Journal of Health Politics, Policy and Law

Regulating Managed Care:
Pulling the Tails to Wag the Dogs

David M. Frankford

The health-policy airwaves are currently filled with talk of regulating managed care to protect "consumers." In the name of "consumer protection" numerous states have already passed legislation to regulate managed care and a multitude of proposed additional state legislation, as well as major bills in Congress, are pending. In this essay I question much or most of this enterprise, concluding that little public good is likely to be attained from regulation of managed care within the confines of our current market-based, competitive private insurance system (including the spillover effects on major public programs like Medicare and Medicaid and the patchwork of public programs aimed at smaller populations, such as the Children's Health Insurance Program). The allied culture and institutions of dog-eat-dog are simply too deeply entrenched. The essay concludes by noting briefly that the same may be true in a compulsory, universal system in the United States and by noting that scholars like me who are sympathetic to "consumer protection" are put into a real dilemma because we must be concerned that managed care regulation serves to legitimate and reinforce a system that is inequitable and prudentially flawed. ¹

We can divide actual and proposed managed care regulation into two groups, one termed procedural and the other substantive. By "procedural" [End Page 1191] I mean regulating the manner in which managed care organizations (MCOs) make decisions like those concerning coverage and medical necessity. Examples include legislative and administrative requirements specifying (1) the manner in which MCOs are to notify providers and beneficiaries of adverse decisions; (2) the internal procedures to be used in making those decisions, including the existence and nature of internal review; (3) whether external review is to be available and if it is, the identity of the external reviewer and whether or not the review is de novo; and (4) the timing of internal or external review. By "substantive" I mean regulation of the substantive standards used in decision making. As an example, colleagues and I have recently proposed that Congress amend the Employee Retirement Income Security Act (ERISA) and mandate the substantive standards that would govern medical necessity decisions (Rosenbaum et al. 1999). Other prominent examples include laws that require the inclusion of particular benefits (e.g., mental health parity), limitations on the amount of risk that can be "downstreamed" to providers (e.g., 25 percent of all patient-care income), and regulation of the standards that can be applied in the selection and deselection of providers (e.g., whether "economic credentialling" is permitted).

For the most part, procedural reforms dominate current state and federal regulatory agendas. However, as my colleagues and I have written, "No amount of procedural protection can help patients if insurers are given broad power to determine [substantive standards]. Such 'protection' would be an instance of winning a battle but losing the war" (ibid.: 229). For that reason, in this essay I will focus only on regulation of the substantive aspects of managed care.

We must first examine the process of competition that managed care involves. Pioneers like Paul Ellwood did not necessarily include market competition as part of their ideal for managed care, although they often did. Regardless, that is the current reality and very much part of the resultant backlash. What is the nature of this competition?

The word "competition" is an abstraction, and we must always be careful to specify what competition consists of. It can take many forms. There is, first, competition over price. Producers attempt to gain market share by lowering price to the extent they can remain profitable. When people write or talk about "competition," they are often using the word in this sense, but that price competition is certainly not the only form competition can take. Competition can occur, second, over many nonprice dimensions of a product. Producers might, for example, not lower price [End Page 1192] but improve quality, offer extended warranties, have longer showroom hours, and the like. They might also, third, compete--more precisely, try to limit competition amongst themselves--by aiming their products for certain geographic markets, certain types of customers, and the like, all strategies of "market segmentation." As an example, a producer might attempt to find a market niche selling specialized books to an Upper East Side clientele and thereby essentially exist in a "market" that is somewhat separate from that in which the megachains operate.

Reasons that explain which type of competition exists in a given market need not detain us; the point is that at least since the rise of commercial health insurers, to the extent competition existed, insurers have competed largely over price--premiums. The introduction of greater degrees of competition through relaxation of the antitrust laws and similar efforts in the 1970s and 1980s, leading to and coupled with the destruction of community rating, has intensified and made more widespread the degree of price competition.

