but some markets get “close enough.”) In particular, health care markets (e.g., for hospital services) deviate considerably from perfection. Consequently, competition may either increase or decrease social welfare. It’s an empirical question.

    In “Is hospital competition socially wasteful?” Daniel Kessler and Mark McClellan [1] (ungated pdf available) take up that question. They study elderly Medicare beneficiaries hospitalized with cardiac conditions between 1985 and 1994. Since this period overlaps dramatic change in the nature of insurance (moving from indemnity coverage to HMOs), the authors are able to explore the impact of network contracting and managed care on the relationship between hospital competition and spending and outcomes.

    We find that, before 1991, competition led to higher costs and, in some cases, lower rates of adverse health outcomes for elderly Americans with heart disease; but after 1990, competition led both to substantially lower costs and significantly lower rates of adverse outcomes. Thus, after 1990, hospital competition unambiguously improves social welfare. Increasing HMO enrollment over the sample period partially explains the dramatic change in the impact of hospital competition; hospital competition is unambiguously welfare-improving throughout the sample period in geographic areas with above-median HMO enrollment rates. Furthermore, point estimates of the magnitude of the welfare benefits of competition are uniformly larger for patients from states with high HMO enrollment as of their admission date, as compared with patients from states with low HMO enrollment.

    The analysis is based on an assumption that the relative distance from a patient’s residence to two different hospitals is uncorrelated with unobservable factors that also affect cost and health outcomes. This type of “differential distance” instrument is not universally accepted as valid. One concern a critic might have is that hospitals may select their relative locations (or arrive at them via closings, mergers, etc.) in ways that do depend on important, unobservable patient factors. Obviously a hospital is not sited based on the characteristics of a single patient, but they may be sited based on the characteristics of a population and with knowledge of competitors’ locations, which some worry could introduce a systematic bias.

    Having written all that, let’s, for the moment, take the results as given. How does competition among hospitals drive spending up in the pre-1991 period? The classic answer is “the medical arms race.” When insurers were relatively weak, pass-through entities, as they were before the era of managed care, hospitals did not compete on price. They competed on technology and amenities. In trying to one-up each other to attract patients, they drove cost and price up. Insurers paid it so what did the hospitals care?

    The incentives of managed care (network contracting, among other things), wrung some of this from the system. Hospitals were rewarded for low prices. In a competitive market, HMOs would cut hospitals out of their network if prices rose too quickly. In such a market, technology and amenities that didn’t drive costs down were no longer profitably offered.

    This medical arms race story parallels the cost shifting one I’ve told before. Like cost shifting, the medical arms race is slowed by insurers with market power in markets with sufficient competition among hospitals. As hospitals continue to consolidate and integrate with other providers (e.g., as encouraged by the ACO movement), I wonder if the medical arms race will return. If the literature supports it, an enterprising researcher could write a review article of the medical arms race literature that is the analog of my Milbank Quarterly cost shifting one. The conclusion might speculate on what we might expect in the coming years.

    I’ve thought about working on such a paper myself, but I’m not likely to, not on my own or soon anyway. So, it’s just sitting there for the taking. However, for that reason, economic theory suggests that this cannot be a valuable opportunity. On the other hand, as in the social welfare of competition in hospital markets, economic theory can be ambiguous.


    Kessler, Daniel and Mark McClellan (2000). “Is Hospital Competition Socially Wasteful?” Quarterly Journal of Economics, pp. 577-615.

    • “However, for that reason, economic theory suggests that this cannot be a valuable opportunity.”

      Should I go back and pick up that $20 bill? More seriously, I dont think people are totally aware of the degree of consolidation going on. It is not just hospitals. Specialties are consolidating. A single GI, cardiology or anesthesia group may cover many. many facilities. Groups like Mednax which concentrated on consolidating neonatologists are not trying to take over other specialties. My last few throwaways (real newsletter also) have had articles about how to sell your group profitably to one of these consolidating entities.


    • -I can’t help but wonder how helpful you actually find standard economic theory (neo-classical equilibrium theory) helpful as a guide for policy evaluation and design.

      I think it’s been an enormously useful enterprise for evaluating a small-subset of idealized exchange interactions, but it’s almost entirely worthless as a tool for constructing or evaluating policy choices in a world that has virtually nothing in common with the set of computational exercises at the heart of post-Samuelson neo-classical econ.

      Do we really need recourse to platonic formalisms to determine whether the outputs from a competitive enterprise conducted in a world of imperfections is more likely to deliver higher quality outputs at a lower cost than a statutory monopoly in every case except those where there’s actually technically sound justification for a monopoly?

      Is it really that tough to evaluate the relative attributes of the Trabant and the Mercedes and determine which process led to the better car?

      In your opinion, is neo-classical econ, and the sub-domain of Welfare Economics actually a sufficient tool for evaluating policy choices? If not sufficient, how about necessary?

      Is there something in particular about health care that makes these tools more useful? Is there any realm of economic activity that actually delivers tangible outputs that has used the formalisms embedded in welfare economics to deliver a better product at a lower price? If not – what does that say about the prospects for using the precepts of welfare econ coupled with an inherently faulty political apparatus to improve the quality and reduce the cost of health care?

    • Hospitals are incredibly complex. They not only compete for patients, but the doctors who use their facilities are often not employees, meaning to be the best, they need to compete for the best doctors. Usually by offering first class facilities.

      In the Boston market, which has plenty of hospitals, a health care plan would have trouble not covering the perceived best hospitals. So, being one of the best, provides some protection from price pressure.

      What the premier hospitals did was to buy up non-premier hospitals and bargain on price as a group, getting the price premium for the non-premium hospitals. Competition decreased and prices went up.

      Anyone who’s been in the situation of needing emergency care and having to read the fine print of their insurance policy while waiting for the ambulance is going to be vocal about their dissatisfaction.

    • Good post. The article is eye-opening and, if it holds up, I hope it gains wide currency as an antidote to the “managed care in the 90s was all about denying care” mantra.

      But I was momentarily stunned by the question “I wonder whether the medical arms race will return.” It did return, all the way back in 1998 or 1999. Three pieces of evidence for that:
      1. Cost trend shot upward after being more or less under control for about 5 years.
      2. Aetna (or United, I forget which was first) announced that they would no longer engage in hardball patient-directed tactics like aggressive UM. Fewer required referrals, etc.
      3. Insurers began moving away from capitation and back to almost entirely fee for service payment. (For hospitals, per diem vs DRG vs percent of charges)
      4. insurers decided to return to being more of the pass through entities of old, profiting by taking a cut off the top and competing on the size of the network rather than the degree of medical cost control. As you know, larger networks are more expensive, especially when it means all significant hospitals must be included and the hospitals can use that for leverage.
      5. There has been a building boom for hospitals going on for about a decade. The arms race doesn’t just take the form of fancy new equipment, but fancy new facilities.

      The backlash against managed care was directly responsible for #2, 3 and 4. I know it was widely discussed in the industry by 2004, when the explosion in costs had brought pressure again for new ways to manage care, but the industry was not in a good position to do so.

    • Oops, I meant five pieces of evidence, not three.

    • I would love to write a review on this topic if no one else is already doing it.