Abstract

Health reform is complex, and simple solutions tend to be elusive. The speakers at the Strategic Levers for a High-Performing Health System symposium presented valuable insights about two key issues confronting health system reform: strategic purchasing and equity. Given the dollars at stake - $148 billion annually and counting in Canada, nearly $100 billion of which comes from the public purse - and the fact that we are often talking about life-and-death decisions, it is hardly surprising that interests, power and perceived legitimacy affect how we choose to proceed. What can we conclude from what we heard?1

Strategic Purchasing

Narrowly understood, strategic purchasing can be considered synonymous with prudent buying: getting good value for money, specifying with some precision the terms of a transaction, managing risks and negotiating good prices. In effect, it shifts the argument from who should finance care (Figueras et al. 2005; Marchildon et al. 2004; Mossialos et al. 2002) to who should deliver it (Deber 2004). Strategic purchasing also entails addressing the incentives inherent in various payment mechanisms. As the symposium's speakers clarified, these considerations, in turn, involve a series of questions, including the following:

  • Who decides about what? Ron Sapsford's discussion of the extent to which such decisions should be decentralized touched on the roles of Ontario's new Local Health Integrated Networks (LHINs) and Ministry of Health and Long-Term Care. Reinhard Busse observed similar issues in Organisation for Economic Co-operation and Development (OECD) countries.

  • What criteria will be used? Quality? Volume? Cost? Outcomes? Innovation? Satisfaction? And how will performance be measured? Adalsteinn Brown commented on key issues involved in such activities.

  • What issues are entailed in deciding whether systems should rely on public delivery or shift to private providers? Gerry McSorley described the shift as it unfolded in the United Kingdom (UK) and Reinhard Busse addressed the transformation in the OECD systems that use public delivery.

  • What are the issues involved in using for-profit or not-for-profit delivery? Joe Murphy discussed this topic as it applied to British Columbia (BC).
  • What is incorporated in the delivery model? Paul Williams discussed sector-specific models (emphasizing home care) and David Levine addressed Quebec's integrated models.

Putting it all together, we are prompted to ask, "When might strategic purchasing work?" Here is the short answer: "It depends.

"The concept of strategy suggests that purchasers must make choices, invest and buy coherently and should ultimately be accountable for achieving their declared ends. For a number of reasons, this has proven to be a very tall order in healthcare (Belli 2004; Figueras et al. 2005). One must consider, for example, the characteristics of what is being purchased, for whom and the processes used to determine how purchasing is done.

One element of the "what" question relates to the outcomes sought. At an aggregate level, once a certain level of spending has been achieved there is no clear relationship between per-capita spending on health - which is itself a function of the mix of services, volume provided and price paid for each - and such aggregate outcomes as life expectancy and infant mortality rates (Kanavos and Mossialos 1999; Leon et al. 2001; Retzlaff-Roberts et al. 2004). Even considering the specific items, the huge body of research attempting to estimate the cost per Quality-Adjusted Life Year (QALY) has its own controversies, including how best to capture softer outcomes (Birch and Donaldson 2003; Daniels and Sabin 1998; Deber and Goel 1990; Donaldson 2004; Drummond and Sculpher 2005; Gold et al. 1996; Menzel et al. 1999).

An optimist might suggest that strategic purchasing could help to deter or eliminate demonstrably inefficient purchases that have appeared otherwise impervious to change (Fisher et al. 2000; Wright et al. 2002). Others might note that the first problem is specifying what is to be purchased in broad terms. Strategic purchasing will be enormously influenced depending on whether one seeks to buy the greatest health impact per dollar (a utilitarian approach), to meet needs (however defined), to enhance equity or even to pursue additional policy goals such as encouraging job creation or research, building healthy communities or respecting clinical autonomy. Similarly, concepts such as "value for money" can imply cutting costs by a variety of methods, including lowering wages and salaries, improving productivity, rearranging the division of labour or reducing utilization, not all of which are compatible with each other and which might or might not be widely endorsed. Indeed, to the extent that some cost cutting has replaced full-time with part-time/casual positions and has otherwise affected recruitment and retention, the resulting shortages of skilled labour (particularly, but not exclusively, in nursing) have often meant that short-term savings have translated into higher long-term costs (Alameddine et al. 2005, 2006; Simoens et al. 2005).

