In their contributions to this section, both Gregory Marchildon and Stefan Ackerby insist on the value of universal single-payer health insurance. Also, they are sanguine that the politicians and administrators responsible for managing Canadian and Swedish health systems will, over the coming years, make reasonably wise cost-benefit decisions for the system and that taxpayers will agree to pay the requisite taxes.

In terms of equity and efficiency, the case for a well managed single-payer health insurance system is strong. It overcomes a variety of market failures associated with private health insurance, and assures equitable access to a crucial service. But – and this is a major qualification – success ultimately rests on the quality of the managerial and political decisions undertaken. There are powerful pressures on government to avoid cost-effective decisions. We are less confident than Marchildon and Ackerby that good decisions will prevail. Hopefully they are right and we are wrong.

Trends in health care spending

In the context of population aging, simply extrapolating the status quo into the future will result in substantial increases in health care spending. In this article we project the magnitude of these shifts in spending trends, and look briefly at some of the politically controversial reforms that will probably be necessary to “save medicare.”

Canada is a relatively young country; it has a smaller over-64 population share than all but two other G8 countries. However, it is aging, and doing so at an accelerating rate (see figure 1). While the over-64 share increased by only three percentage points between 1990 and 2010, on the basis of our projection (details discussed below) it will rise another nine points by 2030. A second demographic trend will also have adverse fiscal implications: the population share in the active working age years (18–64) is shrinking. For this cohort, which pays the overwhelming majority of taxes, we project an eight-point decline in population share between 2010 and 2030.

In Canada, decisions governing public health spending are primarily made by the ten provincial governments – subject to the loose constraints imposed by the Canada Health Act and potentially by the courts. The federal government is the equivalent of a small province because it finances the health care of residents in the northern territories, on-reserve registered Indians and to some extent registered Indians living off-reserve. Ottawa finances a quarter of provincial health care spending through equalization and the Canada Health Transfer, while provinces finance three quarters through own-source taxation.1

The history of public health care budgeting over the last four decades can be divided into three periods (figure 2). The first, from introduction of provincial single-payer systems to the early 1990s, was characterized by fiscal drift. During this period neither Ottawa nor the provinces paid sufficient attention to fiscal discipline in the administration of their respective social programs. Politicians indulged in bouts of mutual recrimination. The provinces protested that Ottawa erected national standards but unilaterally restricted its share of costs as the federal debt worsened. Ottawa replied that the provinces were violating the spirit of cost-sharing agreements by irresponsibly expanding provincial social programs that could be cost-shared with Ottawa.

By the early 1990s, Canada had become one of the most seriously indebted OECD countries. There followed a half-decade of fiscal crisis. While public opinion was deeply divided over specific decisions, a majority favoured fiscal restraint. Federal and provincial governments that undertook this exercise were rewarded with subsequent reelection.

In a recent interview, Lucien Bouchard, premier of Quebec in the second half of the 1990s, was asked to name the political accomplishment of which he was proudest. His reply: “As premier of Quebec, my greatest accomplishment was elimination of the provincial deficit. It was incredibly difficult … We were approaching the abyss, with Quebec debt about to be relegated to junk bond status and the province paying ominously high interest rates.” Another question: What was your greatest error? Bouchard: “I often think of the program to retire nurses.”2

Bouchard’s replies nicely illustrate two points. Having underestimated the importance of fiscal discipline prior to the 1990s, provincial politicians discovered that preserving a single-payer health insurance system required “incredibly difficult” political decisions. Some of these decisions were eminently justifiable in terms of cost-effectiveness. Senior politicians and health administrators had long understood the rationale for closing many hospitals that were too small to provide sophisticated services or were located in communities with declining population. The hospitals had remained open as a result of local lobbying. The second point is that many decisions made in haste proved not to be cost-effective. Quebec was not alone in offering inducements to nurses to retire early, and subsequently regretting the loss of skilled health workers.

Among OECD countries, Canada stands out for the success of its fiscal rehabilitation in the 1990s. However, following the half-decade of fiscal crisis, Canadian provinces entered the third period, characterized by a return to fiscal drift. Perhaps the decline in spending increases in 2010 and 2011 implies a new attention to fiscal discipline, but it is too soon to know.

