Canada’s public health care system is widely regarded as a model for equity of access. This is traditionally credited to a system characterized as having a single-tier structure, i.e., one offering the same degree of access, at least for acute care services (hospital and medical), to everyone regardless of ability to pay. In Canada, it is not possible to purchase supplementary insurance to facilitate alternative provisions of the services covered by the public insurance system. One result is that, in recent years, many Canadians have characterized the system as being resource constrained with major concerns about waiting times for elective surgery and other issues.

In June 2004, the Supreme Court of Canada heard a case brought by Dr. Jacques Chaoulli and others designed to test the prohibition on alternative financing of insured health services. This is, in effect, a test case for one-tier medicine or rather for the possibility of imposing two-tier options on the Canadian system. The justification for the case is that the current system does not provide sufficient access to health care.

Much of the discussion about parallel two-tier health care systems seeks to characterize the health care market as a normal market with the same responses for both supply and demand that might characterize any consumer product or service. The first segment of this article will review the market attributes of health care and identify the characteristics key to discussions about methods of health care financing and delivery. The next segment will examine the basic details of the Chaoulli case and the “Health Care Guarantee” proposed by the Kirby Committee as a solution to waiting times for services. The final segment will discuss the implications for the health care system of legal requirements for a Health Care Guarantee.

In order to evaluate options for funding health care, one must examine what makes the market for health care different. For example, it is important to note that health care is a derived demand. The goal of most persons is to achieve a certain health status: they are prepared to consume health care but would prefer to avoid it.

As well, the health care market is not one market for the consumer, but many: medical care, hospitals, diagnostic services, pharmaceuticals, longterm care, etc. All have different characteristics but most are subject to problems known as market failures.

The essential characteristic of the publicly insured health care market in Canada is that the doctors serve as the gatekeeper, mediating between the requirements of the patient and the capabilities of the system. The physician has a dual role as an agent for the patient as well as an entrepreneur delivering services. The patient comes to the system for services with no explicit knowledge of the associated price, value or even volume. The doctor assumes the role of patient’s agent to see that the patient receives the required care. The public insurance system covers the full costs without the patient or even the doctor being fully informed of all of the costs.

An efficient market requires that both parties to a transaction be fully aware of the behaviour of the other party. Obviously, that does not occur in many health care situations. Such asymmetries in information can give rise to market failure. These take a number of forms, both in the financing and delivery of health care.

Table 1 provides a summary of the major forms of market failure in health care financing.

In many countries, particularly those in Europe, there are often multiple sources of health insurance for individuals. With multiple sources of health care insurance, a form of market failure known as adverse selection arises because of the asymmetry of information between the purchaser and seller of insurance. If a person knows more about their health risks than a prospective insurer does, he or she will purchase a policy that is inefficient from the point of view of the insurance provider. This will result in higher rates for those that are good risks who will stop purchasing insurance or move to a cheaper program. Without intervention, the ultimate market result is limited insurance. In fact, the private insurance market, without regulation, provides very little incentive for firms to cover all risks. The result is that universal coverage is very difficult to achieve.

Cream skimming is a specific case in which insurance providers attempt to attract the low-risk persons and leave the higher-risk persons to other providers. Efficient private insurance markets will sort persons according to risk and likely leave many persons uninsurable at reasonable prices.

The issues of asymmetric information and moral hazard receive the most treatment in health economics literature. For example, if health insurance is involved, there may be incentives to over utilize the health care system or to avoid preventive care that might not be covered by insurance. This becomes complicated by the agency problem of the fee-forservice physician who benefits from delivering the extra care. In this case, moral hazard applies not to the person receiving the care but to the deliverer of the care. These problems exist with public or private insurance.

Private insurance schemes normally have risk-based premiums. Such systems deny insurance coverage to some by being prohibitively expensive. The incentives in private insurance are necessarily oriented toward risk minimization and profit maximization. In a recent article in the North American Actuarial Journal, “Designing a World-Class Health Care System,” Howard Bolnick describes the risk management tools of private health insurance.

• Bad risks may not be able to obtain insurance, making universal coverage impossible to attain (risk selection).

• Bad risks may pay more for their insurance than good risks do, making the sick pay more for their insurance than the well (risk rating).

