The Rise of Managed Care in the United States: Lessons for French Health Policy

Victor G. Rodwin1
Published with Permission from: Health Policy Reform, National Schemes and Globalization. Ed. C. Altenstetter and J. Björkman. London: Macmillan; New York: St. Martin's Press, l997.
Managed care is not merely a set of fashionable administrative technologies for controlling the use of medical care and the growth of health care costs; it has become the name for a transformation in the way health care is financed and organized in the United States. The common characteristic of managed care organizations (MCOs) is the increasing integration of insurer and provider roles, which has reinforced vertical integration in the health sector. Indeed, the shift from a traditional framework of solo physician practices and independent community hospitals to integrated systems that provide a full range of services financed by capitation payments and relying on state-of-the-art medical and management information systems is a distinguishing feature of the American health care system today.
       The French national health insurance (NHI) system stands out in stark contrast to this evolution of the American health care system. More than most European as well as other OECD nations, France remains a model of 'unmanaged,' solo-based, fee-for-service private practice, altogether separate from a mix of public and private hospitals among which the patient is free to navigate while remaining eligible for reimbursement under NHI (Rodwin, 1981; Rodwin and Sandier, 1993). The question is whether France can escape the kinds of changes sweeping the health care system in the United States? Critics of managed care would argue that French NHI is a superior form of health care organization for controlling costs and for achieving equitable access to good quality care. But the development of new strategies and tools for managing health care systems - MCOs - may be the only affordable way to deliver appropriate health care services in the twenty-first century. This chapter reviews the quiet revolution of managed care in the United States, highlights some of controversies it provokes, and discusses lessons for French health policy.
THE QUIET REVOLUTION OF MANAGED CARE IN THE UNITED STATES
While President Clinton's campaign for universal health care insurance coverage made many headlines, it failed to be enacted. Yet there is another, quieter aspect of health care reform that is transforming the institutional landscape in the private sector: the rise of managed care (Iglehart, 1992, 1994a, 1994b). Over the past decade, large employers have demanded that insures provide not merely indemnity coverage but also a range of benefits within clearly defined budget constraints. To satisfy these demands, private insurers have developed a range of interventionist strategies and tools such as selective contracting with networks of providers (PPOs or IPAs); utilization review - for example, prior authorization of non-emergent hospital admissions and aggressive review of lengths of hospital stay; pharmaceutical benefits management (PBM); introduction of practice guidelines and physician profiling; and outright acquisition or formation of health maintenance organizations (HMOs) and point-of-service plans (POS). POS plans have the special attraction of giving their beneficiaries a choice between using health services within an established network with very low or no copayments, or paying significant copayments and then receiving indemnity reimbursement for out-of-network use. Like HMOs, most POS plan networks rely on primary care doctors to serve as gatekeepers or coordinators.
       Beyond the market for private employer-based insurance, the Omnibus Budget Reconciliation Act (OBRA) has since 1981 encouraged states to initiate a variety of reforms for Medicaid through the budget-waiver application process. Among the fastest growing Medicaid reform initiatives, primary care case management (PCCM) is an example of a managed care strategy that offers beneficiaries access to a primary care provider who either provides services directly or authorizes referral to specialty services. Over ten per cent of Medicaid beneficiaries were enrolled in such programs in 1992 (Hurley and Freund, 1993), nearly 25 per cent in 1995, and the number of new entrants continues to soar (PROPAC, 1995a:16). While managed care for the Medicare population still represents less than ten per cent of the market (PROPAC, 1995b:70), the key congressional commissions charged with reporting to Congress on Medicare have tracked developments in the evolution of the health sector and recognize the importance of adapting the Medicare Program to changes in the organization of medical care (PROPAC, 1995a, 1995b; PPRC, 1995). For example, the Physician Payment Review Commission (PPRC) notes that, given the growth of managed care for the privately insured population, it will be necessary to explore how Medicare beneficiaries cane be assured access to physician networks working for MCOs. Since equity of access to medical care is an essential goal of Medicare, PPRC is concerned that the Medicare population not be excluded from a growing component of the health sector (PPRC, 1995).
        The term 'managed care' has been used to cover a myriad of alternative strategies and tools used by employers, insurers and medical groups to contain costs, improve performance and increase coordination of services for enrolled subscribers of a broad ranged of health care plans. There are already so many taxonomies of MCOs - each reflecting different judgments about which characteristics matter most - that it would be presumptuous to expect evaluation studies to result in useful inferences about managed care in general (Hornbrook and Goodman, 1991; Weiner and de Lissovoy, 1993; Welch et al., 1990). That is why Miller and Luft (1994a), in the best comprehensive review to date of research findings on the performance of managed care plans, proceed with due caution. Miller and Luft (1994b:1512) note that 'compared with indemnity plans, those managed care plans most resembling HMOs had lower hospital admission rates, shorter lengths of hospital stay, the same nor more physician visits per enrolee, less use of expensive procedures and tests, greater use of preventive services, mixed results on outcomes, and somewhat lower enrolee satisfaction with services but higher satisfaction with costs.' They emphasize that the problems of unmeasured selection bias as well as diverse and rapidly changing health plans make it difficult to reach well-founded generalizations about managed care. This is unfortunate in trying to differentiate the impact of diverse managed care plans on various dimensions of performance. It is less important for understanding how managed care is reforming the health care sector. What matters, in this regard, are indicators of health sector transformation and familiarity with some central ideas behind managed care.
