UBC Faculty Research and Publications

What seems to be the problem? : the international movement to restructure health care systems Evans, Robert G. Nov 30, 1992

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-Evans_Robert_Restructure_health_care_systems.pdf [ 1.53MB ]
JSON: 52383-1.0048258.json
JSON-LD: 52383-1.0048258-ld.json
RDF/XML (Pretty): 52383-1.0048258-rdf.xml
RDF/JSON: 52383-1.0048258-rdf.json
Turtle: 52383-1.0048258-turtle.txt
N-Triples: 52383-1.0048258-rdf-ntriples.txt
Original Record: 52383-1.0048258-source.json
Full Text

Full Text

WHA T SEEMS TO BE THE PROBLEM?THE INTERNATIONAL MOVEMENTTO RESTRUCTUREHEALTH CARE SYSTEMSR.G. EvansHPRU92:8D November 1992"What Seems to Be the Problem?"The International Movement to Restructure Health Care SystemsRobert G. Evans*Department of Economics, andCentre for Health Services and Policy ResearchUniversity of British ColumbiaNovember, 1992Paper presented to the summer meeting of Spanish health economists, "Econornia y Salud delSystema Sanitario Espanol'', Universidad Internacional Menendez Pelayo, Santander. Spain,August 3l-September 5, 1992.* This work has been supported through a National Health Scientist award from Health andWelfare Canada, and a Fellowship from the Canadian Institute for Advanced ResearchThe Centre for HealthServices and PolicyResearch was established by the Board ofGovernors of the University of British Columbia in December 1990. It was officiallyopened in July 1991. The Centre's primary objective is to co-ordinate, facilitate, andundertake multidisciplinary research in the areasof health policy, health services research,populationhealth, and health humanresources. It bringstogether researchers in a varietyof disciplines who are committed to a multidisciplinary approach to research, and topromoting widedissemination and discussion of research results, in these areas. TheCentre aims to contribute to the improvement of population health by beingresponsive tothe research needs of those responsible for health policy. To this end, it providesaresearch resource for graduate students; developsand facilitates access to health andhealth care databases; sponsors seminars, workshops, conferences and policyconsultations; and distributes Discussion Papers, Research Reports and publicationreprints resulting from the research programs of Centrefaculty.The Centre's Health PolicyResearch Unit Discussion Paperseries provides a vehicle forthe circulation of (pre-publication) work of Centre faculty, staff and associates. It isintended to promotediscussion and to elicit comments and suggestions that might beincorporated within revised versions of thesepapers. The analyses and interpretations, andany errors in the papers, are those of the listed authors. The Centre does not review oredit the papers before they are released.A completelist of available Health Policy Research UnitDiscussion Papers and Reprints,along with an address to whichrequests for copies shouldbe sent, appearsat the back ofeach paper.An 'Epidemic' of ConcernWhere are peace and tranquillity to be found? Public expressions of dissatisfaction withhealth care systems are growing throughout the developed world: everywhere there isunhappiness. Calls for change, from many different quarters, are increasingly being followed byproposals for and programs of reform. The pattern seems virtually universal -- and quite recent.The flood of official enquiries, reports, and proposals is largely concentrated within the last fiveyears, beginning in the mid- to late 1980s.'Of course health care systems have always been a source of contention, in everycountry.' Providers must negotiate their claims for resources with the rest of society, and theytend to greet limitations on those claims with "orchestrated outrage" (Evans et aI., 1989a). Criesof "crisis" are a standard attention-getting mechanism by those whose incomes, careers, andprofessional satisfaction depend upon the continuous expansion of health care expenditures(Evans, 1990a, 1991).Yet for roughly a quarter century, at least, there has been a general consensus in mostcountries that the broad outlines of system organization and funding were sound. Conflicts havebeen over the size of the annual increase, and the priorities for expansion. The current tone isdifferent. Present debates go much deeper, calling into question long-standing structural featuresof national systems. And the proposals for change are more fundamental."The "present discontents" in health care organization thus present us with the interestingphenomenon of a break in the historical pattern of discussion, with proposals for more or lessradical reform occurring roughly simultaneously in many countries in the developed world. Theapparent universality and the timing are inconsistent with explanations in terms of the "failure" ofany particular mode of organizing and funding health care.The same questions are being raised in countries with very different systems, from"socialist" Sweden and the U.K. through Canada, the Netherlands, Germany, and France. Evenin the U.S., always the "odd man out" in the developed world in matters of health care finance(Abel-Smith, 1985), there has been since 1988 a marked increase in the level of expresseddistress and in the attention received by proposals for major reform.'2"Costs": Exploding Rhetoric, Controlled RealityMuch concern focuses on the costs of health care. But there is no correlation betweenofficial unhappiness and the actual level of expenditures on health care. The rhetoric of"uncontrollable costs" emerges in high-cost Sweden and Canada, low-cost New Zealand and theU.K., and most countries in between. More importantly, while the relentless escalation of healthcosts is "viewed with alarm" in every country, such costs have in fact escalated much less rapidlyduring the 1980s than in any of the three previous decades."International cornparative data cornpiled by the OECD (Schieber et aI., 1991, 1992) showvery clearly the general rnoderation of cost escalation during the 1980s. Figure 1 displays thevariation arnong the OECD countries in share of GDP spent on health care in 1970, 1980, and1990. Figure 2 shows the average of these shares, annually frorn 1960 to 1990, and includescornparable data for two outliers: the U.S. where no control on costs has ever been established,and the U.K. which has always been tightly controlled.It is apparent that something quite irnportant happens in the middle of the period. Thetrend in the OECD average parallels that of the U.S. until the rnid-1970s, with the U.K. diverging.But frorn then on the OECD average switches over and parallels that of the U.K., while the U.S.becornes the divergent case.[Figures 1 and 2 about here]Abel-Smith (1992) concludes "The rnain rnessage from the experience of the EuropeanComrnunity is that it is technically possible to control health care costs by governrnent regulationof supply [ernphasis added]" ...."in Europe, regulation works." It also works in Canada (Evans et.91., 1991). This observation is worth ernphasizing, because the cornparison of rnany countriesat a single point in time could lead one to the irnpression that national expenditure levels werelargely dictated by income levels and beyond the reach of public policy. Although an "incomeeffect" based on individual consumer choices is both theoretically and institutionally rneaninglessin this context, the phrase could be used both to suggest a rnechanical relationship, and to shiftattention from public to private decisions.FIGURE 1.I-IEALTH CARE SPENDING OVER GDPGECD Countries0.14·..,--------- .-0.12 .... (), 1 .0.08 .0.06 , . .. .. '. .. .. .. .. . . ': , ..,0.04 .. .. ..•.Ol-f---AL BE DE FR GR IR JA NE NO SP SZ USAU CA FI GE IC IT LX NZ PO SW UK MEANSOURCE: OECD DATA FILE, Provided by Getzen From Pou11ier.'---_19_70 _ 1980~ 1900JFIGURE 2.HEALTH CARE SPENDING OVER GDPAverage of GECD Countries, 1960-19900.13 --._--::I-0.1 2 ~¥'./ ./-¥0.11 :~.+--+: ..y'+-~o~~>~(-L--,+-+-+,y._"T0.08 :.*:": .+:.._+--+. "_ III II III... 11II .. III III"~ -~O. 07 .-- ' ' ;~:.+.;.¥ ·")ii~:.tI·.~.".' ' .~:~: ~1~1L:=~.;·~::*::?-"';::~::,,;~··*~~%~c"'::~..- .. -it: *.~*.= '"0.04 --~.•..*.*..:.~ :.: .0.03 ,. 1 1 1 1 1 1 1 1 1 1 1 1 1 ,-'-1"""-1·'1--.1.......,1'-'-1"""-1---'IC--,-----r--r--.....,--,--,---,-'60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90SOURCE: OECD DATA FILE, Provided by Getzen From Poullier.._--_._~-:,*,- U.K.u.s.I3Getzen and Poullier (1991) have shown that expenditure levels in the OECD countries canin fact be predicted remarkably well over time, by a simple statistical reiationship to nationalincome and general price levels. But in their framework, public policies are the channel throughwhich aggregate incomes influence health spending. Most of the countries of the OECD havetried, under the pressure of deteriorating economic conditions, to limit the growth of healthexpenditures. Most, not all, have succeeded. In several countries a break from past expendituretrends coincides in time with the introduction of specific control policies."An Interpretation: The Health Care System as NeoplasmSo if the problem of cost containment has been "solved" everywhere outside the UnitedStates, why is everyone making such a fuss about costs? The explanation offered in this paper,for both the general unhappiness and the particular focus on costs, starts from Wildavsky's (1977)observation: "The Law of Medical Money states that medical costs rise to equal the sum of allprivate insurance and government subsidy"(p. 109). Health care systems, as presently structuredin every modern industrialized society, have no internal stabilizing mechanisms, or limits to theirgrowth. Professional and technological imperatives common to all systems, and independent ofthe arrangements for organization and funding in anyone country, create a continuous pressurefor more resources, regardless of the ievel currently provided.All health care systems take as their mandate, "meeting needs", yet "needs" are defined(and re-defined) in such a way that they can never be met at any finite level of expenditure. AsWennberg (1990) points out, the energy, innovativeness, and "desire to help" of providers of carecreates a continuing supply of new and potentially useful interventions, each of which representsa new "need". More effective processes to evaluate the effectiveness of these interventions mayweed out many whose potential is overstated or illusory, but will never catch up with the newflow.'Thus health care sectors expand their share of social resources, public or private, untilthey meet with effective external constraints. Such constraints have now been developed andimposed, in most developed countries outside the United States, through direct or indirectgovernment regulation of capacity and budgets. In their absence, the projections in the UnitedStates are for continuing growth, more or less indefinitely (Waldo et aI., 1991).4While technically possible and proven, as Abel-Smith points out, the maintenance of thesecontrols is however becoming increasingly costly in political terms. The general decline in ratesof economic growth in the 1980s (figure 3) has resulted in the progressive tightening ofconstraints on health care sectors. This growing confrontation between the internal urge to grow,and the ever more constricting external environment, is a source of increasing political friction andheat. Those responsible for regulating and paying for health care are responding with increasingefforts to restructure their systems so as to moderate the internal growth pressure, to build insome internal governors. Additionally, they may seek to broaden the coalition of thoseresponsible for containment, or simply to "dump" the problem onto someone else.8[Figure 3 about here]The universality and simultaneity of recent efforts to "reform" health care systems is thusinterpreted as triggered by the unusually severe and prolonged deterioration -- also recent andinternational -- in general economic conditlons." In a sense, the immediate problem is not inhealth care systems themselves, but in the rest of the economy. At a deeper level, however,there is a fundamental problem with all modern systems -- the absence of internal mechanismsto limit expansion.This characteristic is submerged when resources are plentiful, but emerges as a politicallydangerous reef when the economic waters are low. If the water is expected to be low for theindefinite future, either something must be done about the reef, or it will pose a continuing threatto all other public policies (and their proponents). This paper offers a highly schematic historicalaccount of the evolution of health care policy, highlighting the central theme of:"internal expansion-> external constraints-> escalating tension-> proposals for change".The claim that the expansionary drive is internal, an inherent characteristic of all modernhealth care systems, is central to this interpretation. We must therefore consider also acommoniy-offered alternative view, that health care systems are driven to seek expansion byexternal forces in the rest of society, over which they have no control but to which they mustrespond. Dynamic versions of the "patients are sicker; we must meet the needs" argument arenot, we believe, supported by the evidence. All of the "external" forces turn out on closerFIGURE 3.GROWTH IN REAL PER CAPITA INCOrVlEDECO Countries, Average Annual Rates0.07--...------O. 06- ---- --- -------------------------------------------------------------- ----------;- ----------- ----- --------- ------------------------------------- -------­O. 05 -- ----------------- ----------------------------- -- --------- ---------- ---------1---------- --------- ------- ---- ---------------------------------------Source: Schieber et a1. (1992)~__.,,;~;~~.--1966~86-- ---198-0-90~~~~~iL,J~~i~-----------------------5examination to be either not very forcefui, or not extemal. They are largely projections ofprocesses going on within health care systems themselves, expressions of the intemal growthdynamic.The paper will then offer a brief interpretation of general trends in utilization and costswhich are common to several countries, suggesting how these feed into increasing politicalpressures on the "controllers", and will finish by suggesting common themes in the variousnational proposals for change.System Perfonnance • Quite Good, Really:Before addressing what is happening, however, it is important to consider what is nothappening. We argue that the present pressure for reform is not a response to a recentdeterioration in health care systems themselves. But if health care systems were themselves"sick", what might be the symptoms? Normative criteria for a "good" system will inevitably varywith the analyst, but it seems reasonable (to this analyst) to include on any list:(I)(ii)(iii)(iv)Public AcceptabilityEffectivenessEquityEfficiencyIs there evidence that health care systems in the developed world are deteriorating, onsome or all of these criteria?Public Acceptability:Until recently, relatively little has been reported in the academic literature about whatordinary people •• patients, voters, taxpayers .. thought about their health care systems. This isnow changing quite rapidly; much more extensive polling infonnation is an outcome of as well asan input to the drive for reform UL9.:. Blendon et aI., 1990; Blendon and Edwards, 1991).Interestingly, the widespread "official" unhappiness with health care systems among thoseresponsible for regulating and paying for them does not appear to have a counterpart at the publiclevel. In most countries, most people are strongly supportive of or at least reasonably satisfiedwith their systems (Blendon and Taylor, 1989; Blendon et aI., 1990).106The level of satisfaction is by no means uniform. It ranges from very high in Canada andSweden, to very low in the United States and Italy (figure 4). Moreover there is a striking(positive) correlation between the level of per capita expenditure on health care, and the level ofpublic satisfaction .