• Bad risks may be treated differently at renewal or have restrictions on moving from one insurer to another, leading to charges of heavy-handed rating practices by insurers and problems such as “job lock,” where employees cannot move from one job to another without the threat of losing their health insurance (renewal risk management).

The essential conclusion from an analysis of health care funding is that private insurance for health care has a number of issues and will not provide universal coverage. This is true whether it is supplementary or primary insurance. Public insurance appears to be the most effective approach to achieving universal coverage.

Market failures in health care also occur in the provision of health care services, not just their financing. These failures are largely related to the relative market power of the participants. In a 1995 article “Abnormal economics in the health sector” in the journal Health Policy, W.C. Hsaio prepared a useful summary of key issues for the provision of clinical services (i.e. doctors and hospitals).

Proposed and current health care reforms focus on many aspects of this basic typology. For example, the introduction of clinics and nonphysician clinicians could alleviate some aspects of the service gaps. However, the dominant role played by physicians in management of acute care is likely to continue. Even the substitution of nonphysician clinicians does not really eliminate the control of the physician. This is particularly true if physicians are able retain strong control over pricing and management. Monopoly elements based on scale economies or professional control can exist in all areas of the health delivery and financing. Spatial monopolies can arise as well in health care because of the economies of scale. Health practices, hospitals, diagnostic facilities and other health care providers require a certain minimum scale to function with any degree of efficiency because of capital and even labour requirements: there may be only sufficient local market for one hospital or one medical clinic. This “lumpiness” of supply may create a monopoly in a specific region.

The agency problems associated with the health care suppliers, such as doctors, are important aspects of asymmetric information. In their role as an agent on behalf of the patient and on behalf of the payer, the health care providers are in a very special position. Reflecting their quasi or true monopoly power, the health service providers have significant control over both their pricing (in the absence of public regulation) and the intensity of supply of their services at the individual and collective levels. As agents for the patient, they control the treatment modality and cost. In addition, health care suppliers, through professional control, have substantial or monopoly control over pricing and terms of service. The existence of health insurance provides an incentive to the health care supplier to over-supply health services.

In the actual provision of services, there is a strong interaction between the moral hazard issues of insurance and the agency problems of the health care suppliers. Compensation on the basis of fee for service for health care providers, coupled with the moral hazard issues for health care consumers, can result in excessive or inappropriate utilization for some procedures, notably diagnostic ones. One of the main concerns often voiced about health delivery is the absence of clear pricing decisions to provide the appropriate allocation of resources and services. Reforms of service-based pricing for hospital-based services are being tried or proposed in many jurisdictions. The Kirby Committee proposed such reforms for hospitals in Canada. However, in a monopolistic supply situation, prospective payment does not eliminate issues of creamskimming, quality control and other market failures.

Compensation on the basis of prospective fees can result in the “up-coding” of conditions or requirements in order to generate more revenue. Several studies have identified this as a significant issue in the US Medicare system after the introduction of service-based compensation.

In a recent essay entitled “Defining the Sharing Community: The Federal Role in Health Care,” published by the IRPP in Money, Politics and Health Care: Reconstructing the Federal Partnership, Keith Banting and Robin Boadway discuss the market failures that are present in many aspects of health care. As types of market failure, they list free-rider issues, externalities, increasing returns to scale, asymmetric information, coordination problems and transaction costs.

Banting and Boadway also draw attention to the problems of administrative inefficiencies that can arise from competition because of additional costs such as advertising and marketing. This applies both to the private financing and delivery aspects of health care. There is also the issue of forgone scale economies and efficiency losses in the insurance system. Competing private insurance systems may not be able to achieve appropriate scale to achieve claims processing and other infrastructure efficiencies as well as economies in purchasing of services. Private systems inevitably require higher expense ratios to cover revenue collection and generation activities. A recent study in the New England Journal of Medicine, “Costs of Health Care Administration in the United States and Canada” by Stephanie Woolhandler and colleagues, indicated that health care administration costs were about three times higher in the US on a per capita basis than in Canada.

The efficiency and economies-ofscale issues provide the strongest arguments for public participation in health care markets as a single payer. A singlepayer system can potentially manage cost control issues better than other more fragmented systems. The monopolistic elements of the health care system are balanced by the monopsonistic control of the single-payer environment.