       It is presumptuous to answer this question dispassionately. it provokes a host of value judgments and ideological predispositions about the proper role of the State in the social organization of medical care (6). For this reason, in the present essay I proceed rather indirectly so as to enable the reader to arrive at an independent judgment. I begin by sketching the broad features of the French health care system and highlighting its virtues. Then I attempt to fill in this image by reviewing the evolution of French health policy. Finally, I analyse the problem of rising health care costs and discuss some unresolved issues of regulatory policy and management based on my experience in working with (he Director of the principal NHI Fund.
Indicators of Health Sector Transformation
The most striking indicators of transformation in the health sector highlight three trends: the massive shift in the delivery of health services from hospitals to the community; the increasing consolidation - both horizontal and vertical - of health care providers and payers; and the rise of managed care. Evidence indicates that hospitals are losing their place at the apex of the health care system. Over the decade 1983-1993, the number of beds in community hospitals declined by ten per cent (AHA, 1994a). Between 1980 and 1993, 949 hospitals closed.2 The number of community hospitals declined by 6.1 per cent; in rural areas by 20 per cent. 3 During the same period, hospital admissions declined by 14 per cent and the total number of hospital-bed days declined by 19 per cent (PROPAC, 1994: 20). Although the number of surgical operations per person continues to increase and new technologies continue to diffuse from university medical centers to community hospitals, the per cent of surgical procedures performed on an inpatient basis within all hospitals declined from 83.7 to 46.2 between 1980 and 1992 (PROPAC, 1994: 15, 20).
       The counterpoint to the decline of inpatient bed capacity is the growth of outpatient health services. The per cent of hospital surgical procedures performed on an outpatient basis increased from 16.3 to 53.8 (PROPAC, 1994: 20); and this is an underestimate for it excludes surgery performed outside of hospitals in ambulatory surgical facilities. If one includes such surgery in the equation, then outpatient surgery accounts for approximately 70 per cent of total surgical procedures (Shortell et al., 1995). Outpatient visits to hospitals have soared over the past decade but estimates indicate that 98 per cent of medical encounters occur in nonhospital settings.4 During the same period, home care visits for Medicare enrolees grew from 26 to 88 per 1000 enrollees between 1980 and 1994 (PROPAC, 1995a: 21).
       The increasing consolidation of hospitals has turned the independent community hospital into an organizational dinosaur. In 1980, 72 per cent of hospitals were free-standing, independent institutions. In 1992, 43 per cent of all hospitals were horizontally integrated as part of multihospital systems; since then, the share has exceeded one-half.5 Most of the remaining independent hospitals have joined one form of alliance or another. For example, between 1985 and 1992, the share of hospitals having established contracts with HMOs or PPOs increased from 37 to 62 per cent.6 The largest horizontally integrated hospital system, Columbia/HCA, manages 311 hospitals - one-half of the beds in the investor-owned sector (Eckholm, 1994).
       Consolidation extends to the physician community as well, with the doctor in solo private practice becoming a kind of anachronism - charming, like abandoned villages of the middle ages, but no longer critical either to the technical capabilities of medicine or to the broader health care economy. As recently as 1987, over half of all physicians were self-employed in solo or two-physician practice; by 1993, this figure had dropped to 37 per cent.7 The remaining physicians are split roughly between two practice settings: group practice or other patient care such as hospital-based settings.8 But even among the 37 per cent in solo practice, many are affiliated with PPOs and IPAs. Even without formal affiliations, an increasing number of doctors are linked to indemnity insurers and Medicare by electronic communication systems. In 1998 the Medicare Transaction System will require all physicians to file Medicare claims electronically and accept electronic funds transfer payments (Borzo, 1995). Roughly 11 per cent of physicians work in multi-speciality groups.9 Roughly nine per cent have contracts with management services organizations (MSOs) that provide them with a variety of administrative support services such as billing, group purchasing and other aspects of office administration.10 Other solo practices and groups will increasingly be acquired by such enterprises as Mullikin Medical Enterprises, owned by 200 California physicians, or directly by hospitals and health insurance plans.
       Beyond such signs of consolidation, the most significant indicator of health sector transformation has been the growth of organized delivery systems through contractual relationships or direct vertical integration (Shortell et al., 1994).11 Such systems - MCOs - have been largely promoted by the health insurance but also by providers themselves in the form of integrated delivery systems (IDSs), physician hospital organizations (PHOs), and a whole set of new organizations that seek to improve coordination among networks of physicians, hospitals and other purveyors of health care services. In a recent survey of 1200 hopitals by Deloitte and Touche, 71 per cent said they were joining an integrated system that included other hospitals, outpatient units or physician group practices (Freudenheim, 1994). The more vertically integrated systems - largely HMOs of which 69 per cent are now for profit - are themselves undergoing industry consolidation as mergers and acquisitions are rapidly turning the industry into an oligopoly in which the ten largest HMOs enrol roughly 60 per cent of the population in HMOs.12