- with one outstanding exception. The United States, the "odd man out" again,has by far the highest level of expenditure, and is at the bottom (among those surveyed) in publicsatisfaction.[Figure 4 about here]Moreover there seems to be an important distinction between the level of satisfactionpeople express as patients, with the care they themselves receive, and as citizens, with theoverall system that serves them. Most people, everywhere, express a high level of confidencein and satisfaction from those who care for them. Even in the United States, people seem to bevery happy with the quality of their own care (Taylor and Leitman, 1991). The differences acrosscountries emerge when people are asked about the system as a whole. [In the United States,the direct cost of care to the individual is also a considerable source of dissatisfaction.]Thus the universal pressure for "reform" is not being driven by general publicdissatisfaction with present arrangements. The political pressure on payers and regulatorsreferred to above is not in most countries .- the United States is as usual an exception .. a callfor major structural change, but rather an expression of concern about perceived threats to a"good" system. Proposals for radical reform, as in the U.K.. New Zealand, or the Netherlands,have met considerable public resistance.These observations should serve both to focus our attempts to understand the currentsituation, and to appreciate the constraints on the possibilities for change. A satisfied public isclearly one aspect of a "good" system -- whether or not an external observer would regard theirsatisfaction as justified _. but it is also a source of inertia."Yet the universality of official concern and pressure for reform is also a fact, and as suchmust have causes and consequences, whether the various national systems themselves areperforming well or badly. Moreover, many of the concerns expressed about health care systems,at least outside the United States, have to do not with their past performance, but with theirFIGURE 4,SATISFACTION WITH HEALTH CARI=Selected DECO Countries--II SI2200-,------,-------,-----,------- -,----,-.1400 18000.15 ------------------------------------------ ----------------------- ---------------------------------. --. -- --------.---------------------- .... -.lilT0.1 I600 10000.6-·~--·(]) 0.55 ------------------------------------------. _. III_CA --- -- ---. --- --------- ---. ----------- -------s(J) 0.5 ---- ---- -------- ---- -- ----- ---- ----. -- --- -- ------ ---- --- ----- ------- ----.. -- -.. ----- ----- ---- ---- ---- ---- -- ----- -- -- - -- -. -- ---- -- ---.-- --- -.~ -NEo 0.45- ---.------------------ ------------------------------------------------------------- --------------- ------- -- ------. -. -------- -----.--.-S 0.4 9_1;_1II_IIFB________________________~ __ ._. . . .E(])t5 0.35 -- -- - ---- ---- -- ----- -----------IiiAL - ----- --- ------ --- -- - -. ------. -- ----- -- --- -. ----~ -sw(f) 0.3 ---- --- -- ------- ---- --- ----- -ii1JA- ----- --------- ----- -- ---- ---- --- ---. ---- ---------- -- -- ----- ----. ---- ------ ---- -- ------- ---- - --..c -UK 'o 0.25 -. ------------------------------------- --------.--- ---- -------------- .. -------------- -- ---- .. --------- -- -. ----- -----------------------to 0.2 -...sP---------------------------------------------------- -- ---.Q.o.....0..800 1200 1600 2000 2400Per Capita Spending, in $US, 1989SOURCE: (Blendan et al. 1990, 1991)7sustainability in a world which is perceived to have changed. General satisfaction with currentarrangements, however well grounded, is no guide on this question.Effectiveness:A satisfied public is not, however, the only test of whether or not a health care system"works". The complex web of regulation, which in every developed society constrains orsupersedes the free marketplace, is constructed on the presumption that individuals are not ingeneral able to recognize and protect their own interests in transactions involving health care.Consistent with this, one might not wish to rely on public opinion alone to evaluate health caresystems;The most obvious criterion for a "good" health care system is that it should restore,maintain, and promote health. Individuals do not normally seek out and use health care servicesfor their intrinsic pleasures, quite the contrary. They undergo diagnostic and therapeuticinterventions in the anticipation of some (probable) improvement in their health (relative to whatit would otherwise have been). In the same way public debates over health system policy areframed (not always accurately) in terms of expected health benefits. The "orchestrated outrage"of providers consists in large part of, in the British expression, "shroud-waving". "People willsuffer and die, if our proposals/demands are not accepted." Payers respond with allegations ofinefficient and ineffective care, inflated incomes, or simply inability to pay.Of course, the determinants of health go far beyond the activities of the medical caresystem, and the impact of much of health care on health ranges from unproven, through non­existent, to negative. This has been understood since antiquity, and the evidence in support israpidly accurnutatlnq (MacKinlay and MacKinlay, 1977; Roos and Hoes, 1992; Marmor, 1992).But these observations do not negate the obvious power of appropriate care to prevent and curecertain illnesses, or alleviate the effects of others, or the equally obvious fact that peopleindividually seek out care, and collectively support health care systems, in the anticipation ofthese benefits.One can often identify individuals whose lives have been saved, prolonged, or improvedby timely and appropriate medical intervention. But when we look at the health status ofpopulations, there is no similar connection. This may reflect the inadequacy of current measures8of population health status. It is customary to deplore, and then to display, aggregates such aslife expectancy and infant mortality, while conceding that these reflect many features of the socialenvironment which have little or nothing to do with health care.Such aggregates may carry significant information. Comparative studies of eastern andwestern Europe are showing that the slow and steady improvement in life expectancy to whichwe in the west have grown accustomed, is by no means a law of nature. In several easternEuropean countries, life expectancies have been falling in recent decades. But again, thesedivergent trends seem to have much more to do with general social conditions than with theavailability and quality of health care (Hertzman, (1990).Efforts are being made to study international trends in deaths from specific causes forwhich medical intervention is believed to be effective. These indicate that, among the membersof the European Community at least, there is no correlation at all between level of average percapita health spending, and mortality from "remediable" causes (Mackenbach, 1991). On theother hand, American analysts are finding that those members of their population who haveminimal or no insurance coverage, receive fewer services for particular conditions, and havepoorer health. The extent to which this is a causal relation, however, is open to some question.What one can say, is that there are differences in overall health status from one nationalpopulation to another, but that there is no evidence that these are connected with overall levelsof either expenditure, or real resources, devoted to health care (Culyer, 1988; Schieber et aI.,1991).'2 On the other hand, there is some reason to believe that the pattern of distribution ofthese resources is important; in particular the organization of maternal and child health servicescan have a significant impact on health outcomes at the aggregate as well as the individual level(Marmor, 1992).Thus the form of organization of health services may matter, even if the overall level ofexpenditure does not (within the currently observed range), and even if other factors outside thehealth care system also exert powerful influences on population health -- influences which havereceived much less policy attention. Some systems do appear to achieve better results thanothers.9But there is no suggestion in the data that the health care systems of the developedsocieties are in any sense "in crisis", with respect to their ability to maintain their contributions tothe health of their populations. Nor is this surprising, since despite the widespread rhetoric of"cutbacks", most if not all health care systems are still expanding, albeit much more slowly(excepting always the United States) than in earlier decades. The tone of the rhetoric is sharper,in some countries at least, and the allegations of crisis are more frequent and more strident. Butthe underlying patterns and trends in health status provide no basis for the widespread urge toreform. Our health is not threatened.On the other hand, as noted above there is growing evidence both that there are powerfulsocial factors affecting health status that lie outside the health care system (gjl, Townsend et aI.,1988; Dutton and Levine, 1989; Canadian Institute, 1991; Wilkinson, 1992), and that much ofcurrent health care is of questionable effect on health. The broadening appreciation of thisgrowing evidence could contribute to a sense of dissatisfaction with the health servicesthemselves, particularly among those who pay for them.The two phenomena probably are connected, but there is a question of timing. Healthservices researchers have been making these points for nearly three decades. The evidence isgrowing in both weight and quality, but the essential message has not changed in that time. Sowhy would a quarter-century or more of work result in this sudden and simultaneous responsein the late 1980s? A threshold or "critical mass" effect? But policies are often based on quiteweak information; it is not generally the case that some minimal level of secure evidence isrequired. Nor, for that matter, does strong evidence necessarily lead to appropriate policy.An alternative view, which fits the observed pattern better, is that other factors havechanged such that those responsible for health policy and health care payment are suddenlymore interested in the message, and therefore in supporting the research. It is not accumulatinginformation which has led to the pressure for reform, rather the pressure for reform has createdan audience for previously available information -- and more besides (Lomas, 1990).Equity:Most people in the developed world, even including the United States, seem to feel thathealth care should be available to all citizens on the basis of need, regardless of their financial10resources." Ordinary people and political commentators alike take pride in the fact that, in thewords of the Canadian legislation, "all medically necessary care" is available to all "on equalterms and conditions" and without financial barriers. An equitable system, as a normativejudgement, is one in which appropriate care is available at need, and in which the unavoidableburden of illness per se does not attract further financial burdens to the individual who suffers it.Yet as Culyer (1991b) has pointed out, even in those countries which profess moststrongly in both rhetoric and policy their commitment to equity in health care, there is no veryclear consensus as to exactly what this means. Still less effort is made to generate empiricalinformation which would indicate how successfully the objective, however defined, is beingachieved. Even in self-consciously egalitarian Sweden, the establishment of a system of carewhich is (more or less) accessible to all, without financial barriers, is taken by most to meet thecriterion of equity. Questions of the actual distribution of services received, their relation to"medical needs" however defined, and the distribution of health outcomes whether or notcontingent on care, are of much more limited interest.If our public commitments to equity are sincere, then we should be more explicit aboutwhat we mean, in particular so that we could know how closely we are coming to our objectives,and whether correctives are needed. One need not claim -- and should not, because it is not true-- that systems such as those of Canada or Sweden have achieved a perfectly equal matchbetween need and use. All systems are to some degree unequal in access -- and a fortiori inhealth outcomes."The growth in our conceptual and empirical understanding of patterns of health care use,and of health status within populations, is in effect setting higher standards of equity for our healthcare systems to meet. But can this be interpreted as a deterioration in system performance, suchas to motivate the pressures for reform which we now see? Or is there any reason to believe thatour systems are becoming less equitable over time, in any absolute sense?On the old version of the criterion, a negative answer is easy. The predominant form ofhealth system organization remains one of public or quasi-public provision or payment, with noor negligible financial barriers to access. Some "reform" proposals seek to modify this pattern,and for a variety of good or (more often) bad reasons to create unambiguously less equitable11systems. But this emphasizes that the pressures for (real or phoney) reform come fromelsewhere, and threaten the objective of equity as traditionally defined. They do not emanatefrom failures to meet that objective.But as we consider in more detail not just the obvious dimensions of access -- financialand to a lesser extent geographic -. but also the patterns of actual use and their relation to"need", matters become less clear. Explicit financial barriers have the signal advantage that wecan all recognize them. "Needs" do not always share this advantage. And we may find that thereis disagreement over priorities for care, such that some regard equity of access, relative to need,as decreasing over time even as others believe, on the same evidence, that it is improving.We will return to this question below, in the context of the changing patterns of care of theaged and of other medically vulnerable groups. It may be, however, that Culyer (1991b) is in factquite wrong in urging the Swedes (or anyone else) to be more explicit about their values. If weall mean different things (a possibility which he clearly recognizes), greater explicitness may makethe measurement and evaluation of equity impossible, and policy much more difficult. Consensusand a sense of common purpose emerge perhaps only from fuzziness and a certain degree ofambiguity.Efficiency:The fourth criterion which was offered above for a "good" health care system wasefficiency, by which is meant the technical efficiency which is so often taken for