Controlling health costs remains a central concern in all countries. It is absolutely not clear that competitive market outcomes are achievable or can produce the desired results with respect to efficiency and equity. A recent report, “Are Market Forces Strong Enough to Deliver Efficient Health Care Systems? Confidence Is Waning” in Health Affairs, by Nichols et al., found a significant pessimism about the ability of private market forces to improve the efficiency and outcomes of the US health system. Specifically, concerns were raised with respect to:

• Provider market power

• Absence of potentially efficient provider systems such as integrated delivery systems serving an enrolled population

• Absence of employer-based incentives to efficiency such as definedcontribution health care programs

• Inefficient health plan competition because of high fixed costs and barriers to entry

The availability of health labour resources appears to be a concern in many countries. Constrained resource availability can result in higher labour costs. Competition, particularly if inefficient or ineffective, does not necessarily result in appropriate resource supply or allocation. In the US, a 2001 Health Affairs study, “Economic and Demographic Trends Signal an Impending Physician Shortage,” by R.A. Cooper and colleagues, highlights an impending shortfall in physicians in that market with consequent price and supply issues. A 2003 study in the same journal, “The future of the nurse shortage: Will Wage Increases Close the Gap,” by Joan Spetz and Ruth Given, forecasts significant wage increases for nurses in the US because of diminished supply.

Well functioning markets promote efficiency and maximize the gains from trade, one of the major goals of economics. Unfortunately, the health care market fails all tests of this kind.

In an 1999 issue of the Journal of Health, Politics, Policy and Law devoted to criticisms of the managed care revolution in the US, Thomas Rice introduced his discussion, “The Microregulation of the Health Care Marketplace,” of the complex regulations in managed care with a quotation from G.B. Shaw’s preface to his play “Doctor’s Dilemma”:

The most eloquent criticism of fee-for-service medicine is not attributable to Alain Enthoven but rather to Bernard Shaw… who wrote, “That any sane nation, being observed that you could provide for the supply of bread by giving bakers a pecuniary interest in baking for you, should go on to give a surgeon a pecuniary interest in cutting off your leg, is enough to make one despair of political humanity.”

Shaw goes on to define the incentives for an unnecessary operation to raise the incomes of the physician, an example of supplier-induced demand.

It is customary in discussions of the relative merits of two-tiered health care to cast health care as if it was equivalent to any other product or service. In fact, some analysts suggest that the delivery of health care should be no different than the delivery of bread. Yet, the market for bread is relatively straightforward with none of the market failures discussed above.

First, it is easier to create a standardized unit of bread and to deliver it to uniform standards of quality than to do the same for health care. The resources associated with delivering a particular health service or procedure will vary substantially with the complexities of the case. In particular, co-morbidities, other health indications of the patient, may complicate the treatment for the primary indication. Standard estimates of the impact of the complexity of a case are used to drive some prospective payment systems (PPS). However, it is much easier to estimate the cost of producing a unit of bread than of doing even a simple surgical procedure.

In the market for food at the local store, the agency problem doesn’t exist and within limits the asymmetric information issue does not arise. The customer decides how much bread he or she wants, not the baker. There is sufficient competition from the bakers that an adequate supply of bread (even imported bread) is available with reasonable documentation on the label as to its content and nutrition. Since the customer is paying for the bread directly, there is no moral hazard for either the consumer or the baker. There is no state agency paying for the bread. The customer is not tempted to ask for more bread than needed because of the direct payment. The baker is not tempted to suggest to the state agency that the customer needs more bread than necessary or a more expensive type of bread. More importantly, the bakers are not allowed to collude to set prices or restrict access to the baking ovens or other forms of control. New bakers have entry to market unless there is a guild that attempts to restrict access. But also, the bakers are not dependent on the single-payer for the flour to make the bread or to build the bake ovens.

There is usually a queue at the bakery in peak hours. It is usually considered appropriate for persons to wait their turn to buy the bread. All are considered to have the same need for bread. The queue is managed on a first-in-first-out (FIFO) basis. Usually, in most bakeries, a rich person is not permitted to jump to the head of the queue. In the health system, as will be discussed below, need, or severity of health condition, varies by individual and varies over time. This requires more complex queue management both for equity and clinical efficiency.