       In 1994, 17 per cent of Americans were enroled in HMOs, and the rate of growth of enrolments since 1993 was 10 per cent (Hammer and Schwartz, 1994). As a proportion of privately insured employees in firms with over 200 employees, the 17 per cent figure increases to 25; and if the definition of managed care is enlarged to include PPOs and POS plans, another 25 and 15 per cent, respectively, of this population may be classified as under managed care (KMPG, 1994). Thus, without even including 'managed indemnity plans,' 65 per cent of insured employees in large firms were already enroled in managed care in 1994. As for physician participation in managed care plans, 75 per cent have contracts with an IPA, PPO or HMO. This figure ranges, by specialty, from 50 per cent for psychiatry to 80 per cent and above for obstetrics-gynecology, internal medicine, pediatrics and radiology (Emmons and Simon, 1993).
Ideas Behind Managed Care
Beyond the quest to evaluate the impact of the latest MCOs on various indicators of access, cost, quality, performance and outcome, what is often overlooked are the intellectual foundations of managed care. Ideas - even metaphors - matter in framing the debate on health care reform (Annas, 1995). The language of managed care is strangely reminiscent of the health planning literature of the 1960s and l970s (Rodwin, 1984). Health planners then advocated population-based analysis and focused on issues of access as well as need; the new managers of MCOs emphasize variations in medical practice and appropriateness of care. Health planners sought to rationalize the health system by eliminating duplication of facilities and capital expenditures and introducing accountability through consumer representation (Marmor and Morone, l981). The new managers of MCOs seek economies of scale in the provision of expensive high-tech services through selective contracting and organizational accountability through the design of new information systems and mechanisms for total quality management.
       When the trend towards managed care is evaluated in historical perspective, four ideas are likely to be remembered. First, the empirical studies of Wennberg and Gittelsohn (1973, 1982) on small-area variations in medical practice drew attention to the considerable uncertainty surrounding the efficacy of much standard medical practice. Also, these studies revealed the importance of supply-side organization, including specialty distribution, in determining the use of medical care. Among the most important effects of this work is the attention now given to outcomes research and its dissemination as well as to the preferences of patients in medical decision-making.
       Second, studies of the RAND Corporation on the appropriateness of various medical procedures provided methods for evaluating variations in medical practice and assessing the quality of medical care.13 One significant finding from this work is that differences in appropriateness do not explain geographic variations for coronary angiography, carotid endarterectomy and endoscopies (Chassin et al., 1987). Most important, however, at least for the new managed care professionals, is the rationale provided by RAND researches for identifying and trying to eliminate "inappropriate" medical procedures.14
       The third idea behind managed care is the concept of economies of scale in the provision of medical care, and their relation to indicators of quality This idea has received renewed attention and provided an important rationale for 'disease management,' case management (particularly for the elderly and chronically ill) and selective contracting with specialty institutions delivering high-tech medical care. Studies in this area range widely by topic. Some important examples related to heart surgery have established that average costs as well as mortality decrease with higher volume (Finkler, 1981; Luft et al., 1979; Showstack et al., 1987). Such studies provide the foundation for 'medical value purchasing' (Couch, 1991).
       Fourth, studies on the applicability of total quality management (TQM) to health care organizations has resulted in operational changes, not just in fashionable and lucrative consulting assignments for MCOs. Whether these changes involve 're-engineering,' 'work redesign,' 'continuous quality improvement' or 'patient focused care,' the important idea is the contribution of industrial engineering to health care organizations (Berwick, 1989; Merry, 1990; Lathrop, 1993). The relevant metaphor that permeates this literature is the comparison of the health sector to an industry. This comparison was made years ago by Dr Arnold Relman (1980) who was worried by the trend. But the current literature on TQM suggests that much good may also come of it (Laffel and Blumenthal, 1989).
CONTROVERSIES PROVOKED BY THE RISE OF MANAGED CARE
The rise of managed care, in the Unites States, has provoked much controversy and unresolved debate around at least three topics: the corporatization of medicine; physicians' conflicts of interest and quality of care; and 'any-willing provider' laws.
The Corporatization of Medicine
The investor-owned sector is rapidly becoming dominant throughout the entire managed care industry. Moreover, the growth in the size of MCOs, through horizontal and vertical integration, has resulted in industry concentration resembling oligopoly in both the investor-owned as well as the non-profit sector. This means that competition is likely to occur less on price than on quality, innovation, marketing and image, all of which are hard to assess. As industry concentration increases, employers and consumers may well find themselves at the mercy of organizations which are more accountable to their stockholders than to the consumers they serve.
       One example of this problem is the case of the Xerox Corporation that has contracts with about 200 HMOs. The manager of health care strategy and programs fears that the prices paid in HMO acquisitions will lead stockholders to demand big returns on investment, which will raise premium costs as the managed care industry becomes less competitive (Anders and Winslow, 1995). The response to this concern points to the formation of national employer purchasing coalitions to buy managed health care benefits from national HMOs around the country. Thus, more than 100 HMOs recently bid for the chance to offer standardized health coverage to 240,000 employees of ten big companies (Freudenheim, 1995c).
       Another controversy related to corporatization concerns the growth of administrative costs and profit. Between 1970-89 health care administrators increased almost sixfold while the number of physicians only doubled.15 An industry report by the Sherlock Company indicates that the percentage of premium revenue allocated to administrative costs and profit ranges from 16.5 to 27.1 per cent (Freudenheim 1995b). Cash compensation for chief executives in the investor-owned HMO industry averaged $255,000 and some of the cash compensation and stock options for the chiefs of highly profitable HMOs ranged from $2.8 million to $15.5 million a year (Freudenheim, 1995a). The extent to which health care revenues are actually shifting from clinicians to managers is not clear, but there is no doubt that conflict rages between the advocates of professionalism versus commercialism.