In our fragmented insurance "system," competition over price has caused and accelerated the use of a second form of competition, namely, market segmentation, which in this context means cherry picking among types of customers and geographic areas. Risk selection is the obvious example of the former, in which through various means insurers avoid certain types of individuals or groups. Keeping the "good risks" in and the "bad risks" out offers a means of reducing premiums and imposing costs on rivals, which causes rivals' prices to rise and their concomitant loss of market share. Related are the extremely high premium differentials between group plans and individual ones, among different sized group plans--meaning that small employers in particular are paying more--and, last, the redlining of certain economic sectors (like the arts community, that vessel of HIV) to the extent that activity is legal and can occur undetected. By contrast, market segmentation by territory is well illustrated by the withdrawal of Medicare HMOs from selected areas. Despite their whining that their Medicare capitation rate is too low, the HMOs are cherry picking the program, staying in lucrative local markets but withdrawing from less profitable ones.

To some extent, these forms of market segmentation are due to bargaining-power differentials and the use of experience rating. A large employer with thousands of relatively healthy "covered lives" in tow brings much more to the table than a small employer's group of fifteen (which is in an even worse position if one of the fifteen is catastrophically ill). [End Page 1193] In this essay I can ignore these causes of market segmentation because premium regulation is probably necessary to eliminate the differentials and such regulation is unlikely in the current political context.

By contrast, some of the responsibility for the market segmentation rests on insurers' strategic use of information against different buyers, and much of the managed care regulatory effort is focused on these "information problems" in "imperfect markets." In theory, a workably competitive market can be policed by the existence of a few well-informed buyers. In such markets, producers are unable to determine which consumers are ignorant and which are knowledgeable. In other words, they cannot differentiate--the technical term, aptly, is discriminate--among types of buyers according to the buyers' divergent levels of sophistication and knowledge. Hence, any firm that attempts to discriminate among consumers in this way will lose market share to competitors. As a result, the price-quality mix of the market is driven by the decisions of the well informed--that is, where the market "clears." However, the market for managed health insurance fails to conform to the theory.

It is important to be clear about the information asymmetry here. Much of the writing in health economics and other literature has concerned the information asymmetries between patients and providers or providers and insurers. That's not what I'm discussing now. Instead, the information asymmetry here concerns the manner in which MCOs arrange goods and services and administer benefits. Regardless of stipulated price and formally stated contractual obligations, the plans use many subtle means to reduce the amount of insurance they actually provide. Strategies include cutting back on coverage through coverage determinations that are undisclosed and often invisible to the plans' customers and its providers; aggressive use of medical necessity determinations to deny benefits, again most often in a process that is invisible; and the careful construction of provider networks to avoid coverage of costly populations (e.g., avoiding black doctors to avoid their patients), another activity that is very hard to detect. Further, the plans fight tooth and nail to keep these strategies under wraps. They will not disclose their "proprietary data" to anyone unless they are absolutely forced to do so; and they have spent large resources lobbying the Congress and the Department of Labor to maintain this status quo. There also seems to be a recent trend of insurers' settling litigation, and a quid pro quo of settlement is often that the records are sealed. Evidence of actual practices, then, stays hidden. True, some of these activities are being revealed in press accounts, which is part of the "backlash," but the plans can successfully argue-- [End Page 1194] aided by many health service researchers and policy analysts--that this evidence is "anecdotal," since the MCOs so successfully restricted access to the information that buyers, regulators, and researchers need.