The bread market attributes can be explicitly contrasted to health care. The agency problems and issues of asymmetric information and moral hazard were discussed above. More importantly, the supply constraints imposed by significant barriers to entry and major elements of monopoly control obviously differentiate the production of bread from the acute health care system. The arguments for two-tier medicine explicitly emphasize the ordering of queues on the basis of ability to pay, a concept that is not accepted at most bakeries. The barriers to entry are much stronger in the health care sector than in the bakery industry. The management of entry to the health care field is sufficiently controlled so that it is difficult to see major supply enhancements coming through higher prices.

There is very little doubt that Canadians are concerned about access to the health care system. There are broad concerns about waiting lists for access to primary and acute care services. There are also more far-reaching concerns about insured equitable access to a full range of services such as home care, pharmacare etc. However, the reports of both the Kirby Committee and Romanow Commission indicated that waiting lists and times were a major concern. While there are serious issues of measurement for waiting lists for elective procedures, survey evidence indicates increasing public concern for and impact of waiting lists. One of the major recommendations of the Kirby Senate Committee, in their final volume The Health of Canadians— The Federal Role, Final Report on the State of the Health Care System in Canada, Volume Six: Recommendations for Reform. was to implement a health care guarantee:

When this maximum time is reached, the insurer (government) pay for the patient to seek the procedure or treatment immediately in another jurisdiction, including, if necessary, another country (e.g., the United States). This is called the Health Care Guarantee.

The Kirby Committee suggests that the alternative is to allow the development of a parallel system of supplementary insurance to purchase additional access. These access provisions imply either simply a re-ordering of the queues (with consequent ethical issues) or a separate tier of supplementary health care supply available only to those willing to pay higher prices. The committee report cites the existence of supplementary access in a number of European countries as an indication that two-tier systems can exist without the failure of the public tier.

This issue is about to be tested in the Supreme Court of Canada. The Chaoulli case, initially tried in Quebec, seeks to allow someone to purchase covered services outside of the standard system to achieve enhanced timely access. Simply put, they want to have the option of private funding of health care services. These services would be available separately from a default public health care system. The legal arguments are based on a Charter challenge under section 7 that security of the person is compromised by the lack of alternative services when waiting lists exist. The court accepted the argument that a potential section 7 violation of life, liberty and security existed. However, initially, the claim by Chaoulli was rejected on the basis of fundamental justice because the existence of private health insurance would compromise the integrity and viability of the public health insurance system. The case is being appealed to the Supreme Court on the grounds that the Quebec courts made a number of errors in law. In an article in the Fraser Forum in 2002, “A Court Challenge Against Government Monopoly in Health Care,” Edwin Coffey outlined these claims of error:

These include a misinterpretation of section 7 of the Charter relating to its protection of pure versus incidental economic rights, the latter having been infringed by section 15 of the Quebec Health Insurance Act and section 11 of the Quebec Hospital Insurance Act according to the first court’s judgment. Moreover, differing interpretations of the level of threat required (real, potential, or imminent) to qualify as a section 7 infringement of the right to life, liberty and security were not taken into account in the lower court judgment.

Michael Kirby and nine other senators are appearing as interveners in an appeal of case to allow supplementary health insurance and services. The general line of discussion is that persons should be able to obtain access to health care with the same flexibility of choice that they purchase movie tickets or loaves of bread. Legal counsels for the project are Stanley Hartt and Patrick Monahan, who laid out their original arguments that restrictions on private health services violated the Charter in a C.D. Howe study in 2002 entitled “The Charter and Health Care: Guaranteeing Timely Access to Health Care for Canadians.” In presentations on the topic, Professor Monahan defined the concept of a health guarantee as being able to purchase bread elsewhere if the government didn’t bake enough for a person soon enough.