       This conflict is not just about money. It is also about control. Dr Arnold Relman, one of the most respected defenders of professionalism, claims: 'There's never been a time in the history of American medicine when the independence and autonomy of medical practitioners was as uncertain as it is now. . . business men and their agents will begin to exercise unprecedented control over the allocation of medical resources' (Eckholm, 1994). Donald Light (1994:1198) tells British readers that 'executives and investors (in managed care) set policies behind closed doors that affect the nature and extent of services available to patients and their primary care doctors. One might call this MBA-managed care rather than clinically managed care.'
       Another view is that the medical profession will reorganize itself in response to such assaults on its professional dominance and clinical autonomy (Freidson, 1989). Higher strata of physicians will set the standards and practice guidelines while lower strata will deliver care on the front lines. Woolhandler and Himmelstein (1994) quip: 'Why should doctors and nurses manage care; do chefs run McDonalds?' One response is that resources are tightly constrained and, in the long run, there is no alternative to health care rationing (Schwartz and Mendelson, 1992). But, in the short-run, the critical question is whether it is better to have physicians, policy 'wonks' or some combination of the two as medical directors to make key management decisions about what is appropriate care (Welch, 1991).
Physicians' Conflicts of Interest and Quality of Care
The issue of physicians as medical directors highlights the controversy around physicians' conflicts of interest and quality of care. Since managed care organizations have multiple and conflicting goals - for example, reducing expenditures, increasing efficiency, eliminating inappropriate use, improving the quality of care and patient satisfaction - important trade-offs must be made (Rodwin, 1995). In their role as agent for their patients, physicians generally face at least two kinds of conflicts of interest: those stemming from financial and other personal interests and those stemming from divided loyalties (Rodwin, 1993).
       Medical directors in managed care organizations typically build consensus among doctors about treatment protocols appropriate for an enroled population and assure them that exceptions are made only with reasonable justification. They do not normally act as agents on behalf of individual patients. But, if their income depends on how well the organization is doing, they face financial conflicts of interest stemming from divided loyalties between their patient populations and their boards of directors or stockholders. In the case of investor-owned corporations, these conflicts exacerbate the clash between professionalism and commercialism. Depending on the method of compensation, primary care physicians acting as gatekeepers in MCOs may also face a variety of financial conflicts of interest. Even without direct financial incentives, the ways in which MCOs structure the delivery of health care, control resources available to physicians, and implement practice guidelines may place doctors in the uncomfortable position of informing their patients about medical options excluded by the health plan at the cost of jeopardizing loyalty to their organization.
       These problems, of course, are not unique to managed care organizations. They affect doctors practicing fee-for-service medicine on patients covered by indemnity insurance and they are present in national health service systems. What has changed with the rise of managed care is that financial as well as organizational incentives now favor underprovision rather than overprovision of services. This shift can threaten the doctor-patient relationship (Emanuel and Dubler, 1995) and it can affect the quality of care, or at least perceptions of the quality of care. There is already a backlash against 'supermarket' medicine and 24-hour hospital stays for normal deliveries (Chernaik, 1994; Nordheimer, 1995). Arguments are now made that the primary care gatekeeper is essential in protecting patients from 'overtreatment' and in coordinating complex medical servies (Franks et al, 1992). Also, 'report cards' on MCOs are increasingly in vogue to assist employers and consumers in evaluating quality (NCQA, 1993).
       The American Medical Association (AMA, 1990, 1995) considers these problems sufficiently serious to warrant special reports on financial incentives to limit care as well as on ethical issues in managed care. Since these reports will surely not make the problems go away, there is likely to be a movement to assure consumer protection in MCOs and to seek legal remedies and public policy responses.
'Any Willing Provider' Laws
The controversies around corporatization and physicians' conflicts of interest and quality of care are unlikely to threaten the rise of managed care. Quite the contrary, they are likely to stimulate industry improvements and public policy responses that protect consumers and regulate MCOs. In contrast, the controversy provoked by 'any willing provider' laws could stunt the growth of MCOs or, at the very least, change the conditions under which they now proliferate.
       'Any willing provider' laws represent a backlash by providers and consumers who miss the 'good old days' when payers stayed out of the management of health care and restricted their actions to timely reimbursement functions. Since providers now worry about being excluded from MCOs and consumers fear that their freedom to choose could be restricted or subject to financial penalties, there is a movement among the US states to pass legislation that would require MCOs to contract with 'any willing provider' who accepts the terms of the organization (PPRC, l995:Appendix D).Variations of this legislation emphasize due process laws to preclude MCOs from excluding or 'de-selecting' providers without explicit criteria or proper grievance procedures. Some variations emphasize 'essential community provider' laws to assure that providers serving the most vulnerable populations are not excluded. Other variations emphasize 'freedom of choice' laws which would allow enrolled consumers to leave their network and receive reimbursement for these services if the provider agrees to accept it.