A second element of the "what" question relates to the production characteristics needed to generate particular goods and services. Classical economics assumes that markets, which balance supply and demand, can ensure efficient outcomes. But this, in turn, presupposes perfect (or at least reasonably perfect) competition, whereby no single buyer or seller can dominate the prices to be paid. It also implies that the supply of services can expand or shrink to balance changing demand, which further requires the presence of excess supply that can be called into service if demand increases. As Alan Hudson noted in his presentation, additional capacity can come either from existing providers operating more efficiently or from new ones, which in turn suggests something about how easy and desirable it is for new providers to enter the market and old ones to exit it. Clarifying where strategic purchasing might work thus leads to the need to examine what economists term contestability, measurability and complexity (Deber 2004; Preker and Harding 2000).

Contestable goods are defined as being characterized by low barriers to entry and exit from the market. In contrast, non-contestable goods may have any or all of high sunk costs, monopoly market power, geographic advantages and asset specificity (a technical term used to refer to difficulties in redeploying assets from one use to another). For example, the equipment and skills needed to perform open-heart surgery have few alternative uses, and few policy-makers would wish to "waste" highly skilled and trusted providers merely because they had not won a particular competitive contract, or to risk the erosion of clinical skills by allowing volume to fall below a critical mass. The ability to contract selectively is larger for contestable than for non-contestable goods.

Measurability relates to how precisely the inputs, processes, outputs and outcomes of a good or service can be measured. Again, monitoring performance is easiest when measurability is high. It is relatively simple, for example, to specify the performance desired for laboratory tests. In contrast, it would be more difficult to specify the activities to be expected of a general practitioner. Selective purchasing is simplified for measurable goods.

What might be done, however, if measurability is low? The transaction costs of monitoring can be high. These issues are not unique to healthcare. Examination of other fields, including military procurement, suggests "the more completely rules, obligations, and procedures are defined in order to enforce accountability, the higher the price in time, money and flexibility" (Donahue 1989: 108).

Complexity refers, somewhat confusingly, not to goods themselves but to the extent to which they are "stand alone" or must be coordinated with other elements of care. Laboratory tests might be highly measurable, but they are also less useful if their results cannot be delivered promptly to clinical decision- makers. Again, it is difficult to hive off goods that must be integrated with other services. McSorley accordingly noted the need to pay careful attention to the extent of "unbundling" deemed desirable.

Competitive markets may also be hard to sustain, particularly when only one purchaser exists. In such cases, potential suppliers are likely to request guarantees of volume. Taken together, this implies that it is easier selectively to purchase cleaning services than open-heart surgeries. Murphy noted that BC has been relatively successful in contracting housekeeping and food services, both of which are relatively measurable, contestable and non-complex. As McSorley suggested, for less-contestable goods one would also make a case for longer-term contracts, as opposed to encouraging continuous competition. And as Williams showed, when goods have production characteristics that are incompatible with competitive markets, costs can go up rather than down, as occurred with home rehabilitation in Ontario (Randall and Williams 2006).

The literature suggests that not-for-profit providers have a better record of providing services in the interest of clients if this requires going beyond the precise terms specified in contracts (Deber 2004). When measurability is low, this willingness to do more can produce superior outcomes. As the literature has noted, it is important to distinguish between high-trust and low-trust models; at a certain stage, it might be wisest to adopt the concept of stewardship and encourage those providers who, because they have goals other than profit maximization, can ensure needs are met even if purchasers have not clearly specified them (Saltman and Ferroussier-Davis 2000; Saltman et al. 2002).

Indeed, there are also issues concerning how to balance competition with the sort of cooperation required to achieve better integration and coordination. Such balancing has been a major focus for health reformers, leading to questions such as, "Do we wish to lose key providers?" and "How do we distinguish between encouraging best practices and protecting intellectual property?"

The "who" question is also important - are we purchasing for a population or for the small proportion of people who are sick and use healthcare services? In general, health expenditures are heavily skewed: the lowest-spending 50% of the population accounts for less than 5% of expenditures on hospitals and physicians (Berk and Monheit 2001; Deber et al. 2004; Forget et al. 2002). There is a risk that the wrong payment mechanisms will provide a strong disincentive to serve such high-cost clients, particularly in a competitive market. There are also ongoing issues about whether purchasers should be meeting needs or demands.