granted ineconomic theorlzlnq." Are services provided at minimum cost, or some reasonableapproximation thereto?" Since resources are always scarce, and have opportunity costs, healthservices researchers with an economic background have been particularly concerned to evaluatethe health services on this dimension. Their findings have frequently been critical. In spite ofthis, efficiency considerations have not commonly been central to debates over health policy, orcritiques and defenses of particular health care systems.Instead, both data and debate give pride of place to "health care costs", which are always,in the English language discussions at least, "spiralling", often "out of control". At the same time,according to providers of services, or as the Americans increasingly call them, "vendors", "thesystem is underfunded", even as, according to payers, it is threatening to bankrupt the nation.12In the course of this debate, more information is emerging about health care costs in differentcountries, how they compare, and how they are evolving over time (Schieber et aI., 1991, 1992).But much less is known about the comparative efficiency of different health care systems."Indeed even the factoring of costs into internationally comparable price and quantitycomponents is still at a rather preliminary stage. Such analyses are hampered by the fact that"quantity" data tend to measure intermediate products like pills, tests, visits, and days ofinstitutional care. But the connections between these, and the treatment of particular conditions,let alone health outcomes, is very variable across countries as well as over time. [Technologyis not standardized; and no one knows where the isoquants are. Back there somewhere, to thesouth-west.]Yet from an economic perspective, "costs" per se are a peculiar focus for attention.Rational decision-making requires that one consider the benefits that are expected to flow froman action or choice, and compare them with the value of the alternative opportunities given up.Arguments over "costs" should then be subsumed under one or more of the four heads above:acceptability, effectiveness, equity, and efficiency.One may argue that costs are "too high" because one believes that, at the margin, healthservices are ineffective or harmful, or have such small beneficial effects as not to justify their(opportunity) costs. Conversely, one may hold that "the system is underfunded" becauseinterventions which are powerfully effective in improving health and well-being are being deniedto people because their costs exceed the resources made available to pay for them. Theargument turns both on effectiveness and on opportunity cost, but in practice it is usuallyperceptions of the former which are decisive.Alternatively (additionally) one may argue that costs are "too high" because the peopleproviding care are overpaid relative to the rest of society (again, taking account of theiropportunity cost), and/or because services are being produced inefficiently, at excessive(resource) cost. The former is an equity issue which surfaces less often than it should; equityconsiderations arise between provider/vendors and the rest of society, just as much as amongthe population of actual or potential users/payers. The latter is an issue of technical efficiency,13although to the extent that incomes of providers embody a significant quasi-rent component it hasan equity aspect as well."Distributional Concems -- and Irrelevant AnalysisBut there is in fact some logic in bundling all these diverse issues under the generalheading of "costs". On most of the component questions, the differences of opinion andinterpretation tend to split quite neatly, placing those who pay for services in opposition to thosewho are paid. All expenditure is simultaneously and by definition income for someone; total costsequal total incomes. Those who pay are more likely to view health care as relatively ineffective(at the margin anyway), provided by people who are over-paid, and in ways that areunnecessarily costly; those who are paid tend to disagree. Thus "costs" provide a simple andconvenient polarization for the fundamental conflicts of interest which run through theinterpretation of each of the four criteria offered above."The economic perspective, by contrast, completely suppresses this fundamental divisionby framing the decision process as if there were a single social actor (or simple aggregate ofactors) deciding how much to spend upon what sorts of services. Arrow (1973) pointed this outclearly in one of the original theoretical analyses of the welfare effects of charges to patients:"I ignore distributional effects and assume a single person in the economy. Theinteraction between distribution and insurance needs separate analysis." (p.1)Subsequent economic analyses continue to rely on this assumption, but are rarely, if ever,explicit.This habitual, automatically adopted assumption implies that it is possible, for analyticpurposes, to treat all individuals as identical in all relevant respects. Yet conclusions orrecommendations based on analyses of such a hypothetical "economy" are not merely"approximations", arrived at by more convenient analytic methods. They are completely vacuous-- meaningless -- as guides to policy (or anything else) in a world where it is precisely thedifferences among persons which generate all the problems."If one admits the obvious, that people do differ substantially in their relevant characteristics-- some are vendors of services, others are users and directly or indirectly payers, some are old,some young, some are healthy, others have severechronic diseases -- then it is equally obvious14that there is no such thing as a policy without predictable distributional consequences, oftensubstantial ones. An economist who pretends to a "value-free science", and who claims not tomake, or be able "scientifically" to make, interpersonal comparisons, cannot logically or honestlymake any policy recommendations at all. Any such, in the real world, always have redistributiveimplications.None of the above is new; it is all clearly worked out in the historical literature ontheoretical welfare economics. But it represents a serious dilemma for the economist who seeksa rewarding career in the study of health care, or any other applied field. If you do not have, andcannot have, anything to say, people will not long seek (or pay for) your advice. But if youradvice is based upon expertise in a supposedly value-free "economic science", a sort of socialphysics which excludes the possibility of making interpersonal comparisons, (rather than on, say,a slowly and tediously acquired partial familiarity with actual health care systems), then it doesnot in fact bear on any real-life problems at all.What to do? Redefine the problem! Assume that everyone is, for the purposes at hand,identical. This solves the distributional dilemma, preserving that appearance of interpersonalneutrality and scientific objectivity in one's conclusions which is so important for successfulmarketing. For such professional advantages the total unreality of the assumption, andconsequent irrelevance of any conclusions, is a small price to pay. Accordingly some economists,particularly those with a "free market" ideology, continue to offer analyses and policyrecommendations based on a conceptual apparatus which assumes away all distributional issues.But they must wonder sometimes, why so much of the debate and conflict in the real world, isabout total costs."On the surface, the rhetoric of cost escalation with its counterpoint of underfunding seemsto provide an adequate explanation for the universality of distress in health care systems.Certainly reform proposals draw heavily on these concerns, and refer only to a much lesserdegree to the specific criteria offered above. As noted above, however, the widespread pressurefor reform did not corne during the decades of most rapid escalation, but much later, after severalyears of relative stability. This stability appears to be a straight-forward consequence ofincreasingly aggressive and successful efforts by governments in most countries to place limitson health cost escalation.15Thus the perception of health care systems and their costs as political or collectiveproblems is not mechanically related to their absolute level, their rate of growth, or even theirshare of national income. The "present discontents" arise from a dynamic and historical process,which has been roughly similar in most developed societies.Shifting Assumptions and Policies: A StOry in Three ChaptersIn the period from the end of the second world war, to some time (depending upon thecountry) between the late 1960s and the late 1970s, there was a consensus among providers,payers, and the general public in virtually all the developed societies that expansion of the healthcare sector was a social priority. Various institutional mechanisms, differing from country tocountry, were created or modified to bring this about -- with considerable success. But the rapidrates of general economic growth characteristic of this "Phase I" period meant that growingresources were also available to meet other priorities, public and private.Several shared assumptions about health care systems underlay the consensus in PhaseI. The linkage between health care and health was seen as tight; if improving health was a socialpriority -- and it was -- the obvious policy was to put more resources into health care. Health carewas provided only in response to well-defined needs, which were themselves finite, so that therewould be a natural upper limit to the commitment of resources. And finally, responsibleprofessionals could be relied upon to organize their services in an efficient manner, and to acceptreasonable incomes for themselves. Thus health policy was quite simple: "Give to theprofessionals the resources they need, and they will finish the job."The transition from Phase I to Phase II was marked by growing awareness, amonganalysts and regulators of and payers for health care, that each of these assumptions is unsound.Health care is not the only, or even the most direct, route to health; "needs" for health care canbe expanded more or less indefinitely; professionals are no different from anyone else inregarding a fair income as just a bit more than whatever they now receive; and managing anyenterprise for efficiency is a difficult and demanding job which no one will do if it is not forcedupon them.Policies which essentially left it to providers to define the resource needs of the system,and then assigned other public or private institutions (or individuals) the task of meeting them,16thus lead inevitably to continuing expansion without limit. Moreover two decades of rapid growthhad resulted in health care systems which took up a significant share of national incomes.Continued rapid growth on this larger base placed increasing pressure on other social priorities.Accordingly those responsible for health policy in all countries began increasingly to lookfor ways to contain costs, at least to a roughly constant share of national income. This wouldleave room, in a qrowlnp economy, for the pursuit of other priorities. The rhetoric of costescalation and cost containment dates from the late 1960s, at least in North America. But thisrhetoric reflected a break in the earlier consensus; neither nor the general public appear to haverevised their assumptions from Phase I.Many providers, in all occupations, continue to believe that they should, could, simply begiven the resources they (claim to) "need", and left alone to do the job. The public appear toaccept the legitimacy of payers bargaining over provider incomes, and (more uncertainly) theneed for overall fiscal control. But the belief that all health care simply "meets needs", and thatproviders are best able to define these needs, is still strongly embeddedDuring the second phase, the health care system was restricted to a roughly constantshare of a still growing national income. Payers and providers entered into an on-goingrelationship of negotiation and conflict. But this relationship generally led to mutually acceptablesettlements. If providers' ambitions are not fully met at anyone time, nevertheless there arealways some additional resources made available for system expansion or increase in incomes.And anyone who is denied this year, may hope to be at the head of the queue next year.The transition from Phase I seems to occur at different times in different national systems:at the beginning of the 1970s in Canada, in the middle in Denmark and Germany, at the end inSweden, for example (figure 5). And the subsequent history is also variable; in Denmark andSweden the ratio of health expenditure to GDP falls in the 1980s, in Canada it bounces up duringthe deep recession of the early 1980s and then flattens out again, in other countries it staysrelatively flat. [France, seems to be an exception, though not to the extent of the U.S.][Figure 5 about here]FIGURE 5.I,EALTH CARE SPENDING OVER GDPSelected GECD Countries, 1960-19900.1--.---------------------------,,B'~~~O.O~ ;'~~.:..~~~~~)i.~~~~~~ .~, *0.08 ..· ·.· · m ••• :b~" Ai ;t(. ':OK /i :~.':":::*().()E) ~~~~...................... ~ .O.05 .~"!t~ ..~;~/. ...*'-+---¥0.04 ;*:. .~O. 03~--y-,---,-.----,--,--.--r-,-----,--.-,-T-,--,-,---,----,----,--,--.--r--.-----,--.-..---r-,---y--,-'60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90SOURCE: OECD DATA FILE, Provided by Getzen From Poullier.