       At the federal level, the AMA has proposed a Patient Protection Act which calls for federal standards to certify MCOs. MCOs would be required to disclose the criteria they use to select and exclude physicians from their networks while physicians would have the right to appeal their exclusion. Iglehart (l994a) notes that this proposal 'was largely engineered by state medical societies seeking relief for beleaguered fee-for-service physicians.' Much like the state-level legislative activity, this federal initiative represents a strong force in favor of what Weller (1984) and Enthoven (1993) refer to as 'guild free choicer The growth of MCOs has created a constituency that emphasizes choice of MCO rather than choice of doctor and that is one of the most important controversies provoked by the rise of managed care.
LESSONS FOR FRENCH HEALTH POLICY
France has prided itself as a leading international center for the conception and circulation of new ideas. In the field of health policy, however - at least since the mid-1970s - most new ideas have come from abroad (Kimberly and de Pouvourville, 1993). Although they have circulated widely in policy circles, they have rarely reached the top of the health policy agenda (Hurst, 1992). In France, government commissions and national task forces have never devised the kinds of comprehensive proposals for health care reform advanced in Canada, Great Britain, the Netherlands or even the United States. With the possible exception of establishing social security and NHI in 1945 and reforming public hospitals in 1958 under President Dc Gaulle during the first year of the Fifth Republic, health policy has advanced laboriously in small increments.
       The reasons for resistance to comprehensive reform of the health sector in France may reflect the fragmented nature of French medical trade unions, the curious weakness in this sector of an otherwise strong central state, or simply that French NHI is rather well designed (Rodwin, 1992). Such considerations, however, exceed the scope of this chapter. Only two questions guide the following analysis: Is there, despite the system's strengths, a reasonable case for health care reform in France in light of existing problems? And, will the incremental steps taken during the past 15 years push French policy-makers to explore models of managed care?
The Case for Health Care Reform
The French NHI system combines a public-private mix in the provision of medical care with employer-based financing that covers virtually the entire resident population. Patients have unrestricted access to primary care doctors, specialists and hospitals. There are no queues for tertiary hospital services, no 'patient dumping' arising from financial barriers to care, and no public complaints about rationing health care. Moreover, France spends only 9.1 per cent of its gross domestic product (GDP) on health care compared to almost 14 per cent in the United States, and health indicators such as life expectancy at birth and infant mortality place France well ahead of the United States (Rodwin and Sandier, 1993). The French health care system is at the unrestricted, 'guild free choice' end of the managed care continuum. From the points of view of achieving universal coverage and cost control, allowing unlimited choice of doctors by patients, and preserving clinical autonomy and fee-for-service payment as well as a public-private mix in the provision of health services, French NHI is a model to cherish. So, why should French policy-makers pursue health care reforms, let alone seek lessons from American experience?
       When health policy-makers in the United States learn about the French health system, they are generally struck by a model of macro-management: price controls such as a national fee schedule for physician reimbursement; global budgets for hospital operating expenditures; capital controls on new medical technologies and health care manpower, including specialty distribution; and, most recently, expenditure targets for various categories of health expenditure. They are also impressed by the fact that France has the largest commercial hospital sector in Europe as well as balance-billing for so-called 'sector 2' physicians. Almost one-third of French physicians have opted to join 'sector 2' which means that they must pay higher health insurance premiums but, in return, can charge their patients in excess of nationally negotiated fees as long as they do so with 'tact and measure.'
       When health policy-makers in France learn about the American health system, they are impressed by the practice of micro-management. Once they recover from the shock of noting that over 15 per cent of Americans have no health insurance, French policy analysts are intrigued by American administrative technologies: measures of hospital case mix, such as diagnosis-related groups (DRGs); valuations of physician procedures, such as the resource-based relative value scale (RBRVS) for physician reimbursement under medicare; management information systems for hospitals and other health care organizations; an enormous amount of health services research; and innovative forms of medical care organization, such as multispecialty group practices and the full gamut of MCOs.
       Despite the notable attributes of the French health system when viewed from the outside, its policy-makers and analysts within are well aware of serious flaws. Over the past decade, a profusion of government commissions and national task force reports has recognized a range of fundamental problems with the French NHI system.
Five Problems
To begin with, there is a problem of equity that stems from the introduction of balance-billing ('sector 2') in 1979, the increase in copayments (le plan Seguin) in 1986, and the prevailing but erroneous policy assumption that medical practice is largely homogeneous. Lachaud and Rochaix (1993) have shown copayments to be a highly regressive form of health care financing in France. Although 80 per cent of the population have complementary insurance policies to cover these expenses, the 20 per cent who do not benefit from this coverage are the lower income groups who probably need it most (Bocognano et al., 1992). The inequities created by the system of copayments and balance-billing are exacerbated in still other ways. This is documented by studies on practice variations in France as well as by studies of regional inequalities in per capita health expenditure and in differential life expectancy (De Kervasdoué and Polton, 1989; Rodwin et al., 1992). Although these findings are not new, their clarity lead one to re-examine a myth upon which French NHI was originally established: that universal coverage would result in equity of access to health care.