In general, waiting lists are presumed to be the result of some resource allocation issue. It is generally considered relatively inefficient to have sufficient idle resources that waiting times are minimal for systems with random patient arrivals with random requirements. The cost of the idle capacity is considered an inappropriate use of resources. Lowering the wait time requires increased idle resources as well as some policies to limit access to the system. The latter is accomplished by changing the clinical guidelines for patient intake. One approach, such as adopted in Sweden, is to allow patients to choose an alternative hospital in a different district that has capacity. Interestingly enough, according to studies reported by the OECD, persons chose to remain with their local hospital and accept the consequent waiting time. Initially, the guarantees appeared to provide strong stimulus for efficiency moves. In a 2002 OECD review of health reforms in a number of countries, “Improving the Performance of Health Care Systems: From Measures to Action (A Review of Experiences in Four OECD Countries),” the author, Zeynep Or, noted that the capacity of waiting-time guarantees in providing long-term solutions to deal with problems of excess demand is limited.

In a 2003 OECD study by Luigi Siciliani and Jeremy Hurst, “Explaining Waiting Times Variations for Elective Surgery across OECD Countries,” the major conclusion was that the availability of doctors has the most significant negative association with waiting times. Canada has generally had a lower number of physicians than many other countries. According to OECD data, there were 2.1 physicians per 1000 population compared to an OECD average of 2.9.

Supplementary insurance effectively creates a second market for physicians’ services. Physicians can improve their income by maintaining waiting lists that encourage more affluent patients to jump the queue. This was identified as a significant concern in the UK by Sanmartin et al. in a study on waiting lists, “Waiting for Medical Services in Canada: Lots of Heat, but Little Light,” published in the Canadian Medical Association Journal in 2000.

More importantly, there is a strong incentive for the physician to offer more services in the private market than in the regulated market. According to a 1999 article, “A Tale of Two Bounties: The Impact of Competing Fees on Physician Behavior,” in the Journal of Health, Politics, Policy and Law by Thomas Rice and colleagues, this appears to have been the outcome when Medicare fees were reduced in the US. The result was a shift in emphasis by the physicians to services covered by private insurance. In an article to be published in the same journal, “How Does Private Finance Affect Public Health Care Systems? Marshalling the Evidence from OECD Nations,” Caroline Tuohy and colleagues came to the conclusion that impacts of a parallel system on resources or incentives are such as to increase pressure on the public system and to raise public-sector waiting lists and times.

It is likely that competition for resources in the private system would lead to further increases. The issue would be whether additional resources would be available in terms of new facilities and more importantly a larger supply of doctors. It is unlikely that the popular support would be available for increasing sources to the health care system without serious prospects for improvement in access rather than just increased compensation to providers. One of the challenges for a health care guarantee is the existence of the spatial monopolies and barriers to entry in health care delivery. This provides significant market power to the existing health care providers that could be more easily exploited in a multi-payer market place. As discussed above, a large single-payer insurance system provides a counterweight to provider market power. Smaller competing insurance systems would be less able to constrain provider costs.

A queue is a standard feature of many systems, particularly those with random arrivals of queue entrants. Patients present themselves to the system at random times with random severities and complexities of illness or injury. In the absence of infinite capacity to process the queue of patients, a queue of finite length is a necessary feature. The waiting time in the queue is a shadow price that helps to determine the pace at which persons enter the queue and hence impacts the length of the queue. As the wait times get shorter, the number of persons in the queue, that is, the length of the queue, may actually increase. In particular, waiting lists serve as a form of rationing for elective procedures with wait times as a shadow price.

Allocation of health care resources on the basis of need is an important principle of clinical effectiveness. Triage is the term used for the sorting process in queue management. It is actually a term appropriated from the French expression for a railway marshalling yard where rail cars are put into the right order for the train. This implies that a health system queue can never be treated like a first-in first-out queue. Queues are used to manage entry points in the system. Patients will be allocated positions in the queue appropriate to their health status and available system resources. Specific guarantees of treatment within a specific time would imply that persons with lesser conditions might be treated in an inappropriate order to meet the constraints of the guarantee. In another OECD study released in 2003, “Tackling Excessive Waiting Times for Elective Surgery: A Comparison of Policies in Twelve OECD Countries,” Jeremy Hurst and Luigi Siciliani cite a number of countries (Denmark, Norway, Sweden) that have abandoned or seriously modified guarantee programs. They also cite a British study that indicated that 20 percent of specialists noted having to treat patients out of their clinical order to respect guarantee times.