[ II Canada . I Denmark -,*:--- Germany D· Sweden17A third phase begins roughly with the 1980s, in which the share of national resourcesgoing to health care remains constant but the (real, per capita) total grows much more slowly, ifat all. The common experience of most industrialized countries, during this decade, has been amarked decline in overall economic growth, and in many cases a sharp rise in unemployment andgreatly increased pressure on public budgets. For the first time in living memory, health caresystems are not expanding, or at least not very much. And although they are not being cut backin total, resources for new needs and priorities must be found by reductions somewhere else -­cutbacks.But health care systems have an enormous internal momentum. The training institutionshave been for many years generating an increasing supply of personnel; the pressures forprofessional upgrading and credentialling are constantly adding to training requirements andincome expectations, and to the sheer numbers of different occupations and specializations; andthe research establishments turn out a steady stream of new drugs, devices, and techniqueswhich, contrary to their standard claims, are typically more expensive than the ones they replace,or are largely add-ens.All of these forms of intangible capital -- both human and acquired "know-how" -- are theresult of costly investments, and their owners expect to earn a return commensurate with that costand risk. But these expectations of ever more and better "places at the health care feast" (inReinhardt's graphic analogy), which were fed by thirty years of expansion, now run head-on intopayers' efforts to impose global constraints. Whatever the implications of these constraints forthe health of the populations served, or the efficiency of the systems serving them, theirimplications for the career and income aspirations of both present and would-be future providersof services are unambiguously negative.Cost Control Hurts....Lack of Control Hurts MoreThe result is, understandably, a marked escalation in the level of political conflict, and acorresponding marked increase in interest, at the highest political levels, in other and perhapsless uncomfortable ways of organizing and paying for health care. The triggering factor has been,not the escalation of health care expenditures, but the increasing political costs of maintaining andtightening, in response to deteriorating general economic performance, the controls onexpenditures which have in fact been relatively effective.18The expansionary dynamics of health care, on this interpretation, are not peculiar to anyone country or form of organization. They are rooted in the professional ideologies, technicalcapabilities, and economic incentives which, unlike funding and delivery systems, are commonthrough the developed world." Whatever the level of expenditure, tighter controls, lead to moresevere political tensions, and a greater risk that something will break.But what does that mean: "Something will break"? Logically, there seems to be arelatively limited range of possible outcomes from this situation. First, general grow1h rates couldincrease, as inexplicably as they previously dectlned." The health care sectors of thedeveloped countries could then begin to expand again, within a constant share of nationalincome, and everyone could go back to the policies of the 1970s. This would be nice. Failingsuch an external rescue, however, and assurning that real per capita grow1h rates remain low in(recent) historical terms, the conflict between internal expansion and external constraint willcontinue.That in turn raises two possibilities: either the cost control processes will break down, orthey will not. If they break down, health care costs will begin once again to take up an increasingshare of national incomes, financed either from public or from private budgets. The forrnerimplies some cornbination of increasing taxes, increasing deficits, or reduction in other publicservices -- continuing for the foreseeable future. The latter implies the re-introduction of directcharges to patients, not as mere token payrnents, but as a serious contribution to health carecosts.But "taxes on the sick", however gratifying to those whose fundamentalist religion takesthe peculiar form known as "neo-classical economics", will not actually raise much money, partlybecause sick people tend to be or to become poor, and partly because they and others becornerather distressed by the obvious inequity. Out of pocket charges might finance a few years ofexpansion, but for the longer terrn serious reliance on private funding implies private insurance,with its well-known gross inequities and inefficiencies. It will also require the rnassive publicsubsidies without which private health insurance on any serious scale is impossible -- or at leasthas never been seen:'19In other words, the United States, The key point is that the health care system of thatcountry is not an accident; it is the natural outcome, through forces well documented and wellunderstood, of a grim determination to rely as much as possible on private funding, confrontedby the realities of health care,Thus a break-down of the current mechanisms of cost control, in an environment ofcontinuing economic stagnation, appears to lead either to an inequitable, inefficient, and veryexpensive U,S,-style system, or to further pressure on public budgets which in most countriesare already believed, rightly or wrongly, to be at the limits of political tolerance,This may appear to be a rather extreme or apocalyptic posing of the alternatives,particularly to those (usually providers) who believe that just a little bit more private funding wouldbe a good idea, It would put a little extra money in their own pockets, and could permit theirwealthier patients to buy preferred treatment while still drawing primarily on public funds -- likethe British NHS, only with more resources, [The idea also tends to be attractive to wealthierpatlents.] But the point about any system with "just a little" private financing, is that its stabilityactually rests on the continuation of effective global cost control over the remainder of the healthbudqet, with all the conflict that implies -- just as in the NHS, Private funding survives, in a stableenvironment, only because it is not very important.If private funding is to serve as a lever for the continuous expansion of the health caresystem, in the context of a no-growth economy, it must either continue to grow as a proportionof the total, or else reach a scale at which it undermines the ability of the public or quasi-publicfunding system to contain its own costs, Then both grow in parallel -- as they have been doingin the United States,On this interpretation, the reason why we observe at present only two kinds of systems,many with no or trivial private funding and relatively stable costs, and one with significant privatefunding and unstable costs, is because given the dynamics of health care, those really are theonly alternanves."The consequences of a return to uncontrolled escalation, however financed, explain thevery consistent pattern among the different national efforts to reform health care systems, All20emphasize that what is needed is not more money, but more effective management andaccountability for the resources already being used. And almost all see this as coming, not fromshifting the purchase of care back to the private marketplace, but from re-structuring theorganization of the delivery system. The ideas of Alain Enthoven have had a particularly broadinfluence in Europe, being encapsulated in the notions of the "purchaser-provider split", and"managed competition" among providers.The articulation and application of these ideas, however, is still in a state of flux, and itwould be premature to offer much more than speculation about their effects. Instead we retumto the multi-phase scenario sketched out above, to expand on a couple of related points whichmay provide further clarification for the direction of current reforms. These are the growthdynamic in health care, and the nature of the political pressures motivating reform.Diversionary Tactics: "We're Just Trying to Meet the Needs"It is a central tenet of this analysis that the pressures to expand health care systems arisewithin those systems, or rather within the world-wide professional and technical culture of modemhealth care. The providers of services of all types -- clinical, supportive, technical, research -- donot merely or even primarily respond to external pressures, well-defined needs, but rather areconstantly re-defining and extending the agenda of health care. Both in research and in theirclinical activities they develop new grounds for intervention by seeking out new problems or re­interpreting old ones." This view is now, I think, dorninant arnong those responsible for healthpolicy and payrnent, and the evidence for it is very strong.But there rernain dissenting voices, particularly arnong providers, because it has alwaysbeen a powerful political argurnent, in the endless struggle for "rnore", to clalrn that one is merelyresponding to the obvious and externally determined needs of others. The system is simply thereto serve the patient -- and does so. More resources are needed because needs have grown -­"our patients are sicker, and there are rnore of them" -- and we rnust respond.It is not difficult to forrnulate this as a circular argument, logically impenetrable. If aservice has been provided, it must have been needed, because providers only provide whatpatients need. Therefore, if utilization and costs are rising, needs must be increasing. And since21the purpose and obligation of health care systems is to meet needs, clearly more resourcesshould be provided.Nor can any external observer ever know for certain what the patient's needs were.Every case is different, only the clinician on the spot can have enough information to know whatis to be done. So the claim that utiiization is evidence of need cannot be refuted. [Economistsshould recognize a parallel with the claims of certain economic theories .• the Panglossian trapis the same, and may be similarly rnotivated.]But the scientific tradition is also strong in medicine, and progress in health servicesresearch, particularly in clinical epidemiology, has demolished any empirical basis for this styleof argument. It survives only as a form of faith, or as a public relations tactic. Instead, those whoargue for "more" support their claims by reference to external forces or trends which are allegedto generate new and greater needs. In economic terms the implicit claim is that, for a variety ofreasons, the marginal benefit to be anticipated from further expansion of health care is itselfincreasing, and faster than the opportunity cost.This claim is commonly presented under three heads: the aging of the population; theprogress of technology; and the rise in public expectations. The physical and mental vulnerabilityof the population is increasing, on average "people are sicker"; moreover, whatever the health ofthe population, new interventions make possible greater improvements; and finally, in effect,"tastes are changing" such that people attach a higher value to whatever benefits health care canyield. For all these reasons the health care sector should receive an increasing share ofwhatever national resources are available, through whatever channels are least clogged byexternal constralnts.i'Yet on examination, each of these claims of "external pressures" tends todissolve (Evans, 1985).Aging Populations?Looking first at the very popular "aging" argument, it is readily apparent that internationaldifferences in health care costs bear no systematic relation to population age structure. Theyoungest countries •. Canada and the United States •. have the highest costs, and the "aging"which so concerns them (and Japan) has already occurred in much of Europe.22And when one examines utilization and cost trends in individual countries, they show acommon and unambiguous pattern. Care of the elderly absorbs a growing share of the healthbudget, but this is almost entirely a result of increases in the per capita rates of use (age­adjusted) by elderly people. Increases in their numbers, or their ages, have much smaller effects(Barer et aI., 1987; Evans et aI., 1989b; Hertzman et aI., 1990; Barer et al., 1991; Gerdtham andJonsson, 1991). Any claim that demographic forces have played a major role in the pastescalation of health costs, in aggregate, is simply false?" Projections for the future yield similarresults.One could, of course, hypothesize that the elderly -- at each age -- are on averagebecoming sicker. (Why? Well, they must be, because they are using more care...) On this thereis much less evidence. What there is suggests that today's elderly may indeed be somewhat lesshealthy, on average, than their predecessors (Verbrugge, 1984; Barer et aI., 1987). But there isno evidence that this accounts for most of the increase in utilization and cost per (elderly) capita.Changing patterns of care, not changing numbers and needs, are the drivers (Black et aI., 1992).Growing Public Expectations?A quantitative assessment of the "aging" claim is relatively easy using generally availabledata on population age structu res and age-specific utilization rates. The "technology" and "publicexpectations" claims cannot be addressed as conclusively. Unlike demographic change, however,these are not really forces external to health care systerns. Public expectations, in particular, arebased in large part on information generated by people and institutions within those systems.The extensive and well-funded campaign in the United States to promote cholesterolscreening for the general population is a leading example (Moore, 1989)?9 The marketing ofdrugs and "high-tech" procedures directly to the public, often in subtle and apparently objectiveways, is an expanding activity in North America. The expectation is that patients will then requestthese from doctors who are also subject to a massive barrage of advertising, and whoseeconomic interests, at least, are consistent with compliance.More generally, current arrangements for the delivery of care, in many countries, are suchas to reward those who successfully encourage unrealistic expectations, and offer thecorresponding product or service." And these then become embedded in the international23culture of medical practice. After the fact, as the clinical epidemiologists have so clearly shown,it is difficult enough for the clinician to know whether a particular manoeuvre was effective, letalone for the patient.By contrast there is very little incentive, except for the professionalism of the individualphysician, to discourage such expectations and provide for the patient or the general public arealistic view of what is possible. Some do, but there is little profit in it.From recent research, however, a certain amount of information is beginning to emergeas to what people really do want, and it turns out that when aware of what the options really arethey often want less, not more. Wennberg (1992) and his colleagues have shown that patientsoffered surgery are much more risk averse than clinicians -- not surprisingly _. and when givena more complete and understandable description of risks and benefits, very often choose"watchful waiting" instead. Molloy and Guyatl (1991) have confirmed widespread impressions,that patients at the end of life do not want the "heroic" interventions which are often (thoughprobably inaccurately) blamed for much of the escalation of health costs. Given the choice, mostof them reject cardiac resuscitation. The rapidly growing interest in living wills and the "right todie" supports a similar conclusion.We do not at this point have enough information for a general conclusion, but what thereis, is consistent with the possibility that a serious effort to communicate to patients and to thegeneral public, what health care can and cannot do, its benefits and risks, would deflate the"public expectations" argument rather rapidly. But why should any provider do this?"The economic advantages for providers of "doing more" are obvious; there is an inherentconflict of payer and provider interests, and which is most aligned with patient interests cannotbe determined in general or a priori. It depends upon the circumstances. But the important pointwhich emerges from Wennberg's work is that apart from economic issues, providers and patientshave significantly different attitudes toward risk. And providers, not surprisingly, believe in whatthey do. So they may, in good professional conscience and acting as the patient's agent,recommend as "best for the patient", services which the patient, if better informed, would refuse.He concludes that the agency relationship, which is at the heart of the process of health careutilization, is in this respect at least fundamentally flawed.24The Technological Imperative?On the third claim, the role of technology, it is obvious that new drugs, equipment, anddevices of extraordinary scientific sophistication are becoming available, apparently at anincreasing rate, and that they are expensive. But it is equally obvious that these are not the resultof some external process, outside