Others have undertaken similar projection exercises to ours.3 What we add to the projections of others is sensitivity analysis on “service intensity,” a variable that should become an important parameter in public discussion of health budgeting. In what follows, we disentangle the impact of changes in service intensity from the impact of an aging population on increases in the ratio of public health expenditure to GDP:

• Shift in age distribution of the population. We project the age distribution of the Canadian population under a set of conventional assumptions with respect to fertility, life expectancy and interprovincial and international immigration (see Detailed projection assumptions box for more detail). Aging of the population generates two effects, both serving to increase public health spending as a share of GDP. First, there is an expenditure effect arising from the larger population share of high-needs elderly patients. Second, there is a GDP effect because aging of the population reduces the share of the population in the active labour force cohort (ages 18–64). We assume real GDP per person in this cohort grows at a constant rate.
• Service intensity. A portion of change in per capita health care spending can be explained by population aging, but only a portion. The remainder, the change in per capita health care spending that would take place if the age distribution remained the same, we refer to as a change in service intensity. If per capita spending in each age cohort remains constant, then any change in per capita health spending reflects change in the population distribution, not in service intensity. Prior to the new millennium, population aging was of minor importance; hence changes in per capita spending were essentially due to changes in service intensity. Service intensity may rise for many reasons: introduction of new medical technologies, increase in compensation for health care providers at a rate above the assumed inflation rate, increase in ratio of health care workers per capita, expansion of the scope of health insurance and so on. We allow this parameter to vary.

Projection results

Figure 3 illustrates projected provincial health spending as a share of GDP, under the assumption that the average per capita change in real service intensity for the years 1998–2009 (2.3 per cent) persists to 2050. Provincial health spending rises from 7.7 per cent of GDP in 2012 to 21.8 per cent. If, hypothetically, the share of all age cohorts remained unchanged relative to the base year 2012 distribution, the increase would be from 7.7 per cent to only 13.0 per cent.

Sidebar: Detailed projection assumptions

Service intensity

Annual real rate of increase, 1998–2009: 2.26%

Health service inflation, average annual
1997–2011 change in current government expenditure implicit price index:       2.59%

Population projection

Total fertility rate:    1.5 children per woman

Life expectancy progressively
rises to reach by 2050:

for women:                                    87.5 years

for men:                                          82.5 years

Immigration: number and age-sex distribution relative to 1996–2011 average:  unchanged

Economic growth

Annual real increase in productivity
per capita among working-age cohort: 1.37%

It is highly unlikely that provincial governments over the next four decades will extrapolate recent rates of service intensity increase. If provinces reduce the rate by three quarters (to an annual increase of 0.6 per cent in real terms), provincial spending in the absence of population aging would decline. However, even under this severe constraint on rate of growth of service intensity, an aging distribution induces rising provincial spending relative to GDP: from 7.7 per cent of 2012 GDP to slightly under 10 per cent of 2030 GDP and slightly over 11 per cent of 2050 GDP.

From now to 2030 the demographic projections are near-certainties. The key parameter that politicians and officials must assess is service intensity: will politicians be able to reduce its rate of growth dramatically, and do so without the backdrop of fiscal crisis that faced Canada in the early 1990s? Maybe yes, maybe no.

From a peak in mid-1990s to 2012, Canada reduced its taxing effort by nearly seven percentage points of GDP.4 As a result, Canada fell from the middle of the second quartile among OECD member states, ranked by taxing effort, to near the bottom quartile. Over the last two decades, U.S. taxing effort remained within a much narrower range. Accordingly, Canada approximately halved the gap in taxing effort between the two countries. Given these trends, there may be a political willingness among provincial electorates to accommodate – somewhat – higher public health care spending by increasing their taxing effort.

Are Canadians willing to increase taxes? The World Values Survey affords evidence to the effect that Canadians – and residents of other majority Anglo-Saxon countries – are among the more adamant of those in OECD member states in not wanting government to “do more” than at present.5 In its 2005–08 round of the survey, respondents were asked to indicate the extent of their agreement or disagreement with the following propositions: “people should take more responsibility to provide for themselves” and “the government should take more responsibility to ensure that everyone is provided for.” Respondents could express complete agreement with the statement that people should take more responsibility to provide for themselves (scored at 10), compete agreement with the contrary position (scored at 1), or degrees of agreement with each by choosing a number between 1 and 10.