       Second, the French NHI system is suffering a crisis of legitimacy which erupted due to the decline of trade union membership and the ambiguity of decision-making authority between the NHI funds, the state, and health care providers (Lépinay, 1991). The decline of trade union membership has resulted in a situation where the Board of Directors of the principal National Fund for Salaried Workers (CNAMTS)16 is composed of trade union representatives - dominated by Force Ouvrière (FO)17 - who represent less than 10 per cent of French employees. The business community also represented on the Board, has traditionally played a rather passive role (Silow-Carroll and Meyer, 1995). Despite periodic complaints by French employers that payroll taxes are too high and that NHI expenditures not strictly related to the workplace should be covered by the French state, power has been shared between the President of the Board, who represents FO, and the Director of the CNAMTS, who is appointed by the government. While the NHI Funds are claiming the right to intervene more aggressively in the management of the services they finance, the French state is unlikely to relinquish its powers. A report of the National Planning Commission (Soubie et al., 1993) suggested that the state finance the health system through general revenue funds (fiscalisation) and increase its control and accountability for how health services are delivered. Such a reform, however, is unlikely due to the delicate balance of power between the NHI Funds, the state and health care providers.
       The NHI Funds - the CNAMTS as well as the Funds for agricultural workers (MSA)18 and for the self-employed (CANAM)19 - finance roughly 75 per cent of all health expenditures but, for two reasons, they have neither the authority nor the financial incentive to manage their funds judiciously. First, they are closely supervised by the state in their negotiations with representatives of the health professions and private hospitals and are obligated to fund global budgets, set by the Ministry of Health, for public hospitals. Second, they are not really held accountable for balancing their funds because they have no say in setting the payroll tax and because the state is responsible for annual deficits. In France, the Ministry of Health wields extensive control over all public hospitals and strong regulatory authority over the whole health care system. However, it pays a minute fraction (only 1.5 per cent) of all health expenditures. Furthermore, Parliament has no official responsibility for how funds, raised through payroll taxes, are allocated. Finally, French physicians are deeply divided among competing trade-unions. Sanctioned by the state to represent the medical profession in national negotiations, they actually represent less than 20 per cent of all practicing doctors. As clinicians, physicians dispense services and maintain enormous authority; they tolerate no intrusion into their medical practice and bear almost no responsibility for the resources expended by their clinical decisions. The problem of creating some Cartesian order and accountability out of a NHI system characterized by institutional complexity political sensitivity and a crisis of legitimacy has eluded French policy-makers for more than thirty years (Rodwin, 1982).
       The third, perhaps foremost, problem of French NHI is that the state has failed to contain both the level of health care expenditures as a per cent of GDP and the growth of real expenditure increases. France is the highest health care spender in Europe and among OECD countries comes third after the United States and Canada.20 In 1979, per diem rates and estimated days of hospital use were regulated in all public hospitals (two-thirds of all beds). In 1984, global budgets were imposed but this policy has not overcome lax enforcement and shifts in resource allocation to private hospitals and ambulatory care.21 In December 1990, Health Minister Durieux initiated important negotiations, known as la politique contractuelle, with the health professions. He began moving French NHI from an open-ended system (in terms of volume) with stringent price controls to a system with explicitly agreed targets for specific categories of health expenditure - laboratory tests, private-duty nursing, private cliniques, pharmaceuticals and physicians (Rochaix, 1992).
       Fourth, the state's financially driven, 'accounting-style,' budgetary regulation of public hospitals as well as of ambulatory care perpetuates a veil of opacity over what health care services are being provided. This results in a failure to exercise managerial control (micro-management) over the allocation of health care resources. Although US-devised diagnosis-related group (DRGs) measures were adapted for France in 1983, most public hospitals still cannot produce a measure of inpatient case mix and this indicator still has no role in the determination of hospital budgets. Likewise, for ambulatory care, the NHI funds have no patient encounter information. There is no coding system for diagnostic categories and all procedure codes are still aggregated by so-called 'key' letters.
       Although the rationale for such anachronistic practices is to preserve patient anonymity the consequence is that NHI funds have only rudimentary information on how their money is spent. Based on surveys by French medical controllers from CNAMTS, there is mounting evidence of waste and inappropriate medical care utilization. Upon retiring, the national medical director of CNAMTS issued a report (Béraud 1992k), largely inspired by the ideas of Wennberg and Gittelsohn (1973, 1982) and researchers from the RAND Corporation, in which he noted that French physicians perform too many appendectomies and cholecystectomies, and prescribe too many useless drugs and laboratory tests. Moreover, Dr Béraud accused French physicians of engaging in delinquent billing practices.
       The fifth problem concerns the darker side of laissez-faire, the anarchic character of the French health system which leads to difficulties of service coordination and quality assurance (De Kervasdoué 1993). This problem, all the more evident for the elderly and chronically ill, was systematically addressed in a 1981 national task force on health care organization (Gallois and Taib, 1981), which analyzed proposals to create networks of coordinated care or French-style HMOs (Launois et al., 1985; Launois and Giraud, 1985). The problem resurfaced again in a White Paper commissioned by former Prime Minister Balladur, which notes the absence of appropriate information for strengthening public health services, evaluating the quality and outcomes of medical care and coordinating the multiple actors involved in the health care sector more effectively (Soubie et al., 1994).