Finally, it is important to clarify how decisions will be made, and by whom. A "social good" orientation will lead to a different notion of strategic purchasing than will a "rights" orientation; such factors as decision-making criteria and the entitlements to care under different conditions will vary. Formally, one could achieve consensus by having people express their preferences behind a Rawlsian "veil of ignorance," whereby they would assume they have an equal probability of being rich or poor, healthy or sick. But in real life we are not so ignorant of our circumstances and our preferences might well shift over time and in response to our own, our family's or our community's experiences.

As the symposium speakers suggested, there is no magic solution to the strategic-purchasing issue. Nonetheless, tough cases should not distract us from solving simpler problems. The very term strategic purchasing is loaded, as Williams observed in his presentation. It suggests an ethos of consumerism and choice, and the challenge is to determine whether and where it can be successfully applied. Some guides for genuine strategic purchasing are as follows:

  • Be precise and transparent about exactly what we want to purchase, for whom, how, by whom and why. Similarly, recognize that certain goods and services might not lend themselves to strategic purchasing.

  • Where purchasing is deemed appropriate, ensure there are different consequences, rewards and other signals for good and bad performance, however defined and nuanced they might be.

  • Capacity, authority, and accountability must be aligned. Busse highlighted the role of genuine devolution in using resources more effectively. In Canada, responsibility has already been devolved from the national to provincial governments, but it has proven difficult to shift any meaningful accountability from the provincial to the regional or local levels, even when regional health authorities exist. Who should decide? What some would term democratic control, others might call politicization. To what extent should purchasers be empowered to make allocative decisions and be held accountable for them - as Busse notes is the case in European systems - as opposed to holding governments accountable for meso- and even micro-level events?

Equity

A similar definitional problem arises when we speak of equity. During the symposium, Gwyn Bevan emphasized equal resources per capita, with a stress on inputs and access. Anthony Culyer added the need to pay attention to outcomes, outputs and process. He emphasized efficiency, with the strong suggestion that ineffective healthcare ought not to be provided at public expense. The question of whether effective healthcare should be purchased regardless of cost, however, remained unresolved, and gave rise to the question how would/should cost-effectiveness be incorporated into an equitable health system?

To make progress on health equity, it will be necessary to leave the comfort zones of rhetoric and hand wringing, and to pay attention to language, politics and evidence. Vocabulary matters; "equity" is an elastic term that people with very different notions of distributive justice can all embrace. It can be used to justify a whole host of political agendas. As such, unless it is carefully defined the term confuses more than it clarifies. As Stone (1997) has noted, equity can refer to the characteristics of potential recipients of an item, to the item itself and/or to the processes used to distribute it. During her talk, Jeanette Vega (not represented in this collection) highlighted large discrepancies across various nations. It is fine to follow Aristotle's injunction to "treat likes alike" but who will be classified in the group of "likes"? To what extent should Canadians be concerned about health outcomes in other countries? Indeed, as care in Ontario regionalizes, to what extent should residents of one LHIN be concerned about health outcomes in another? Again, a series of compelling, pivotal issues arise, which we will now discuss.

Precisely what do we want to be equitable about - health or healthcare? And are we concerned with short-term or long-term outcomes? If we focus on the short term, which is where most policy-makers spend their attention, where does prevention fit?

How much inequality of access and/or outcome is acceptable? Indeed, how much is addressable by public policy? Obviously, some people will always be healthier than others, even after adjusting for every conceivable social determinant of health. Genetic endowment, behaviour and luck will all play roles. As population health researchers have ably described, inequality might also be related to class, gender, race/ethnicity and geography, which to varying degrees interact (Evans et al. 1994; Starfield 2006). Socio-economic status (SES), which is about more than money, appears to be the dominant influence on inequality. From the evidence available, to be in favour of wide socio-economic disparities is to be in favour of irreducibly wide health disparities. There is no society-wide consensus on the acceptable degree of general inequality; indeed, this is the appropriately contested ground of democratic politics. The odds against there being a true consensus - implying near unanimity - on these matters are long. This is not fatal to the enterprise of reducing inequalities; it means merely that we should understand the nature of what is required to make it a public policy priority in a democratic context. The essential condition is not consensus but reasonably stable, consistent and sustained majority support.

If one takes a single set of definitions linked to a coherent set of principles, as Culyer suggested, how can the quest for such support be balanced against the recognition that trade-offs are inescapable, that most difficult decisions generate winners and losers and that the losers will, in turn, try to change the rules to increase their probability of winning (Schattschneider 1964)? Participation is one way to do this, which in turn leads to consideration of who is seated at the deliberative table.