The point, then, is that MCOs have superior access to the information concerning their activities, and they are able to differentiate among consumers according to the relative degrees to which different types of customers are at an informational disadvantage. ² This is the strategic use of information, which occurs at multiple levels in the health care system. Admittedly, often the plans themselves do not know what services they are providing under their contracts--Oxford's financial problems are hardly unique because many plans lack adequate information systems. Yet the point remains that to the extent that information exists, plans can and do discriminate among types of buyers. Even large, sophisticated, repeat-playing private buyers routinely complain that managed care is a "black box." Further, Sara Rosenbaum et al. (1998) have amply demonstrated that state and local Medicaid administrators who are engaged in risk-based contracting neither understand nor have the necessary information concerning the composition of the managed care packages they are purportedly buying. If that is the case for large public agencies and numerous large private buyers, the situation is much more serious for smaller businesses (and individuals!) that cannot possibly devote the resources to "buying right" that, say, California and Xerox can. McClure (1990) and others' solution of creating group purchasing to enable relatively ignorant buyers to buy right might help level the playing field [End Page 1195] among consumers, but it is unclear how inclusive these groups can or will be--there's no rapidly rolling bandwagon out there--and alone group purchasing cannot solve the informational asymmetry and agency problem that are the root causes of this form of discrimination among plan customers. The degree of the problem is indicated by the fact that some large, self-insured plans are starting to contract directly with providers, bypassing managed care plans so that they can know what it is that they are actually buying. Additional mechanisms to "facilitate" the market are necessary.

In the Information Age, the primary strategy of managed care regulation to stop this form of discrimination and change the nature of competition is to increase the amount of information available to consumers, individual or collective, public or private, large and small, through the creation of information mechanisms like "report cards" and HEDIS-type measures. However, report cards are charitably described as crude at best (and I do not mean to insult my friends who are working hard on such efforts). A report card now might include, say, eight categories, examples of which are performance measures regarding childhood immunizations, control of hypertension and asthma, and the level of prenatal services. These measures comprise the tiniest tip of a gigantic iceberg below, and do not concern the lion's share of what consumers want and need to know. More sophisticated measures of specific and harder to measure procedures exist, like New York's famous risk-adjusted coronary artery bypass grafting mortality statistics. However, even assuming that these reports are methodologically sound--itself debatable because of questions over the validity of severity-of-illness controls and the gaming of the categories by providers--they are of only very limited utility. For much if not most of health care, the question is whether the diagnosis and proposed course of treatment are right, not just the capabilities of those performing procedures: I want to know whether I should be on that operating table to begin with before I even worry about whether the surgeons who open up my chest have good risk-adjusted outcomes. These methods of scoring cannot answer such questions.

Managed care reform premised on HEDIS and similar measures suffers from the problem just discussed, but these performance measures are far more refined than the simple report cards now being touted. Yet numerous problems remain. First, when those sophisticated buyers who are buying right are about to buy, they must choose among competing plans. The point of HEDIS and similar efforts is to enable interplan comparison by standardizing measures across numerous plans. However, for [End Page 1196] HEDIS-type measures to apply to those plans, the plans must be large enough to enable the methodologically required amount of data to be collected from each plan for valid statistical inferences regarding plan performance. Most plans, however, are not of sufficient scale. The implication is that these types of reporting mechanisms could be valid across plans only if a massive consolidation in the marketplace occurs. Yet, the logic is internally inconsistent because it is contradictory to claim that the elimination of competition protects consumers in a system premised on the supposition that patients are to be protected through competition.

The second difficulty with measures like HEDIS is that currently they rely on unaudited, self-reported data, a situation which is rather uncomfortable. In the ideal, the National Committee on Quality Assurance would prefer that the data be audited, but so far purchasers are unwilling to pay the high price of obtaining those data. Perhaps auditing could be used on statistically representative samples, but the assumption that the samples can be made representative is heroic at best. Third, to be manageable sources of information, these measures must be aggregated into a limited number of categories. Imagine reading a Consumer Reports in which there were thousands of categories. Yet no one currently knows how to accomplish the necessary aggregation. How do we combine performance measures for gall bladder removal with ones for angioplasty? Fourth and related, individual and collective buyers who are relatively unsophisticated are going to be lost trying to use these aggregations, were they even possible, because one needs information about the information to understand the information; and that returns us to the problem of preventing discrimination against different kinds of buyers. Fifth, the whole idea of this enterprise is protection of individual patients, not their collective buyers who purportedly act as their agents. At present millions of employees have no choice among plans, and the idea of throwing information their way has no relevance. Hence, unless we assume that entities like employers are perfect agents for patients, we're back at the basic problem that individuals do not have the information they need to protect themselves--now against the activities of their agents. Unless the labor markets ensure that this agency is workable, then we must have serious reservations. Sixth and last, even if all these problems can be overcome, we must still account for what the economists term "simultaneous production and consumption" of health care services. Consumers need protection at the time services are delivered, including right now, this day, this year. Even if HEDIS-type measures and report cards can be perfected, that utopia is years away. Moreover, there will necessarily be [End Page 1197] a "data-lag": the problems that are harming patients at year one won't show up until some future performance measures appear some later year or, more likely, years. Patients and their families are not going to feel comfortable with the idea that they are foot soldiers in the quality wars, and they are not going to be consoled by the fact that present suffering will lead to some performance measure in the future.