The Emergence of Managed Care, French-Style
There is no silver bullet to solve the five problems identified above. Some are distinctly French (the crisis of legitimacy and the extreme opacity of the health system); others (inequities, rising costs, poor coordination) are pervasive in all OECD health systems. Like Enthoven's (1990) view of what Europeans can learn from the United States - 'glasnost' and 'perestroika' - French policy analysts have been interested in both as reflected in the rise of managed care in the United States. Glasnost points to the importance of improving management information on costs, inputs, outputs, outcomes and evaluating the costs and benefits of medical care. Perestroika points to the importance of restructuring decision-making processes for the health sector and improving coordination between primary health care and more specialized hospital services, between medical and social services and between public health programs and medical care.
       One striking contrast between the health systems in the United States and France is the difference in the number of employees, hence expenditures, devoted to administrative matters. In the United States, the strongest critics of administrative excess estimate that 20 per cent of total health expenditures are 'wasted' on administration (Himmelstein and Woolhandler, 1987). In France, the health system is considered to be 'under-administered.'22 'Although large numbers of administrative personnel in the United States are due to the eligibility and reimbursement requirements of multiple payers, to matters of risk management arising from malpractice claims, and to quality assurance mechanisms required by multiple and overlapping review agencies, a certain share can also be attributed to management and medical information required by MCOs.23 In this respect, France can learn something from the United States.
       In 1992 a new principle - 'medical utility' - was introduced in the national agreements on fees and expenditure targets negotiated between the state, the NHI Funds and representatives of the medical profession. Henceforth, with regard to specific conditions, the NHI funds will reimburse only those procedures that conform to national medical guidelines (RMOs).24 If physicians fail to follow the guidelines, they will face financial penalties. This accord was considered so important that a law was passed in January 1993 that gave it unprecedented statutory authority. Moreover, the accord calls for the development of necessary tools to assure its implementation: a uniform system for coding all medical procedures and diagnoses. The new accord - what the French call a 'convention' - represents a decisive shift in health policy away from the crude, financially driven, budgetary regulation toward what has come to be. called la régulation médicalisée which is, in effect, French-style managed care (Rodwin et al., 1992). Of course, neither national agreements nor laws are sufficient to effect institutional and behavioral change. Only a fraction of medical practice can be formulated in the language of national guidelines which, to date, number less than a hundred. Also, although a ministerial decree has called for the elaboration of a system for coding all medical procedures, it is not yet operational in France. What is more, there is not yet a decree that requires a coding system for diagnoses.
       In 1994, physician expenditures in France increased by only 1.9 per cent in comparison to 6 per cent the year before. Those who brokered the 1992 national agreements attribute this change to the combination of national medical guidelines and expenditure targets but, of course, it may be no more than a short-run response to the sentinel effect. But in either case, it is significant that the NHI Funds and the state are increasingly intervening not merely over the price of medical services but also on their volume, and not merely on financial criteria but, for the first time, on the basis of medical criteria as well. Under President and Prime Minister Juppé, the announced plan is to lower the pressure on doctors and begin examining the performance of public hospitals more closely.
CONCLUDING OBSERVATIONS
Given the rapid diffusion of administrative technologies, it is hard to believe that French NHI can resist the lure of new medical and management information systems. In this sense, French policy makers and NHI executives could learnm much about glasnost in the United States by studying the use of these systems in well-run MCOs. A more difficult question is whether the adoption of such technologies will result in 'perestroika' or changes of organizational form? For example, if the collection of medical and administrative information yields information on low cost and high quality physicians in France, will the American experience of relying on gatekeepers in certain HMOs or of contracting selectively with PPOs, IPAs, or individual physicians in private practice, provide lessons for French health policy? The President of CNAMTS has been trying to negotiate the creation of a PPO - a national 'optional sector' for French physicians - which would allow them to receive higher reimbursement rates in return for pursuing more cost-effective medicine. The national trade-union for general practitioners has proposed upgrading the rolse of GPs by implementing reimbursement incentives that encourage patients to sign up with GPs who would serve as gatekeepers to the specialists. Both proposals represent forms of perestroika, and both make the controversies provoked by the rise of managed care in the United States pale by comparison.
       For the time being, French-style managed care is no more than the intended application of new administrative technologies to a conventionally organized health care system under NHI. As in the United States, the role of hospitals is changing and ambulatory care is growing, albeit not nearly as fast. Also many for-profit cliniques have merged or joined larger chains, but, again, on a smaller scale than in the United States and without vertical integration to other health sector providers. For better and for worse, the rise of managed care in the United States is likely to be a source of useful lessons for French health care policy. Whether MCOs will ever take root, however, even after they are adapted to French soil, is far from certain.
Notes
1. The author thanks Tony Kovener for his rapid and incisive reactions to an earlier draft of this chapter.