Effective queue management is an important issue. It’s about clinical effectiveness and disease management. The clinical picture for an individual queue member will change over time and should lead to an adjustment in their position in the queue. There are several major projects in specific areas of the health system in Canada to improve queue management. Examples include the Cardiac Care Network (CCN) in Ontario and the Manitoba Cataract Waiting List Project. Such techniques will not eliminate the queue. Changes in health care processes and improvements in the management of key diseases such as breast cancer can improve the impact of waiting lists on patient health and satisfaction significantly.

One of the major virtues of service-based funding is that it provides an incentive to treat and, with prospective funding, shifts some of the risk of service costs to the provider. For example, in the US the prospective system of payment is used for Medicare. Under this system, the payment for a specific service is fixed in advance for a specific Diagnostic Related Group (DRGs) on the basis of the complexities of the patient’s case. Prospective determination of payment can provide a strong incentive to the provider to cut costs and possibly service. Of course, without macro budgets, there will be an incentive to increase utilization to raise provider revenue. There may be an incentive to increase the services but also increase the diagnosed complexity of the patient’s case. If the same pool of health care providers operates in the public and private spheres, there will be an obvious incentive to shift activity toward the more remunerative sector. The OECD study cited above documents various forms of restrictions for such activity tried in some countries. For example, in England, specialists with full-time National Health Services contracts were limited to 10 percent private activity. In Finland, specialists working in publicly funded hospitals are not allowed to see private patients (for a specialist visit) in the same hospital. Private patients are treated outside of normal working hours. In Ireland, it is proposed that newly qualified specialists be constrained to work in the public system.

A fragmented insurance market, such as created by a parallel system, can result in cost increases in the health system because of market failures. In a study of European experience, “Private Health Insurance and Medical Savings Accounts: Theory and Experience,” published in 2002 in the book Funding Health Care: Options for Europe, Alan Maynard and Anna Dixon note that the existence of private insurance providers will fragment the market and increase market power of the health care providers with respect to the funders. Because their costs are passed on to the consumer, private insurance providers do not necessarily have the same incentives for cost control that a public system might. In an OECD study of the Australian market, “Private Health Insurance in Australia: A Case Study,” Francesca Columbo and Nicole Tapay note that the private health insurance (PHI) providers do not seem to exert the same cost controls as public health insurance because of a lack of incentives.

Cost control could be a critical issue in the Canadian context because the professional associations, who negotiate fees with the public insurance system, would be in a much stronger bargaining position if they could offer higher priced services in the private supplementary market.

Because of resource limitations, it is not clear that simply increasing financial resources will necessarily reduce waiting queues, either for those waiting for treatment or for those in the treatment process. Additional capacity will likely generate additional demand, since the existing waiting times have served as a deterrent for some of those seeking elective surgery. Increased resources may simply lower the shadow price of queue membership, encouraging additional persons to request treatment. Expanding capacity rather than just increasing compensation will be a necessity to reduce waiting times.

One of the major issues with voluntary insurance is that it traditionally is available to the more affluent segment of society. Thus, any tax subsidies provided by making the insurance payments deductible will also accrue to the affluent and may be significantly regressive. Maynard and Dixon note that supplementary insurance, in the European context, has largely been taken up by the affluent. In other words, it is the affluent that access the extra health system resources.

The issues with respect to health system, capacity does not guarantees and two-tier medicine are relatively simple.

• The existence of a second tranche of income available to a restricted supply of health care resources will inevitably bid up the cost of those resources. In the short run, supply will not be able to expand. Health system resources allocated to the parallel system will receive higher compensation. Inevitably, the public costs will rise to match that compensation or resources will shift. There will be the inevitable issues of agency conflicts with respect to resource management. In other words, there will be an incentive for physicians to optimize their resources toward the more remunerative private system. This would result in lower resources being available to the public system. Unless additional resources can be attracted to the increase. Waiting lists and times in the public system would likely increase.

• Health care guarantees may result in clinical issues in the management of existing resources. Significant absolute guarantees about treatment will likely result in a requirement to treat patients out of optimal clinical order.

• Supplementary health insurance will be of most benefit to the more affluent members of society who can afford it. At best (and possibly worst), it facilitates queue jumping based on the size of pocketbook, not need.