Sustained democratic support confers legitimacy, which entitles the state to act and allocate even if some people do not get what they want some of the time. Here the class divide becomes critical. Canada's healthcare system is a cross-subsidization scheme whereby the healthy and wealthy pay for their own care and for a large proportion of the care of the poor and the sick. Thus far this transfer of wealth has been broadly accepted across Canadian society; approval, however, is contingent on the subsidizers continuing to believe that the system by and large serves them at least adequately, and preferably that it serves them well (Evans 2006). As a result, the system is organized mainly to meet the needs of the middle class (and often for the convenience of providers). If a fundamental reorganization of healthcare to improve access and outcomes among the disadvantaged clashes with the preferences or sensibilities of the middle class, legitimacy will be among the casualties. In this regard, Richard Glazier described the current crisis in Canadian primary healthcare, but also presented data showing some successes, particularly the small or absent SES gradients for many services, albeit with considerable room for improvement for others. But the reform of primary healthcare could also threaten legitimacy: under some models, the main beneficiaries are likely to be the disadvantaged, for whom episodic, conventional care from stand-alone medical clinics is insufficiently effective. To the extent that well-off people seem to be content with conventional care and are not demanding the socially oriented, comprehensive centres promoted by primary healthcare visionaries and population health experts, tension might arise. If they perceive that major primary healthcare restructuring reduces access to or quality of care they receive, their support will wane and, as Vega noted, allowing the rich to opt out of a health system is enormously risky because such withdrawal erodes support for the overall system. Ideally, the needs of well-off members of society will also be better met by the restructuring, in which case there will be no tensions to resolve. But often there will be trade-offs and reallocations - a simple example is locating clinics and other facilities closer to those in need and farther from where the well-off live.

Engaged citizen participation in policy-making and preference-sorting exercises is much in vogue these days, and the literature suggests that it can be fruitful and enlightening. Whatever its virtues, it is neither a proxy nor a substitute for democratic decision- making. Engaged and informed citizens are, by virtue of these very characteristics, atypical; their views and preferences may change during the deliberative process in response to group interaction and an increasingly sophisticated understanding of issues and options. The more engaged and informed they become, the less representative they are of the process of everyday opinion formation and expression. Enriched participation might generate wise and nuanced policy ideas but not a deeper understanding of what the public thinks, or finds acceptable, under the usual conditions of indifference and surface reflection. Healthcare in most industrialized countries is a public realm and, ultimately, subject to democratic forces; citizens can thus exert power without having passed a knowledge test. One could well imagine a citizen-participation process that would rank wait times rather low on the priority list once people had been fully apprised of a system's actual performance, the quality of care of the frail elderly, the state of mental health services and other considerations. Such a ranking would, however, be a lonely voice amid a constellation of interests that insist on making hip- and knee-surgery wait times every decision-maker's top priority.

Are improved payment and incentive systems pathways to increased equity, however defined? As Culyer pointed out, these would be ambitious expectations to thrust upon revised processes - financial incentives and targets can lead to "ridiculous distortions," formulas do not always adhere to cost structures and capitation and similarly granular funding approaches do not always capture fixed costs. As is well recognized, fee-for-service is not the best way to ensure that the emergency room in a small hospital in a remote community is financially viable. Indeed, teaching hospitals were not included in the Quebec models Levine presented and Ida Goodreau showed how the high cost of teaching hospitals affected the funding allocation to those BC health authorities in which they existed.

As Culyer and Vega noted, there is a need for government to set and monitor standards and to ensure that data are available. Professionalism is critical; the most we should expect is that the payment and incentive systems should get out of the way of the advance toward whatever notion of equity we seek to achieve. Here there are reasons for optimism, if only because existing financial incentives tend to discourage first-rate chronic disease management, optimal care of the frail elderly and comprehensive approaches to care for the disadvantaged. Removing the perverse incentives that affect provider behaviour is useful work, and anyone interested in marrying strategic purchasing to increased equity needs to address the matter.

Related to this concern are the mixed signals about payment-for-results vs payment-for-activity. In Canada, we preach the former and practise the latter. After exhorting practitioners to abandon fee-for-service and to spend their time on high-need patients, managers grumble about reduced productivity and demand shadow billing to track how the new-style clinicians stack up against the ostensibly obsolete practices of their fee-for-service counterparts. Increased rates of interventions and throughput are uncritically accepted as improvements in productivity, without consideration of whether the outcomes are worth the cost and whether new thresholds of intervention are defensible. Funders must decide where they stand on the issue and practitioners have a right to expect consistency of both message and policy, as well as clear standards and rules of accountability. In this regard, it is important to clarify whether we are interested in redistribution or just in targeting new money to particular activities. Redistribution presents major challenges to power relationships; new money is much less contentious (Kellow 1988).