and independent of the health care system. The research anddevelopment which generates these innovations is inside, and very much a part of, that system.And it is not obvious that the appearance of an innovation, of whatever sort, automatically bringswith it an obligation to accept and pay for the corresponding interventions. The technology doesnot dictate its own range of applications -- or its prtce." Those result from the choices ofpeople.There is in fact no a priori reason why technological innovations should add to costs. Inthe rest of the economy, they usually have the opposite effect. But new technologies are theresult of more or less rational activity, in contemplation of profit. A bias towards cost-enhancinginnovations presumably reflects a judgement by investors that these are most likely to beaccepted in health care. They do not threaten anyone's place at the feast.When a new technology is developed, the initial marketing stage commonly emphasizesits potential for substituting for other costly interventions, more or less well identified. This seemsto be a standard claim, to blunt or rebuff the concerns of payers. In practice, however, theapplications often expand well beyond those for which effectiveness has been demonstrated, andthe new manoeuvre is commonly performed in addition to whatever forms of intervention werepreviously employed. Unless a procedure is particularly dangerous or very uncomfortable -­pneuma-encephalography, for example -- clinicians are reluctant to give it Up.33Thus the cost-enhancing tendencies of technology in health care are traceable to the waythe health care system functions; they are not a consequence of the characteristics of thetechnology itself. The rapid spread of interest in "technology assessment", again within the lastfew years, is evidence of the realization by governments and other payers that a more carefulevaluation of "What works, for whom, and when?" might significantly mitigate the cost-enhancingpressures associated with new technologies.25But again, it is worth noting that excellent "technology assessment" has been going on inthe U.S. for nearly twenty years, and there cannot have been many in health policy circles inother countries who were unaware of this (Evans, 1990b)." The sudden recent interest, outsidethe U.S., is consistent with the argument above. Global mechanisms of control could, at anacceptable political cost, hold cost escalation to rates supportable in a growing economy. Therewas no need for payers to investigate in detail the content of practice. Now that those moretraditional methods must be more rigorously applied, and are incurring increasing political costs,more "information-intensive" approaches become more attractive. Demands for more money tokeep up with the latest technology may perhaps be met with arguments that the technology itselfis unproven, or overutilized, or inappropriate.Building Political Pressure: The Recruitment of Public ConcernThis brings us back to the second point arising from the historical scenario above. Whatis the nature of the political pressure now being brought to bear, such as to result in suchwidespread calls for reform? After all, simmering and sometimes overt conflict between payersand providers has been observed throughout Phase II, without producing such a response.A simple answer might be that, insofar as the controls on costs are tighter in a low- or no­growth economy, the level of provider anger and distress is greater, and there is certainly sometruth in that. But payers and providers are not the only participants in these negotiations andconflicts. Each of them depends, ultimately, upon the support of the general public.The members of the public play several roles. As actual or potential patients, they useor hope to be able to use health services when needed. As voters they accept or reject thepolicies of governments that either pay for health care, or regulate those who do. And astaxpayers -- or in some cases premium-payers -- they ultimately bear the costs of the system.The objective of providers, or vendors, has been to convince the public to think like patients whenthey act as voters, and to support claims for "more". Governments have tried to remind them ofthe cost. Rarely has anyone invited them to consider the real benefits and risks of the servicesin question.But in fact most people, most of the time, are (fortunately) not patients. Moreover, whilethe future is always uncertain, people also differ greatly in their probability of becoming patients.26The risk of illness and the probability of using care are very unequally distributed. Nor are thosewho do become ill, likely to use the same amounts of services. For most of us, contact with thehealth care system means one or two visits to a doctor, and perhaps a prescription; for a few itmeans very intensive and/or very long-term care.Not surprisingly, sub-groups within the population vary considerably in the interest theytake in health services and health policy. "Consumer representatives" tend to be self-selectedfrom the sufferers from some chronic disease or problem, or their friends and relatives. Theordinary citizen, being ordinarily healthy, takes less interest, wanting primarily to be assured thatservices will be available if needed.What is of particular Interest, however, is that this inequality in the average (per capita)use by different sub-groups within the general population is Increasing quite markedly over time.The phenomenon is clearest among the very old and the very young. We noted abovethe steady increase in age-specific use and expenditure rates among the elderly, relative to therest of the population. In some countries and for some services (hospital days in particular), thiscorresponds to actual declines for people at younger ages. For others, use rates are going upfor all but much faster for the elderly. Neonatal Intensive care has the same impact at veryyoung ages, a rapid build-up in resources used by the very, very young -- and a very smallproportion of them -- along with a sharp reduction in paediatric hospitalization in general.This is perhaps not surprising. For people who are basically in good health, and whoseillnesses or accidents have reasonably well understood remedies, there Is no reason why use orcosts should be going up. Indeed, expert opinion has held for years -- decades -- that hospitalsin particular were grossly overutilized in a number of countries, and that patients would benefitfrom less use. But for the aged, or low birth weight infants, or those with chronic and/or incurableillnesses. the limits to appropriate intervention are much less clear. There is always somethingmore that could be done, if not this year then next, that might be beneficial -- the "needs" arenever met." So that is where expansion takes place.In the process, the professional perception of "need" can become shifted. Instead ofdescribing a remediable circumstance, it can refer simply to an unfortunate circumstance. Do the27sickest people have the greatest needs? On the second interpretation, yes, by definition; on thefirst, not necessarily. That would depend upon whether or not anything can be done to changetheir condition.Professional Arguments, Political ObjectivesThe second interpretation of need, and the concentration of services on the sickest, canthen be reinforced by arguing that the objectives of the health care system should go beyondimproving health, in a narrow sense, to improving quality of life. Additional services for those inmost need, even if they do not result in "cures", or improvements in functional state, may at leastmake the recipients happier. From the earlier but still fashionable WHO definition of health asa state of complete physical, mental and social well-being, it follows that any activity whichimproves quality of life can be regarded as meeting a need for "health" care.This broader interpretation can be very satisfying to providers -- especially those who carefor the increasing population of elderly who are physically or mentally infirm. It is also stronglysupported by those suffering from particular chronic diseases, who are increasingly well organizedpolitically as pressure groups.36 It provides a justification for more servicing; what it neglectsis the competing claims to quality of life of those who do not happen to be the special concemof the health care system.The issue is in essence a distributional one. Do those suffering from physical or mentalillness, whose health status in a narrow sense cannot in fact be improved, have a greater claimon collective resources to improve their quality of life than those, outside the health care system,whose quality of life would also benefit from more services, albeit of different kinds and fromdifferent (generally non-health care) vendors -- you and I, for example? And if so, why?In acting as advocates for "their" patients, providers are also trying both to increase thedemand for their own services -- marketing -- and to impose costs on those (unnamed) otherswho must provide the additional resources (or forego their own claims to service). The providerrole shifts from the professional -- applying expertise to the resolution of health problems -- to thepolitical -- transferring resources to themselves and their clients.28On the second interpretation of "need", then, a system whose total resources are notgrowing should allocate those resources to the oldest, the sickest, the people "in greatest need".It will concentrate its resources more .• even if this does not lead to the greatest impact on health•• and others whose "needs" thus interpreted are less immediate or less severe will be given alower priority. They may have to wait in queues, or simply be crowded out of the system. Alarger share of services will go to a smaller group of people .. which is what is happening •• andthe general public will become increasingly concerned about access, even though in totalresources are not, in fact, being cut back.If at the same time, considerable publicity is being given to alleged technicalbreakthroughs and other marvels of extraordinary expense, one can reinforce the impressionamong the general public, particularly those who do not actually use the system, that they are atrisk of being denied potentially beneficial, perhaps even life-saving, services. Health care is"underfunded", and "rationed". It is no accident that such claims in Canada •. the world's mostexpensive system, after the United States -- give a central place to coronary artery surgery. Thisis of particular concern to middle-aged men -- well represented in positions of authority .• whotend to perceive the procedure (generally wrongly) as potentially llle-savlnp for themselves.The core of the problem is that the priorities being set within the health care system, whenexpansion stops, may not match those of society at large. If the people whom the system isserving most intensively are a smaller and less representative segment of society, and if theservices they receive are less demonstrably effective in "curing" their conditions, and providerather an increasingly expensive form of maintenance, it is not surprising that the generalpopulation might begin to feel that the system which they support politically and financially is notin fact "theirs" any more. With increasing publicity about "cutbacks", waiting lists, and inadequateservices, people understandably fear that the service may not be available when they need it.The growth of markets for purely private care is an obvious next step, enabling providersto draw more money into health care and evade regulatory restrictions on budgets " andincomes. Where such private markets exist, it is remarkably common for the same providers towork in both mainstream and private care, and thus to be able to encourage patients with straight·forward problems and good incomes to "go private". It is hard to imagine any other institution,29public or private, tolerating such blatant confiict of interest, yet in many health care systems it isconsidered "normal".Nor has there been much incentive, in most health care systems as they have beenstructured until recently, for providers to do much about access problems. A long waiting list, inmany systems, indicated a very popular physician or institution, a need for more resources, andin some cases an opportunity to promote private care. The possibility that it might also be aresult of poor management of staff or facilities, co-existing with idle capacity, or inappropriatepatterns of care, was carefully avoided.Indeed, any internal problems, whether occasional errors of judgement or systematicfailures of management, can always be presented to the external world as the unfortunateconsequences of insufficient funding. "If only we had been given the extra money we asked for,none of this would have happened." Who, except perhaps the coroner or the plaintiff's lawyer,is in a position to argue? Certainly not the press.So long as waiting lists were advantageous to institutions or to key people in them, asseems to have been the case in the U.K, and Sweden, for example, no amount of extra fundingcould make them disappear. "Demand was infinite," But preliminary indications from Stockholmsuggest that when patients can shift their institutional affiliations from one clinic to another, andbudgetary allocations move with them, waiting lists can shrink in a matter of months. A changein the way in which budgets are allocated may serve to make internal system priorities moresensitive to those of the public that they are expected to serve.A similar result was achieved through another mechanism in both Vancouver and Toronto,cities which in 1989 were experiencing very long waits for cardiac surgery and loud cries ofunderfunding in the media. In each city a metropolitan surgical registry was set up whichpermitted (did not require) patients to go to the next available surgeon, and which made clearwhere they were on the list and how long they might have to walt." The "crisis" disappeared.Recapturing Delegated Priorities: The Ouest for "Management"This conflict over priorities, between those who provide services (and their long-termclients) and the general public who support and pay for, and hope be able to use (if necessary)30the health care system, is solidly rooted in Phase I of the evolutionary process sketched outabove. In that period, the social consensus in the developed societies in favour of expandinghealth care embodied, almost without discussion, the assumption that the health care systemwould play the same agency role vis-a-vis the rest of society that individual providers did forpatients. The task for the rest of society was that of developing institutions to mobilize theresources to "meet the needs", turning them over to providers of care to do the job, and ensuringthat everyone who needed care would have unhindered access.While systems for funding and delivering care differed, all delegated to, or rather left with,providers the authority to decide what should be done, for whom, and how -- Individually andcollectively to set the priorities for the system as a whole. Were they not the experts? Ineconomic terms, they were to (continue to) make the allocative decisions within health care, inthe role of "social agents".Phase II saw the development of limits on the flow of funds into health care, whichaddressed the allocative