Two final points must be stressed before we sum up. First, government's imposition of risk-adjusted payments to plans--which is currently technically and politically infeasible--will not solve the problem I discuss. Even if payments were risk adjusted, the plans would still have the incentive to reduce the level of services below those implied by the risk-adjusted premium, and buyers would still need information regarding plan performance. Second, all of these managed care regulatory measures, actual or proposed, affect only those who are currently in the market, private and public. None of them can possibly help any of the 43 million or so who have been dropped out. It is a sad commentary on the state of justice and public discourse in this country that our current debate focuses on protecting consumers against the hard edges of managed care. It is as if the 43 million who are no longer "consumers" have been dropped off the planet.

Overall, then, now and in the foreseeable future, the dream of getting plans to compete over quality remains just that, a dream. ³ The dominant form of competition among managed care plans is likely to remain price competition, with increasing levels of variance in the amounts paid by different buyers. This competition, in turn, will continue the process of separating "high-risk" individuals from "low-risk" ones, a process that is part of the "death spiral" of insurance, continuing if not accelerating the expansion of the number of uninsured and underinsured (and just wait to see what happens when--not if, but when--the next economic downturn arrives). The result will be the increasing instability of the private insurance pool, a problem that can resolve only by massive consolidation. Just witness the recent bankruptcy of HIP of New Jersey and the proposed takeover (subject to regulatory approval) of Prudential's health care books of business by Aetna-U.S. Healthcare because of the Pru's deep financial problems. Further, even if stability is gained through massive consolidation--a real possibility--the incentives to dump individuals and groups into public programs and the safety net will continue. [End Page 1198] This pressure on the public programs and the safety net, in turn, creates political instability in the era of "lean" government, which in turn pressures political actors to try to dump the dumped individuals and groups onto someone else. Doing nothing for the current 43 million or so uninsured is one means of such dumping, and suggestions to add to the problem by restricting Medicaid benefit packages, putting in place defined contribution for Medicare, and increasing the Medicare age of eligibility are all examples of these "fiscal" pressures to dump, dump, and then dump some more.

Where does this leave us? Not in a very good situation. Regulation of managed care without compulsory, universal health insurance is unlikely to protect patients very much. I've previously expressed this concern that our health policies are premised on the assertion of power against countervailing power, which leads to a counter assertion of power, and then another in an endless, fruitless process (Frankford 1997). Students of regulation--even fairly conservative ones like then-Professor and now-Justice Stephen G. Breyer (1982)--could have easily predicted a long time ago that these endless cycles of regulation, industry reaction, and reregulation would occur. At bottom, where there is a will, there is a way. The current marketplace almost forces managed care to take its current form; insurers are acting quite rationally given their situation because rationality is always a product of situation. As Deborah Stone (1993) has written, this is the soul of our current insurance industry. To be effective, our regulation of managed care would have to bring about a massive conversion. Nothing of the sort is currently in the policy airwaves nor likely to be so in the near future. As such, the backlash and regulatory reaction only create political and institutional instability.