Among 17 OECD member states, average national scores ranged from 3.5 to 6.4 and were strongly correlated with average taxing effort in these countries over the decade 1999–2008. As would be expected, scores shifted toward “people taking more responsibility” at higher rates of taxing effort and toward the opposite proposition at lower rates of taxing effort. The Anglo-Saxon countries (United States, Britain, Australia, New Zealand, Canada) and Switzerland were outliers. At any given taxing effort, residents of these countries were more inclined to conclude that people should be taking more personal responsibility.

Arguably, the direction of causation should be inverted in the medium term: public opinion imposes a floor and ceiling on the ability of democratically elected governments either to tax and spend or to not tax and not spend. Canada’s average response is 6.0 on the ten-point scale. On the basis of a simple regression across the 17 countries, this response is 1.3 points higher than expected given its taxing effort – presumably because Canada belongs to the Anglo-Saxon subset. A three-percentage-point increase in taxing effort would increase Canada’s average score to the median among OECD countries and increase its hypothetical score in a future survey from its present value to 6.2, placing it close to the apparent ceiling of public tolerance for government “taking more responsibility.”6

Getting better value for money

Our first recommendation is modest: political and administrative elites should discuss aggregate public health spending in terms of an appropriate cap to place on growth of service intensity (as we have defined it) and be prepared to define budgets in terms of such a cap. There is no unambiguously optimal value for health care spending as share of GDP. Debating the allocation of public funds among competing uses lies at the heart of politics in any high-income society in which a universal health insurance system is inevitably the most important single spending envelope.

If we are looking for “good but unlikely to be adopted any time soon” reforms to contain health care costs, it is worth examining what other high-income countries are doing. Comparing Canada with Scandinavian countries or the United Kingdom is more meaningful than comparing it with the United States. Hence the value of Ackerby’s article. Given the importance of the British National Health Service (NHS) to the history of medicare in Canada and given that Canada and the U.K. are both reasonably populous culturally heterogeneous countries, the U.K. is a good initial comparitor.

The U.K. achieves broadly similar population health outcomes to Canada while spending considerably less as a share of GDP (11.4 per cent in Canada, 9.8 per cent in the U.K. in 2011). In terms of life expectancy at birth and containing tobacco addiction, Canada performs slightly better than the U.K.; in terms of infant mortality and potential years of life lost among those under age 70, Canada fares slightly worse. The U.K. and Canada maintain similar ratios of core professional personnel – doctors and nurses – relative to population.

The Canadian system is decentralized inasmuch as each province enjoys great autonomy in program design. At its birth in 1948, the National Health Service was run out of Whitehall as a hierarchical organization not unlike the army. Over the last quarter century U.K. governments, both Labour and Conservative, have decentralized the NHS by creating multiple “trusts,” defined in terms of region and health function. These are led by boards with broad discretion over major financial decisions pertaining to hospitals and remuneration of doctors and other NHS employees. These institutional reforms have been far more extensive – and far more controversial – than any undertaken by their Canadian counterparts.

Several institutional features of the NHS should be of interest to Canadian governments contemplating better value for money:

• Capitation as the dominant means of compensating primary care providers. Whereas a Canadian general practitioner is typically paid via fee-for-service, a GP in the U.K. receives a negotiated amount per patient on his or her roster, the amount varying by age, gender and other patient characteristics. Neither mode of compensation should be evaluated as a means to reduce GP incomes, and both generate certain perverse incentives among care providers. Among self-employed GPs, those in the U.K. realize an average income slightly higher, relative to the national average wage, than do their Canadian counterparts – 3.6 versus 3.1. An underappreciated virtue of capitation is the relative ease of introducing new health professions into delivery of primary health care. As Åke Blomqvist and Colin Busby note, “the bulk of activity in the health system has shifted away from acute care to the monitoring and treatment of chronic illnesses.”7 Clinical pharmacists and nurse practitioners can play a valuable role in cost-effective delivery of primary care. In the Canadian context, their employment in this context has been limited, in large part because of the difficulty of designing a suitable mechanism for their remuneration within a fee-for-service environment.
• Salary as the primary means of compensating specialist doctors. In the U.K. most specialists are employees of hospitals, their salaries negotiated by the relevant hospital trust. Hospital managers are obliged to include the cost of specialists among the costs of other necessary inputs in the course of budgeting. The result has been U.K. specialist incomes similar (relative to the respective average national wage) to most other OECD countries. Not so in Canada. In Canada, nearly all specialists are self-employed as are GPs. Canadian specialists have negotiated relative incomes much higher than salaried U.K. specialists, in part because the cost of specialists is not internal to hospital budgeting.8
• Extensive use of internal markets. Internal markets require that public agencies (trusts in the U.K. context) purchase services from other public agencies. Not without controversy, U.K. – and Scandinavian – governments have made extensive use of this principle. For example, primary care trusts pay a fraction of deemed hospital costs to the relevant hospital trust for patients hospitalized by the GPs working under the trust. The intent here is to provide an incentive to constrain GPs’ resort to hospitalization.

In conclusion, there is no guarantee that Canadians will elect governments, either in Ottawa or in the provinces, willing to experiment with institutional reform of public health administration on the scale that has taken place in the U.K. However, if we do not address the fiscal implications of an aging population in an orderly manner, the result will probably be a replay, at some point over the next decade, of a disorderly imposition of spending caps analogous to those imposed in the 1990s.

Notes

1 Equalization assures that “have not” provinces can generate a predefined threshold level of per capita revenue provided they undertake an average taxing effort on accessible tax bases. Equalization is an unconditional transfer but, since provinces devote nearly half their program spending to health care, it is fair to conclude nearly half of equalization transfers are devoted to provincial health budgets. Adding half of equalization to the explicit federal transfer targeted for provincial health expenditures implies that Ottawa finances approximately a quarter of total provincial spending on this file. For the 2012–13 fiscal year, this estimate of federal transfers was approximately $35 billion. See Canada, Economic Action Plan (Ottawa: Department of Finance, 2012), p. 243.

2 “Confidences d’un ex: entrevue avec Lucien Bouchard,” L’actualité, October 1, 2012, pp. 44–46.

3 Don Drummond, Therapy or Surgery? A Prescription for Canada’s Health System, Benefactor’s lecture (Toronto: C.D. Howe Institute, 2011); Pierre Fortin, Staying the Course: Quebec’s Fiscal Balance Challenge, Commentary 325 (Toronto: C.D. Howe Institute, 2011; Chris Ragan, Canada’s Looming Fiscal Squeeze (Ottawa: Macdonald-Laurier Institute, 2012); William Robson, The Glacier Grinds Closer: How Demographics Will Change Canada’s Fiscal Landscape, E-brief 106 (Toronto: C.D. Howe Institute, 2010).

4 Organisation for Economic Co-operation and Development, Economic Outlook (Paris: Author, 2012), retrieved October 10, 2012, from here

5 World Values Survey, WVS 2005–2008 fifth wave (2011), retrieved from here

6 The OLS results are as follows: y = 2.38 + 1.26×1 + 0.06×2, where y is the predicted national score, x1 takes the value 1 for Anglo-Saxon countries (U.S., Britain, Australia, New Zealand, Canada) and 0 elsewhere; x2 is average revenue / GDP over the decade 1999–2008. The adjusted R2 is 0.44; both regressors are significant at 0.025. See also John Richards, “Idle speculation: Prefunding the boomers’ ‘frail elderly’ care and legitimizing a carbon tax,” in Fred Gorbet and Andrew Sharpe, eds., New Directions for Intelligent Government in Canada: Papers in Honour of Ian Stewart (Ottawa: Canadian Centre for Study of Living Standards, 2011), pp. 227–48.

7 Åke Blomqvist and Colin Busby, How to Pay Family Doctors: Why “Pay per Patient” is Better than Fee for Service, Commentary no. 365 (Toronto: C.D. Howe Institute, 2012), p. 5.

8 Relative to average national wage, Canadian specialist incomes are third highest among 23 OECD member countries with available data (Organisation for Economic Co-operation and Development, Health at a Glance, 2011 , p. 67).