issue of the overall size of the health care system relative to the rest ofthe economy (though that "decision" seems to have been pretty arbitrary, and variable acrosscountries). But the process of budgetary constraint did not include, in most countries, anyattention to allocative or technical questions within the health care system itself. Providerscontinued to set priorities for the use of whatever budgets they were able to extract from the restof society, and to make their own decisions as to how to organize the process of production.There were increasing concerns expressed in the research literature, during the 1970s, aboutevidence of inefficiency in the use of personnel and facilities. But these did not lead to any policyresponse.In many national systems, particularly those which had originally developed as "socialinsurance", there were no policy levers whereby those outside the system could address intemalallocative or technical issues. Some influence could be exercised over capital investment,particularly physical capital, but once the buildings, people, and technologies were in place, theyoperated largely independently of any but financial constralnts."Thus the universal cry, in the current wave of enthusiasm for reform, is "management".Despite the wide variety of organizational arrangements in the different national systems, all have31the common problem that they were built on a foundation of delegated managerial control,decentralized and without accountability. If service priorities are to be changed, and technicalefficiency enhanced -- and these seem to be the current objectives everywhere -- then the locusof control has to be changed, and accountability to those outside the health care systemestabl ished.But who is to manage, to whom should they be accountable, and how is all this to beachieved? To these questions a number of different answers have been proposed, and severalare being tested. In the United States, "managed care" has a long history and has taken on anextraordinary variety of forms. The "traditional" Health Maintenance Organizations (HMOs) havemore recently been joined by prospective payment of hospitals (by DRG), Preferred ProviderOrganizations (PPOs), and utilization review and "case management" by private insurers, allrepresent attempts to transfer some part of the agency role, the control over who does what towhom, from providers to payers. In the U.K., District Health Authorities (DHAs) will purchase carefrom independent hospital trusts, at arm's length, on behalf of "their" patients. At the same time,groups of capitated "budget-holding" GPs will take both professional and financial responsibilityfor the care received, from themselves or others, by their patients. In the Netherlands, the"Dekker Reforms" suggest that insurers might take on this role as purchasers for their enrollees,rather than simply paying bills. In Sweden, experiments are underway in which patients canchoose to which clinic they will assign their capitation allocation, and from which they will receivetheir care.While the forms differ, these various approaches all involve the transfer of some portionof the agency role, the control over what is to be done, must be transferred away from providers.The question: "Whose patient is this?" will have a different and more complex answer in future.The assumption -- hope -- is that purchasers will then be able to transfer "their" patients from oneprovider to another, thus generating competitive pressures for efficiency and effectiveness, limitinginappropriate and costly diagnostic and specialist care, and ensuring that hospitals and specialistsare responsive to their patients' needs. Otherwise, they will take their business elsewhere.Much Promise, Some Understanding, No Solutions YetBut the key questions are:32(i) "Can these new purchaser institutions in fact exercise effective control, or at leastsignificant influence, over the care process, or will clinical "business as usual" reassertitself?" and(i1) "If they can, will their incentives be such as to lead to more satisfactory outcomes, or willefficiency gains be illusory, and/or "quality of care" suffer?"It is far too early to try to answer these questions from the European proposals, but thereis now a good deal of American experience to draw on. The results are mixed, but in general notencouraging. On the positive side, allegations of grave threats to "quality of care" from HMOsare now a generation old, and have never been substantiated. The same charges have beenlevelled at prospective payment of hospitals, and at PPOs, with the same result. There is noevidence that "managed care" leads to worse outcomes for patients, although it may threaten theautonomy of physicians.But the effects of the various forms of "managed care" on overall costs consists to dateof encouraging local successes adding up to general failure (Jones and Jones, 1991; Hadley andLangwell, 1991; United States, 1992).39 Individual sponsoring agencies, in a fragmented andcompetitive environment, have not been able to do much more than harass providers andcompete for a selected low-risk clientele, transferring costs rather than reducing them. And tothe extent that this competition leads to high-risk populations becoming "uninsurable", patientsdo suffer. The problem is not so much incentives to cut corners in clinical care, as to refuse toenroll and serve at all (Friedman, 1991). And such successes as have been achieved are offsetby the high cost of decentralized and competitive administration -- new and well-fed places at thehealth care feast."Separating purchasers from providers does not in itself deal with the first question above.If the providers remain in control of all the care decisions for "their" patients, as they have donein Canada, with the government "purchaser" doing no more than paying -- and arguing about -­the bills, the result may be cost control, but is not management." Germany has also had acomparatively good record in cost control by simply setting firm budgets and pro-rating fees sothat increases in utilization do not translate into increased expenditure. But this creates no directincentive for improved efficiency or effectiveness in the delivery of care.33In addressing the two questions above, there are four places to locate authority andaccountability for health system performance. These are:a. patientsb. governmentc. providers, ord. other non-government agencies (§Jl employers).Each nation's reform proposals blend some or all of these in different ways -- they are commonlycomplementary rather than mutually exclusive.Inthe U.K., for example, DHAs are directly accountable to the central government (b); theyare responsible for managing the care received by the persons in their regions, through contractswith providers. Assigning budgets to Gps, on the other hand, uses one part of the providersystem to influence the others (c); such Gps are in turn accountable to patients (a) who may ifdissatisfied move from one GP to another and take "their" budqets., The Swedish experimentspermit patients (a) to select among clinics which are responsible to county councils (b). In theU.S. staff-model HMOs have a strong internal management structure (c), and must "meet themarket" by attracting patient enrollees (a). But PPOs are often set up by employers (d) thatstrongly influence their employees' choice of providers. Insurers (d) may also list approvedproviders, in the hope of holding down costs and thereby attracting enrollees (a).Recommendations for change in Canada include regional, capitated budgetary envelopes,centrally set by provincial governments (b). But governments must take a much more activemanagement role, in conjunction with professional organizations, in evaluating the effectivenessof care and monitoring the correspondence of actual practice to external standards (b,c),While the development of lines of accountability for the health care system is still at a veryearly stage, certain firm generalizations do appear to be supportable. And all reinforce a pointmade by more sophisticated advocates of managed care, that "state regulation" and "the market"are not alternatives, but must be mutually supportive.For example, all (a) proposals which rely on individual patients choosing among alternativeinsurers (which includes implicit "insurers" -- American HMOs, Swedish clinics or Brttlsn GPbudget-holders) are vulnerable to selection problems." The insurer is strongly motivated to34recruit "profitable" patients, and discourage "unprofitable" ones. [The "profits" need not bemonetary.] A solution, if there is one, will require quite careful regulation. Systems which assignresponsibilities for regional populations -- British DHAs or Canadian provinces/regions -- avoidthis problem, but must then rely on political "voice" rather than individual "exit" choices toinfluence system behaviour (Hirschman, 1970).Second, all systems (even markets) function within an imprecisely specified but criticallyimportant matrix of values, expectations, roles, and "ethics" which define how people seethemselves and others, and what constitutes appropriate behaviour. There is a difference, forexample between "professional" and "profit-seeking" behaviour, although the economic theoristhas difficulty representing it. Proposals for reform which emphasize formal, contractual, arm's­length and self-seeking forms of interaction may erode this framework, or change it in unlookedfor ways.Third, external, politically enforced controls on health care will continue to be needed, bothto restrain the growth of total cost, and to ensure that all citizens remain in the system. Improved"management", howeverapplied, will not be powerful enough to offset the internal expansionarydynamics of health care systems (Wennberg, 1990) -- that requires global caps. Bettermanagement may, however, make the system more effective and more acceptable within suchcaps, and so make them easier to maintain. But the task will also be easier, if the most costlyor least rewarding people can somehow be dropped from the system, explicitly or implicitly.External constraints may be required to ensure that this does not happen."Fourth, the effective transfer of some portion of the agency role from providers to otherinstitutions or individuals will require much better information. This will include both evaluationsof the effectiveness of care, and data on actual patterns, as a basis for bringing the latter intoconformity with the former. But it will also include information on the convenience, accessibility,and general "user friendliness" of different providers.Political Problems Do Not Have Technical Solutions -- That WorkThis underscores the importance of maintaining, through the political process, centralrequirements and constraints. If individuals are to make decisions as to the use of services orinstitutions, then information requirements must be set and met. if on the other hand these35decisions are to be made by politically responsible agencies, then these agencies too need muchbetter information both to know what they are doing and to defend it against the inevitable politicalattack.But these decisions, by whomever, must be made within an authoritative and enforcedglobal, universal framework of expenditures and entitlements, which does not permit cost­increasing evasions whether or not disguised in pseudo-market rhetoric." At least that hasbeen the conclusion of most reform proposals thus far, and it seems a conclusion consistent withwhat has been learned about the way health care systems operate in the modem world. Thetask, in the present stage of health care reform, is to reassign the agency relationship, andreorganize managerial authority and incentives, without losing the centralized control which hasbeen essential for both equity and cost control.It appears therefore that the fundamental principles on which most of the health caresystems of the post-war period have been built, are sound -- as indeed most of the officialenquiries have concluded. But the organizational structure, of managerial authority delegated toproviders without adequate accountability, was unsound, and its weaknesses are now beingsharply exposed in the new and much less favourable economic climate.The challenge for all systems in the western European tradition -- including Canada,Australia, and New Zealand -- is both technical and political. First, what are the appropriate neworganizational structures which will yield greater accountability and efficiency without losing thestrengths of the existing systems? Nobody knows; and a lively interchange of comparativeexperience with new structures should be helpful.But the second, underlying problem is political. Can one assemble the constituenciesnecessary for major reform? This requires one to steer between those who believe, often quitecorrectly, that they themselves would benefit from the erosion or collapse of a universal systemand its replacement with something more fragmented and market-driven, and those who are sofirmly attached to existing systems, again often from interest as much as principle, that theycannot accept change. The former are engaged in deliberate sabotage; but the latter may bringabout the same result."36On the whole the political problem is probably the most dangerous: a race betweeneducation and disaster. Health care research, and the communication of findings, is much morethan just an "academic" exercise.37ReferencesAbel-Smith, B. (1985) "Who is the Odd Man Out: The Experience of Westem Europe inContaining the Costs of Health Care" Milbank Memorial Fund Quarterly 63:1:1-17.Abel-Smith, B. (1992) "Cost Containment and New Priorities in the European Community" TheMilbank Quarterly 70:3:393-416.Anderson, G.M., and D.J. Allison (1990) "The Appropriate Use of Intrapartum Electronic FetalHeart Rate Monitoring" UBC HPRU #90:3D, Centre for Health Services and Policy Research,University of British Columbia, Vancouver (January) (unpublished)Arrow, K.J. (1973) "Welfare Analysis of Changes in Health Coinsurance Rates" R-1281-0EOSanta Monica, Ca.: The Rand Corp.Banta, H.D., and S.B. Thacker (1979) Costs and Benefits of Electronic Fetal Monitoring: A Reviewof the Literature National Center for Health Research, Research Report Series. Department ofHealth, Education and Welfare (PHS) 79-3245, Washington D.C.: DHEW.Barer, M.L., R.G. Evans, C. Hertzman, and J. Lomas (1987) "Aging and Health Care Utilization:New Evidence on Old Fallacies" Social Science and Medicine 24:10:851-62.Barer, M.L., and R.G. Evans (1992) "Interpreting Canada: Models, Mind-Sets and Myths" HealthAffairs 11:1 (Spring) 44-61.Barer, M.L., C. Hertzman. R. Miller, and M.V. Pascali (1992) "On Being Old and Sick: The Burdenof Health Care for the Elderly in Canada and the United States" Journal of Health Policy, Politicsand Law 17:4 (Winter) (forthcoming).Black, C., N.P. Roos and B.J. Havens (1992) "Rising Use of Physician Services by the Elderly:The Contribution of Morbidity" Winnipeg: Manitoba Centre for Health Poiicy and Evaluation(unpublished).Blendon, R.J., and H. Taylor, (1989) "Views on Health Care: Public Opinion in Three Nations"Health Affairs 8:1 (Spring) 149-57.Blendon, R.J., R. Leitman, I. Morrison and K. Donelan (1990) "Satisfaction With Health Systemsin Ten Nations", Health Affairs 9:2 (Summer) 85-192.Biendon, R.J., K. Donelan, A.J. Jovell, L. Pellise and E.C. Lombardia (1991) "Spain's