This puts scholars like me in a difficult dilemma. Do we participate in these "reform" efforts? My personal answer so far has been in the affirmative, as shown by my writings proposing reforms of the substantive standards used in medical necessity (Rosenbaum et al. 1999) and coverage (Frankford 1996) determinations. Yet even then, the possibility that we are legitimating an inequitable and collapsing system should gnaw at us and spur us on to redouble our efforts for large-scale change that involves at least the creation of a universal, compulsory system, and probably the complete displacement of "competition" as the modus operandi even if competition were to be redirected toward quality. For reasons I've written elsewhere (1997) that cannot be rehearsed here, consumerism is ultimately a dead end. We can do better for ourselves, but that task involves chipping away at a culture and allied institutional [End Page 1199] framework that pits dog against dog. Managed care reform seems contrary to that task. It is an instance of pulling the tails to wag the dogs. Try pulling the tail of a large Doberman sometime, and you'll see what I mean.

Rutgers University School of Law-Camden

David M. Frankford is a professor of law at the Rutgers University School of Law in Camden and a professor at Rutgers' Institute for Health, Health Care Policy, and Aging Research in New Brunswick; Faculty Director of Rutgers' Center for State Health Policy; and a member of Rutgers' Graduate Department of Public Policy and Administration in Camden. He is also the book review editor of the Journal of Health Politics, Policy and Law. His primary current research interest concerns the reconstitution of professionalism as the normative integration of professions and community. In prior work, he has focused on the interactions between health services research, health care politics and policy, and the institutions of professions and professionalism.

Notes

Phil Harvey, Sara Rosenbaum, and Sandy Tanenbaum contributed to this essay with their usual combination of wit and good advice.

1. I ask readers to remember that this is an essay with minimal references, not a full-length treatment.

2. Two subsidiary points must be made. First, the informational asymmetry between managed care organizations and their customers exists whether or not the MCO sells insurance, which I've concentrated on in text, or administrative services for self-insuring groups. To decide whether to insure or self-insure, a group's agent, like an employer, must canvass its options, and to do that, it must compare insured products against third-party-administered self-insurance. Hence it needs information regarding how various third parties perform in administering self-insurance, just as it needs information regarding how various third parties perform when they function as insurers. Even though the incentives of the MCO might differ between the two situations, the problems of obtaining information about the chosen agency--the third-party-managed health insurer or the third-party administrator of self-insurance--persist. In both situations, the entities making the choices are choosing agents, and these purchasers will have different levels of knowledge and sophistication concerning the monitoring of the chosen agent's performance.

Second, if the plans were selling a physical good--the proverbial widget--then arbitrage by buyers could, over time, defeat the plans' discrimination against types of buyers along the lines of sophistication and knowledge. Sophisticated, knowledgeable purchasers, who would be able to obtain a relatively low price or relatively high value for a given price, would buy more widgets than they need and resell the surplus to disadvantaged purchasers, thereby undercutting the plans' discrimination strategy. However, such arbitrage is not practical in the market for the sale of the management of health insurance, for it would necessitate the reselling of management services from one plan customer to other plan customers, and because the product being sold is a complex service, health care.

3. I have doubts whether that dream, still fixated on competition as an ideal, is normatively attractive, but that remains outside the scope of this essay.

References

Breyer, S. G. 1982. Regulation and Its Reform. Cambridge: Harvard University Press.

Frankford, D. M. 1996. Food Allergy and the Health Care Financing Administration: A Story of Rage. Widener Law Symposium Journal 1(1):159-265.

------. 1997. The Normative Constitution of Professional Power. Journal of Health Politics, Policy and Law 22(1):185-221.

McClure, W. 1990. Health Care Reform: The Buy Right Strategy. Remarks to the NEA Retirement and Benefits Forum, Clearwater, FL: NEA, 20 October.

Rosenbaum, S., D. M. Frankford, B. Moore, and P. Borzi. 1999. Who Should Determine When Health Care Is Medically Necessary? New England Journal of Medicine 340(3):229-232.

Rosenbaum, S., P. Shin, M. Zakheim, K. Shaw, and J. Teitelbaum. 1998. Negotiating the New Health System: A Nationwide Study of Medicaid Managed Care Contracts. 2d ed. Washington, DC: George Washington University Medical Center, Center for Health Policy Research.

Stone, D. A. 1993. The Struggle for the Soul of Health Insurance. Journal of Health Politics, Policy and Law 18(2):287-317.