Finally, even if there is a stable majority commitment to reduced disparities, there are no guarantees of success. European countries have adopted a population health perspective and some have created holistic frameworks for equity that focus on the social determinants of health as well as the reorganization of care (Department of Health 2003; Mackenbach and Bakker 2003). Yet in some countries disparities are widening, while others are holding the line; nowhere are there dramatic reductions in health inequality (Mackenbach and Stronks 2002; Mackenbach et al. 2003). Canadian research has also identified widening disparities (Brownell et al. 2003), despite the country's strong tradition of population health research and advocacy.

The etiology of disparities is complex and the extent to which they are amenable to public policy solutions is debatable. There have been dramatic reductions in some disparities such as infant mortality and certain communicable disease rates, where the interventions are relatively straightforward and baseline differences were very large. In some cases all ships rise with the tide (e.g., extending clean water to entire communities and effective immunization campaigns). It could be that the generally more egalitarian societies of northern Europe have achieved all of the easier wins and now confront seemingly intractable levels of health inequality. After all, as Marmot and Wilkinson (1999) have shown, there is a gradient of inequality that affects every social class. No one fully understands why upper-middle-class people are healthier than middle-class people, and few are concerned about that disparity. The gap widens as one approaches the bottom quintile of the SES ladder and spreads further toward its lowest rungs.

Thus, before we can strategically purchase equity, we must define what equity we wish to purchase and learn more about the capacity of various forms of capital to get the job done. Again, honesty about willingness to pay for reducing disparities would add a more pointed, if uncomfortable, element to the debate. To cite one example: economists have long maintained that a certain level of structural unemployment is required to make economies function well. That perspective guarantees that a significant group in society will be vulnerable to marginalization and poor health even if there is a strong social safety net. The organization of a great deal of human life involves hierarchy, which seems in itself to create a health gradient regardless of absolute levels of abundance or deprivation. If hierarchy is intrinsic to the human condition - hard-wired into our psyches and sense of worth - then so too, at least to some degree, is health inequality.

While all these complexities are real and intellectually interesting, they should not result in policy paralysis or seduce us into believing that nothing can be done. It is within our power to eliminate poverty even if we cannot make everyone content or provide opportunities for all to reach their full potential. We can make the healthcare system more needs oriented and responsive to the unhealthier end of the gradient (Health Disparities Task Group 2004). And we can certainly develop indicators that are sensitive to SES and other markers of inequality instead of reporting aggregate measures of performance that mask differential effects. Simply reconceptualizing the idea of performance might, over time, create greater awareness of and momentum for addressing the determinants of health, while still enabling recognition of the need to treat the small proportion of the population that at any given time urgently requires care.

Indeed, careful thought suggests that we might not be interested in equality at all. Presumably, an easy way to reduce inequality would be to encourage higher SES people to smoke, drink, eat poorly, drive recklessly and otherwise decrease their health outcomes to the lowest common denominator. Few so advocate. Similarly, should a focus on equity lead us to reject health promotion activities that are more likely to be adopted by those with higher levels of education and thereby increase health disparities? What we are interested in, we suggest, is improvement. Such improvement might well focus upon those sub-populations with the greatest scope for gains; however, the ultimate aim would appear to be improving health outcomes for all, as efficiently as possible.

About the Author

Raisa Deber, PhD
Professor, Department of Health Policy, Management and Evaluation,
University of Toronto

Steven Lewis
President, Access Consulting Ltd., Saskatoon
Adjunct Professor, Centre for Health and Policy Studies,
University of Calgary

References

Alameddine, M., A. Laporte, A. Baumann, L. O'Brien-Pallas and R. Deber. 2005. "'Stickiness' and 'Inflow' as Proxy Measures of the Relative Attractiveness of Various Sub-Sectors of Nursing Employment." Social Science and Medicine 63(9): 2310-19.

Alameddine, M., A. Laporte, A. Baumann, L. O'Brien-Pallas, R. Croxford, S. Wang, B. Milburn and R. Deber. 2006. "Where are Nurses Working? Employment Patterns by Sub-Sector in Ontario, Canada." Healthcare Policy 1(3): 65-86.