CitizensAssess Their Heaith Care System" Health Affairs 10:3 (Fall) 216-28.Blendon, R.J. and J.N. Edwards eds. (1991) System in Crisis: The Case for Health Care ReformNew York: Faulkner and Gray.Canadian Institute for Advanced Research (1991) "The Determinants of Health" CIAR Publication#5 (August) Toronto: the Institute.38Culyer, A.J. (1988), Health Care Expenditures in Canada: Myth and Reality, Past and Future,Canadian Tax Paper No. 82, Toronto: Canadian Tax Foundation.Culyer, A.J. (1991a) "Health Care and Health Care Finance in Sweden: The crisis that never was,the tensions that ever will be", Summary of an International Review of the Swedish Health CareSystem [Den Svenska Sjukvarden] Occasional Paper no. 33, SNS Sweden, Stockholm: SNS.Culyer, A.J. (1991b) "Reforming Health Services: Frameworks for the Swedish Review" inInternational Review of the Swedish Health Care System [Den Svenska Sjukvarden] OccasionalPaper no. 34, SNS Sweden, Stockholm: SNS, 1-49.Danzon, P. (1992) "The Hidden Costs of Budget-Constrained Health Insurance Systems", HealthAffairs 11:1 (Spring) 21-43.Dutton, D.B. and S. Levine (1989) "Socioeconomic Status and Health: Overview, MethodologicalCritique, and Reformulation" in J.P. Bunker, D.S. Gomby, and B.H. Kehrer, eds. Pathways toHealth: The Role of Social Factors Menlo Park, Cal.: The Henry J. Kaiser Family Foundation.Enthoven, A. and R. Kronick (1988) "A Consumer Choice Health Care Plan for the 1990s" (PartsI and II) New England Journal of Medicine 320:1&2:29-37 & 94-101.Evans, R.G. (1985) "Illusions of Necessity; Evading Responsibility for Choice in Health Care",Journal of Health Policy, Politics and Law 10:3 (Fall) 439-67.Evans, R.G., M.L. Barer, J. Lomas et al. (1989a) "Controlling Health Expenditure: The CanadianReality" New England Journal of Medicine 320:9 (March 2) 571-7.Evans, R.G., M.L. Barer, C. Hertzman et at., (1989b) "The Long Good-Bye: The GreatTransformation of the British Columbia Hospital System" Health Services Research 24:4 (October)435-59.Evans, R.G. (1990a) "Tension, Compression, and Shear: Directions, Stresses, and Outcomes ofHealth Care Cost Control" Journal of Health Politics. Policy and Law 15:1 (Spring) 101-28.Evans, R.G. (1990b) "The Dog in the Nighttime: Medical Practice Variations and Health Policy"in T.F. Andersen and G. Mooney, eds. The Challenges of Medical Practice Variations London:MacMillan, 117-52.Evans, R.G. (1991) "Life and Death, Money and Power: The Politics of Health Care Finance"In Theodor J. Litman and Leonard S. Robins, eds. Health Politics and Policy (second edition),Albany, N.Y.: Delmar, 287-301.Evans, R.G., M.L. Barer and C. Hertzman (1991) "The 20-Year Experiment: Accounting for,Explaining, and Evaluating Health Care Cost Containment in Canada and the United States" inG.S. Omenn, J.E. Fielding, and L.B. Lave, eds. Annual Review of Public Health 12:481-519.Faltermayer, E. (1992) "Let's Really Cure the Health System" Fortune (March 23) 46-58.Fein, R. (1992) "Health Care Reform" Scientific American 267:5 (November) 46-53.39Friedman, E. (1991) "Insurers Under Fire", Health Management Quarterly XIII:3:23-27.Gerdtham, U.-G. and B. Jonsson (1991) "Health care expenditure in Sweden - An intemationalcomparison" Health Policy 19:2+3 (December) 211-28.Getzen, I.E. and J.-P. Poullier (1991) "An Income-Weighted International Average forComparative Analysis of Health Expenditures" International Journal of Health Planning andManagernent 6:1:3-22.Hadley, J.P. and K. Langwell (1991) "Managed care in the United States: promises, evidence todate, and future directions" Health Policy 19:2+3 (December) 91-118.Hertzrnan, C. (1990) "Environment and Health in Czechoslovakia" UBC HPRU #90:23D, Centrefor Health Services and Policy Research, University of British Colurnbia, Vancouver (December)(unpublished).Hertzman, C., I. Pulcins, M.L. Barer et al., (1990) "Flat on Your Back or Back to Your Flat?Sources of Increased Hospital Services Utilization arnong the Elderly in British Columbia" SocialScience & Medicine 30:7:819-28.Hirschman, A.O. (1970) Exit, Voice and Loyalty Carnbridge, Mass.: Harvard.Jones, S.B., and J.M. Jones (1991) "Where Does Marketplace Competition in Health Care TakeUs? Impressions, Issues, and Unanswered Questions from the NHPF Site Visit to Minneapolis-SI.Paul (January 14-17, 1991)", (Washington, DC: George Washington University, June).[Liberman, T.] (1992) "U.S. Insurers on the Attack" Consumer Reports 57:8 (August) 587.Lornas, J. (1990) Finding audiences, changing beliefs: The structure of research use in Canadianhealth policy. Journal of Health Politics, Policy, and Law 15:3 (Fall) 525-42.Mackenbach, J.P. (1991) "Health Care Expenditure and Mortality frorn Amenable Conditions inthe European Community" Health Policy 19:2+3 (December) 245-55.MacKinlay, J.B., and S.M. MacKinlay (1977) "The Questionable Contribution of Medical Measuresto the Decline of Mortality in the United States in the Twentieth Century" The Milbank MernorialFund Quarterly/Health and Society 55:3 (Summer) 405-28.Marrnor, T.R. (1992) Japan: A Sobering Lesson" Health Management Quarterly XIV:3 (ThirdQuarter) pp.10-14.Molloy, W. and G. Guyatt (1991) "A comprehensive health care directive in a home for the aged"Canadian Medical Association Journal 145:4 (August 15) 307-11.Moore, T.J. (1989) Heart Failure: A critical inguiry into American medicine and the revolution inheart care New York: Random House.Naylor, C.D., A. Basinski, J. Frank and M.M. Rachlis (1990) "Asymptomatic hypercholesterolemia:a clinical policy review" Journal of Clinical Epiderniology 43:1029-1121.40Poullier, J.-P. (1989) "Health Data File: Overview and Methodology" Health Care FinancingReview (Annual Supplement, December) 111-94.Roos, N.P., and L.L. Roos (1992) "Small Area Variations, Practice Style, and Quality of Care",in R.P. Wenzel ed., Assessing Quality Care: Perspectives for Clinicians Baltimore, MD: Williamsand Wilkins, 223-38.Schieber, G.J., J.-P. Poullier, and L.M. Greenwald (1991) "Health Care Systems in Twenty-FourCountries" Health Affairs 10:3 (Fall) 22-38.Schieber, G.J., J.-P. Poullier, and L.M. Greenwald (1992) "U.S. health expenditure performance:An international comparison and data update" Health Care Financing Review 13:4 (Summer) 1­87.Sheils, J.F., G.Y. Young and R.J. Rubin (1992) "What To Expect from a Canadian HealthInsurance Program in the United States" Health Affairs 11:1 (Spring) 7-20.Taylor, H. and R. Leitman (1991) "Consumers' Satisfaction with Their Health Care" in R.J.Blendon and J.N. Edwards, eds., System in Crisis: The Case for Health Care Reform New York:Faulkner and Gray, 75-102.Townsend, P., N. Davidson, and M. Whitehead (1988) Inequalities in Health London; Penguin.United States, Congressional Budget Office (1992) Staff Memorandum (K. Langwell) (June).Verbrugge, L.M. (1984), "Longer Life But Worsening Health? Trends in Health and Mortality ofMiddle-aged and Older Persons" Milbank Memorial Fund Quarterly/Health and Society 62:3:475­519.Waldo, D.R., S.T. Sonnefeld, J.A. Lemieux, and D.R. McKuslck (1991) "Health Spending Through2030: Three Scenarios" Health Affairs 10:4 (Winter) 231-42.Wennberg, J.E. (1990) "Outcomes Research, Cost Containment, and the Fear of Health CareRationing" New England Journal of Medicine 323:1202-1204.Wennberg, J.E. (1992) "Innovation and the Policies of Limits in a Changing Health CareEconomy" in A.C. Gelijns, ed. Technology and Health Care in an Era of Limits Washington, D.c.:National Academy Press, 9-33.Wildavsky, A. (1977) "Doing Better and Feeling Worse: The Political Pathology of Health Policy"Daedalus 106:1:105-24.Wilkinson, R.G. (1992) "Income Distribution and Life Expectancy" British Medical Joumal304 (18January) 165-8.Woolhandler, S., and D.U. Himmelstein (1991) "The Deteriorating Administrative Efficiency of theU.S. Health Care System", New England Journal of Medicine 324:18:1253-8.41Endnotes1. During this time. each of the ten Canadian provinces. has appointed a Royal Commission or equivalent officialenquiry into health care; the U.K. government has moved from issuing White Papers to a process of restructuring ofthe National Health Service; the Netherlands produced first the "Dekker Reform" proposals and in 1992 the Report ofthe Government Committee on Choices in Health Care; the New Zealand government put forward in the "Green andWhite" paper of 1991 a series of radical proposals for change; the new (1991) government of Sweden has appointeda parliamentary committee to review that country's health care system...and so on.2. Enoch Powell, minister of health in the UK in the early 1960s, referred to the "chorus of complaint that rises dayand night from every part of [the NHS]." In Canada, conferences on "the crisis in health care" have been held withgreat regularity since the early 1970s -- immediately atter the present system was completed.3. These observations lead quite naturally into an appealing metaphor -- are health care systems themselves "sick"? ­- a neat rhetorical inversion which can be dangerously misleading. It presupposes an alternative state of "health", ofsocial homeostasis which has never existed outside the dreams of nostalgic physicians or market-intoxicatedeconomists. Real-lite systems evolve out of the shitting balance of advantages among the interests involved -­providers, patients, and payers -- which are partly congruent but partly in conflict. "Crises" are perceived by whicheverset of interests feels threatened. This situation is well characterized by Culyer (1991a) in reference to Sweden; 'Thecrisis that neverwas; the tensions that ever will be."Worse, the "sick system" metaphor invites the response: "Call the doctor" -- seek out the expert and followhis/her orders. Political, distributional questions are misrepresented as technical or "scientific", with "right" and wrong'answers, to the considerable material advantage of whichever group -- physicians? economists? ethicists? -- cansuccessfully assert its claim to be the relevant "experts". The metaphor thus serves as marketing for professionalservices, while distracting attention from the fundamental political choices which must be made.4. That is not to say that no system has "failed", or that there are no differences in performance on generally acceptedcriteria. Such differences do exist, and at least one system, that of the United States, is in the judgement of manyobservers clearly failing. The level and focus of distress in different countries is to some degree responsive to theirparticular patterns of strength and weakness. But general phenomena must have general explanations.5. A trend which has itself been viewed with alarm, and greeted with considerable outcry, by those for whom thoseexpenditures represent incomes. The providers of care more commonly perceive underfunding.6. While the correlations found by Getzen and Poullier are remarkably close, they do leave room for "non-endogenous"policy.Two countries, the U.K, and Denmark, have exercised tighter control over their health costs than income trends alonewould predict. And both Canada and the U.S., while within 20% of their "predicted" levels of expenditure, show a cleartrend over time relative to that level. The former falls; the latter rises.7. In effect a health care system behaves like a neoplasm, with a blind drive for growth. It seeks constantly foradditional sources of vascularization to support itself, regardless of consequences for the rest of the organism.8. The resurgence of interest in user charges and other forms of "privatization" which move the costs back ontopatientsis an example of such an effort at burden-shifting. Those economists who continue to advocate user charges on theground that they will limit costs, provide a convenient, if specious, public relations cover for such re-distributive policies.The widespread support for such measures by providers of care, whose incomes would in total fall ~ healthexpenditures shrank, indicates their judgement that in fact tapping private sources of finance is a way to underminepublic controls and expand total costs. Indeed, "more private funding" is very commonly offered explicitly by providersas a remedy for "underfunding",9. Efforts to change health care systems have followed the general economic decline by several years. At thebeginning of the downturn, many of those responsible for policy assumed that they were dealing with a normal businesscycle, continuing with "business as usual", and waiting for the recovery. In addition, the political counter-pressureswhich result from stringent controls on health care are cumulative. Slow or no overall growth (which in health care42rhetoric is called "cutbacks") may be bearable for a short time, but a continuing deferral of the return to 'normal" growthleads to growing distress.10. Politicians, whose jobs depend upon knowing such things, emphasize the breadth and depth of this support, andthe constraints it places upon change.11. It may be worth noting, however, that the extreme and prolonged dissatisfaction of the American public with theirhealth care system has thus far led to no significant change. In that political environment at least, unhappinessproduces not change but proposals for change, whose number, diversity, and complexity seem to numb and cripplethe opposition to the status quo.12. Expenditures do not correspond directly to real resources; Gerdtham and Jonsson (1991) find considerableinternational variation in the relative price of health care. The high level of expenditures in the U.S. is largely accountedfor, on their estimates, by the higher relative prices of health care. When these price differences are adjusted for, theyestimate that the average quantity of health services received per capita is the same in Canada as in the U.S. andsubstantially greater in Sweden, while several other countries are within about 10% of the U.S. level. Price dITferencesthen emerge in relative incomes. The institutional arrangements in the United States have produced physician incomesthat are about six times the American average wage, in the Nordic countries the ratio is around two (Gerdtham andJonsson, 1991; Fein, 1992). Other European countries fall in between.13. Americans seem. however, to be quite schizophrenic. They share the general normative position. and seemgenuinely concerned about the circumstances of their tens of millions of un- and under-insured citizens. Vet they alsohold, apparently more strongly than in any other society, that inequality in the distribution of the material benefits of civilsociety is right and just, presumably on the assumption that each person's benefits match his contributions and merit.And they are profoundly ambiguous as to whether these benefits should include health care. Finally, there is anextraordinary degree of antipathy toward governments, and taxation, which nevertheless are directly or indirectly theonly instruments through which equity in health care can be achieved -- or at least approached. The result is, inMarmor's words, a powerful and widely shared commitment to providing all assistance short of actual help.14. One should not, however, be deceived by those who seek to elide the distinction between "every system has someinequities" and "every system has similar inequities". Even on such limited information as is available, ~ is clear thatsome systems are much more equitable than others. Such blurring of distinctions is part of the deliberate myth-makingof the American private insurance industry, now under heavy attack as a massive source of "cost without benefit"(Woolhandler and Himmelstein, 1991). Those whose livelihoods depend on maintaining inequality of access and ofburden -- for that is precisely what risk-based insurance underwriting does, quite deliberately and explicitly -- seek todefend their positions by misrepresenting and denigrating more successful national systems (Liberman, 1992).15. Economists typically distinguish between allocative efficiency -- producing the right mix of output -- and technicalefficiency -- producing that output in the right (least cost per unit) way. More recently, students (and practitioners) ofhealth care have begun to distinguish between "doing things right" and "doing the right thing". 