2. Cited by Shortell, Gillies and Devers (1995), this figure is from AHA (1994b).

3. The time period for these figures is 1980-1992 (PROPAC 1994:17).

4. Cited by Shortell, Gillies, and Devers (1995), these data are from Coddington, Moore and Fischer (1994).

5. Cited by PROPAC (1994:21), these data are from the AHA.

6. Cited by PROPAC (1994:22), these data are from the AHA Annual Survey of Hospitals.

7. Cited by PPRC (1995:243), these data are from AMA (1993a).

8. Cited in PPRC (1995:246), these data are from AMA (1993b).

9. Ibid.

10. Ibid.

11. In creating organized delivery systems, hospitals must decide whether to contract out or integrate vertically. For an economic analysis of this issue in terms of transaction cost theory, see Robinson (1994).

12. This figure is calculated from data on the ten largest HMOs and updated by United Health Care's recent acquisition of Metrahealth for which PPO and POS plans are included (Anders and Winslow, 1995; Quint, 1995).

13. This approach has resulted in a large body of work including Brook and Lohr (1985) and Kosecoff and Chassin (1987).

14. Blustein and Marmor (1992) provide an excellent critique of this approach.

15. Cited by Himmelstein and Woolhandler (1992:63), these data are from the Statistical Abstract of the US and the National Center for Health Statistics. For more data on administrative waste in American health care, see ibid.:80-101.

16. Caisse Nationale d'Assurance Maladie des Travailleurs Salariés.

17. 'Working Force' is the largest public sector and health sector trade union in France.

18. Mutualité Sociale Agricole, National Health Insurance Fund for Farmers and Agricultural Workers.

19. Caisse Autonome Nationale d'Assurance Maladie, the National Autonomous Health Insurance Fund.

20. A good indicator of a country's ability to contain health care costs is the elasticity of real per capita health expenditure (health-price-deflated per capita health spending relative to deflator-adjusted per capita GDP). Among OECD countries for the period 1975-1987, the value of this elasticity coefficient is the highest in France at 3.1 compared to 1.1 in the US and 0.9 in Canada (Schieber and Poullier, 1990).

21. The Commission on Social Protection of the Tenth Plan (de Kervasdoué, 1989:ch.3) notes that the share of public hospitals, as a per cent of health care expenditures, declined from 41.3 to 36.6 between 1980 and 1988. On the shift of resources from hospital to ambulatory care, see Pouvvourville (1988).

22. See Berthod (1987), a special issue of Revue d'Aministration Publique 43 (July-September 1987) entitled 'La Santé Est-Elle Sous-Administrée?' (Is Health Under-Administered?).

23. For a comparison of administrative personnel in French and American hospitals, see Rodwin with Cohen et al. (1992).

24. Réferences Médicales Opposables.