Belli, P.C. 2004. "The Impact of Resource Allocation and Purchasing Reforms on Equity." HNP Discussion Paper, Washington, DC: The International Bank for Reconstruction and Development, The World Bank. Retrieved February 16, 2007. < http://siteresources.worldbank.org/HEALTHNUTRITIONANDPOPULATION /Resources/281627-1095698140167/Chap7BelliFinal.pdf >

Berk, M.L. and A.C. Monheit. 2001. "The Concentration of Health Care Expenditures, Revisited." Health Affairs 20(2): 9-18.

Birch, S. and C. Donaldson. 2003. "Valuing the Benefits and Costs of Health Care Programmes: Where's the 'Extra' in Extra-Welfarism?" Social Science and Medicine 56: 1121-33.

Brownell, M., L. Lix, O. Ekuma, S. Derksen, S. De Haney, R. Bond, R. Fransoo, L. MacWilliam and J. Bodnarchuk. 2003. "Why is the Health Status of Some Manitobans not Improving? The Widening Gap in the Health Status of Manitobans." Winnipeg: Manitoba Centre for Health Policy. Retrieved January 25, 2007. < http://www.umanitoba.ca/centres/mchp/reports.htm >

Daniels, N. and J.E. Sabin. 1998. "Last Chance Therapies and Managed Care: Pluralism, Fair Procedures, and Legitimacy." Hastings Center Report 28(2): 27-41.

Deber, R. 2004. "Delivering Health Care Services: Public, Not-for-Profit, or Private?" In G.P. Marchildon, T. McIntosh and P.-G. Forest, eds., The Fiscal Sustainability of Health Care in Canada: Romanow Papers, Volume 1. Toronto: University of Toronto Press.

Deber, R. and V. Goel. 1990. "Using Explicit Decision Rules to Manage Issues of Justice, Risk, and Ethics in Decision Analysis: When is it not Rational to Maximize Expected Utility?" Medical Decision Making 10(3): 181-94.

Deber, R., E. Forget and L. Roos. 2004. "Medical Savings Accounts in a Universal System: Wishful Thinking Meets Evidence." Health Policy 70(1): 49-66.

Department of Health. 2003. "Tackling Health Inequalities: A Programme for Action." London: Author. Retrieved January 8, 2007. < http://www.dh.gov.uk/PublicationsAndStatistics/Publications /PublicationsPolicyAndGuidance/PublicationsPolicyAndGuidance Article/fs/en?CONTENTID=4008268&chk=Ad%2BpLD >

Donahue, J.D. 1989. The Privatization Decision: Public Ends, Private Means. New York: Basic Books.

Donaldson, C. 2004. Economics of Heath Care Financing: The Visible Hand. New York: Palgrave-Macmillan.

Drummond, M. and M. Sculpher. 2005. "Common Methodological Flaws in Economic Evaluations." Medical Care 43(7 Suppl): 5-14.

Evans, R.G. 2006. "From World War to Class War: The Rebound of the Rich." Healthcare Policy 2(1): 14-24.

Evans, R.G., M.L. Barer and T.R. Marmor. 1994. Why are Some People Healthy and Others Not: The Determinants of Health of Populations. New York: Walter de Gruyter.

Figueras, J., E. Jakubowski and R. Robinson, eds. 2005. Purchasing to Improve Health Systems Performance. New York: Open University Press.

Fisher, E.S., J.E. Wennberg, T.A. Stukel, J.S. Skinner, S.M. Sharp, J.L. Freeman and A.M. Gittlesohn. 2000. "Associations among Hospital Capacity, Utilization, and Mortality of U.S. Medicare Beneficiaries, Controlling for Sociodemographic Factors." Health Services Research 34(6): 1351-62.

Forget, E.L., R. Deber and L.L. Roos. 2002. "Medical Savings Accounts: Will They Reduce Costs?" Canadian Medical Association Journal 167(2): 143-47.

Gold, M.R., J.E. Siegel, L.B. Russell and M.C. Weinstein. 1996. Cost-Effectiveness in Health and Medicine. New York: Oxford University Press.Health Disparities Task Group and Federal/Provincial/Territorial Advisory Committee on Population Health and Health Security. 2004. "Reducing Health Disparities - Roles of the Health Sector: Recommended Policy Directions and Activities." Ottawa: Public Health Agency of Canada. Retrieved January 25, 2007. < http://www.phac-aspc.gc.ca/ph-sp/disparities/drpolicye.html >

Kanavos, P. and E. Mossialos. 1999. "International Comparisons of Health Care Expenditures: What We Know and What We Do Not Know." Journal of Health Services Research and Policy 4(2): 122-26.