'Doing things right"usually refers to the quality of performance, as a clinician would judge it, though ~ could include cost considerations;"doing the right thing" refers to providing appropriate or effective care. Both are embodied in the criterion ofeffectiveness above: a system which provides the wrong care, or the right care incompetently, will not be effective.16. The boundary between "technical" and "allocative" efficiency tends to be ambiguous, if not arbitrary, when oneconsiders actual services. A highly efficient laboratory, owned by a practitioner who uses ~ to perform unnecessarytests --- those which add nothing to the diagnosis, therapy, or outcome, but only to reimbursement -- representstechnical efficiency but allocative inefficiency in the production of tests, and technical inefficiency in the treatment ofthe condition for which the test was (inappropriately) ordered. The labelling of the problem depends upon one'sdefinition of output, though the problem itself is not changed thereby.17. Once again, the United States provides an exception. Considerable work has been done, and more is in process,looking at the comparative efficiency of systems of payment, and it seems quite clear to everyone but the Americaninsurance industry and its spokesmen that multiple, decentralized and unco-ordinated private payers are a hopelesslyinefficient and costly way to organize the financing of care. But the excess cost of this inefficiency in the United States,$60 to $120 billion depending upon who is counting, represents an equivalent sum in incomes and careers at risk."Feather-bedding" has never been an activity peculiar to the public sector.4318. This is of course only a short-run consideration, relevant over, say, one working lifetime at most....19. The public at large are ambiguous, as befits those who have a stake on both sides of the issue. As noted above,public satisfaction tends to be higher in countries which spend more on health care. On the other hand, cltizens havebeen willing, at least up to a point, to support governments which, also as noted above, have been able to contain thecosts of health care.20. Robinson Crusoe does not buy insurance, or negotiate health care budgets. Insurance might be bought in a societyof identical individuals, who each had the same probability of becoming ill, but no underwriting would be necessary -­no risk differentials -- and no equity questions would arise since everyone was equal in health and wealth. Moreover,since everyone is by assumption a part-time doctor, reimbursement is not contentious,21. Oh well, economists have always known that economic policies are commonly misguided because most peopledo not understand economic issues. Except of course in their personal consumption, labour supply, and productiondecisions, which are periectly informed and rational, and may even embody rational expectations based on a fullunderstanding of how economies function....22. Obviously the economic incentives facing physicians who are paid fees for their services will differ from those facedby salaried or capitated physicians, and one would expect their behaviour to differ somewhat as well. But producersof new drugs and devices -- the technology industries -- operate in an increasingly globalized market. And whateverthe form of payment, more is always better.23. "Inexplicably" may be inappropriate. There are a number of explanations, all different.24. In the United States, these include the income tax subsidy of $40 billion or more, the historical favourabie taxtreatment of non-profit insurers, and somewhat less directly, the assumption by the public sector of the highest risks ­- the old and the poor.25. Even in the United States, though the forms and the rhetoric are private, the money is mostly public. Governmentsprovide, directly and indirectly, nearly two dollars for every dollar that is generated from private insurance, but they donot gain a commensurate degree of control.26. Just as an example, the "new paediatrics", the behavioural and psychological problems of children and adolescents,emerged in North America very shortly after the collapse of birth rates in the mid-1960s threatened to reducesubstantially the clientele for the "aid paediatrics."27. These are the most common arguments that "more is needed"; other claims are less international. For example,apologists for the American system have claimed that costs are higher in the U.S. because there are more poor people,and poor people are generally in worse health -- "our patients are sicker" again. Both propositions may be true, butwhat is ieft out is that in the U.S., poor people are much less likely to be insured, and so receive less health care, notmore. The spread of AIDS is similarly pressed into service, despite the fact that, over the population as a whole, ltsimpact on costs is known to be trivial.28. A changing population mix, however, results in a different mix of priorlties. An aging population, in a no-growthenvironment, implies a reallocation of resources among sub-sectors within health care --- more for home care, less forsub-specialists. But this leads to the unthinkable, removing places at the feast. Hence the insistence that changedpriorities require expansion, not re-allocation.29. Reviewers in Canada, on the other hand, have concluded on an assessment of the scientific ltterature that defining25% of the adult male population as "sick" and promoting the extensive use of prolonged drug therapy, is not in facta way to enhance health (Naylor ~, 1990).30. Did the public suddenly "demand" non-steroidal anti-inflammatory drugs, for example? These have recently beentaken off prescription in North America, and are now heavily advertised as substitutes for the much cheaper (but non­branded) aspirin which is in general equally effective -- and whose side-effects are at least known.4431. This general discussion excludes two additional phenomena. There is clearly continuing pressure to expand therange of services which are accorded the extraordinary regulatory and financial privileges that attach to 'hearthservices". Providers on the excluded fringe, and their clients, have an obvious interest in being included under publicprograms. But this is not so much an issue of "public expectations" for hearth care in general, as a standard politicalproblem of: "Who shall benefit from collective support?"Secondly, as Hertzman has pointed out, our societies have only two kinds of institutions •• prisons andhospitals .- for people who are simply no longer able to cope with their life circumstances, and 'give up'. In lessextreme cases, people may turn to health care providers with problems of personality or external circumstances thathave little to do with health as usually conceived, simply because there is nowhere else to go. It is not obvious thatthese "public expectations" are in fact growing, arthough providers often complain of such patients. The appropriateresponse, however, would seem to be to target the underlying problem, rather than to expand the hearth services.32. Is tissue plasminogen activator (TPA) "worth" ten times as much as the alternative "clot-bustinq" drug,streptokinase? The developer, Genentech, hoped that buyers would think so. But its superiority in clinical applicationsturned out to be elusive, forcing Genentech to engage in heavy promotion to support its aggressive pricing. In anothercase, a drug long used in veterinary medicine has turned out to have applications in humans as well •• but the priceis raised a hundred times. "Technology" did not dictate these decisions.33. Tonsillectomy is a classic example. Researchers were raising very serious questions about the appropriatenessof this procedure in the 1930s, and the post-war antibiotic revolution removed most justilications for its performance.But this "ritual surgery" persisted until those who had originally been trained to do it retired and were replaced by a newgeneration. Electronic foetal monitoring provides a more recent example; Banta and Thacker (1979) pointed out overten years ago that it had been adopted in the absence of any controlled trials of its effectiveness. Now the trials arebeing done, and are not showing the procedure to be of general value (Anderson and Allison, 1990). Its indiscriminateuse may contribute to unnecessary caesarian sections. But the technique remains firmly established in obstetricalpractice.34. In Canada, attempts were made to set up a similar process in the federal Department of National Hearth andWelfare in the late 1970s, but there was little interest at senior levels.35. Few "cures" have been found, for age or chronic illnesses, though the range and sophistication of interventionsincreases rapidly. As the old dermatologists' joke has it: 'The patients never die; they never get better; and they neverwake you up in the middle of the night."36. People with AIDS are an obvious example, but advocacy groups forthose with mental illnesses are also prominent.Indeed the politics of disability seems to be blurring the distinction between ensuring that people are not arbitrarilybarred from using the abilities they do have, to compensating them •• at some one else's expense •. for the abilitiesthey do not have. This creates an obvious attraction for the sufferers from any chronic illness to try to re-labelthemselves, for political purposes, as "disabled".37. For Vancouver, the provincial government also paid for a certain number of operations in Seattle, to make up forthe delays caused by a month-long nurses' strike, and a temporary shortage of perlusionists. But patients seemedrather reluctant to take up this alternative.38. Indeed, it may be that some of the recent European calls for "social insurance" payment systems to replace directpublic finance, are motivated precisely by the expectation that this will make tt~ ddficult to achieve systemaccountability and cost control. There may also be opportunities, in the present polltical climate of "privatization", forthe financial services industry, and lts tame management consultants, to get their share of places at the health carefeast, And in any case, the overall funding system is likely to be more regressive •• a clear benefit from some pointsof view. It might even permit the healthy and wealthy to "go private'·· opt out of the universal system·· on terms moreadvantageous than, say, the present NHS in the U.K. Multiple forms of insurance are likely to be more acceptablepolitically, and less transparent in their distributional effects, than multiple tax regimes.39. The traditional staff-model HMOs continue, as they always have, to hold their costs somewhat below the aggregateupward trend, but not to change that trend.4540. Those who make their living by marketing managed care systems have an obvious incentive to report success,and they do. But local success -- for a particular employee group, for example, can be achieved by selection of patientsor providers. If successful managed care systems push high-cost providers on to somewhere else in the overall system,then they can show relative gains while overall system periormance is unaffected. This seems to be what is happeningin the U.S.41. There are, however, indirect incentives, insofar as the professionalism of providers, in a budget-constrainedenvironment, encourages them to use scarce resources efficiently so as to be able to do more for their patients. Thisincentive is not trivial, but it is apparently not sufficient either. The urge to "do more" seems to lead to heavy pressurefor more resources, long before all the inefficiencies are squeezed out.42. Wennberg's approach of providing patients with greatly increased information about individual treatment options,and then permitting them to make the critical decisions, is an alternative way of establishing accountability to patients.In general, most analysts have concluded that patients can never hope to have -- and may not want -- the informationnecessary to make treatment decisions, though they may well decide where to seek treatment, and in most systemsdo. The approach certainiy deserves further extension, but ~ is probably best viewed as embedded within theallernatives at the system level. As noted above, it is being tried in American capitated HMOs, and is unlikely to beas quickly acceptable either in fee-for-service environments, or in those with a strong tradition of medical paternalism.43. Capitated systems, pubiic or private, try to deal with this problem by setting capitation rates that vary accordingto the expected use/cost of the individual. Work in this area is progressing, but the fineness of adjustment which ispossible to date falls far short of what is assumed by hypothetical models of "competitive" systems. Some patients arestill much more rewarding than others, and it may always be so.44. Some of the most committed advocates of competitive market allernatives concede that their position is foundedsolely on faith, having no observational support (Faltermayer, 1992). Others abandon observation altogether, in effectdefining as optimal, whatever pattern of outcomes emerges from (whatever they allege to be) a "market system" -- evenif that system itself bears no relationship to the hypothetical markets of the economic textbooks on which thefundamental theorems of theoretical welfare economics are based~ Danzon, 1992).45. Domestic sabotage also receives encouragement and support from outside. This has three distinct origins. Privatebusiness firms operating in the U.S. market try to create and expand into foreign markets as well, and to promoteconditions favourable to expansion. But there also seems to be an extraordinarily poweriul ideological urge -- themarket as religion -- which lies behind American efforts to "assist" the eastern European countries in particular to movetoward more costly, less equitable, and less efficient health care systems.Most recently, however, the growing realization that the U.S. system is not only very unsatisfactory, but isgetting worse, has pushed health care reform to the top of the policy agenda in that country as well. The defendersof the status guo, the private insurance industry and organized medicine, are therefore conducting an extensive andwell-funded "disinformation" campaign to convince Americans that, however bad their own system may be, anyproposed "foreign" alternative is much worse. The distortions and outright lies that now circulate within the U.S., havea tendency to jump borders as well, particularly to near neighbours like Canada (Barer and Evans, 1992).


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items