Kellow, A.J. 1988. "Promoting Elegance in Policy Theory: Simplifying Lowi's Arenas of Power." Policy Studies Journal 16(4): 713-24.

Leon, D.A., G. Walt and L. Gilson. 2001. "International Perspectives on Health Inequalities and Policy." BMJ 322: 591-94.

Mackenbach, J.P. and M.J. Bakker. 2003. "Tackling Socioeconomic Inequalities in Health: Analysis of European Experiences." The Lancet 362: 1409-14.

Mackenbach, J.P., V. Bos, O. Anderson, M. Cardano, G. Costa, S. Harding, A. Reid, Ö. Hemström, T. Valkonen and A.E. Kunst. 2003. "Widening Socioeconomic Inequalities in Mortality in Six Western European Countries." International Journal of Epidemiology 32(5): 830-37.

Mackenbach, J.P. and K. Stronks. 2002. "A Strategy for Tackling Health Inequalities in the Netherlands." BMJ 325: 1029-32.

Marchildon, G.P., T. McIntosh and P.-G. Forest, eds. 2004. The Fiscal Sustainability of Health Care in Canada: Romanow Papers, Volume 1. Toronto: University of Toronto Press.

Marmot, M. and R.G. Wilkinson. 1999. Social Determinants of Health. New York: Oxford University Press.

Menzel, P., M.R. Gold, E. Nord, J.L. Pinto-Prades, J. Richardson and P. Ubel. 1999. "Toward a Broader View of Values in Cost-Effectiveness Analysis of Health." Hastings Center Report 29(3): 7-15.

Mossialos, E., A. Dixon, J. Figueras and J. Kutzin. 2002. Funding Health Care: Options for Europe. Buckingham: Open University Press.

Preker, A.S. and A. Harding. 2000. "The Economics of Public and Private Roles in Health Care: Insights from Institutional Economics and Organizational Theory." Washington, DC: Health, Nutrition and Population Discussion Paper, Human Development Network, World Bank. Retrieved February 16, 2007. < http://siteresources.worldbank.org/HEALTHNUTRITIONANDPOPULATION /Resources/281627-1095698140167/Preker-TheEconomicsOfPublic-whole.pdf >

Randall, G.E. and A.P. Williams. 2006. "Exploring Limits to Market-Based Reform: Managed Competition and Rehabilitation Home Care Services in Ontario." Social Science and Medicine 62(7): 1594-1604.

Retzlaff-Roberts, D., C.F. Chang and R.M. Rubin. 2004. "A Comparison of OECD Countries." Health Policy 69(1): 55-72.

Saltman, R.B. and O. Ferroussier-Davis. 2000. "The Concept of Stewardship in Health Policy." Bulletin of the World Health Organization 78(6): 732-39.

Saltman, R.B., R. Busse and E. Mossailos. 2002. Regulating Entrepreneurial Behaviour in European Healthcare Systems. Buckingham: Open University Press.

Schattschneider, E.E. 1964. The Semisovereign People: A Realist's View of Democracy in America. New York: Holt, Rinehart and Winston.

Simoens, S., M. Villeneuve and J. Hurst. 2005. "Tackling Nurse Shortages in OECD Countries." OECD Health Working Papers 19. Geneva: OECD. Retrieved February 16, 2007. < http://www.oecd.org/dataoecd/11/10/34571365.pdf >

Starfield, B. 2006. "State of the Art in Research on Equity in Health." Journal of Health Politics, Policy and Law 31(1): 11-31.

Stone, D.A. 1997. Policy Paradox: The Art of Political Decision Making. New York: Norton.

Wright, C.J., G.K. Chambers and Y. Robens-Paradise. 2002. "Evaluation of Indications for and Outcomes of Elective Surgery." Canadian Medical Association Journal 167(5): 461-66.

Footnotes

1 Our paper is based on the live proceedings of the two-day symposium. The contributions that speakers subsequently prepared for this issue of Healthcare Papers closely resemble but do not replicate the contents of their verbal remarks. Thus, there might be slight variations between our discussion here and the speakers' contributions to this collection.