Canada: Selected Issues

This Selected Issues paper addresses some of the key policy and economic challenges facing the Canadian economy. The paper presents a new approach to predicting the business cycle in the context of the Canadian economy. This approach uses a range of parametric and nonparametric tests to gauge the ability of various indicators to predict turning points in the business cycle. The paper also presents a model that links the inflation rate to the business cycle and the rates of change in the exchange rate and in unit labor costs.

Abstract

This Selected Issues paper addresses some of the key policy and economic challenges facing the Canadian economy. The paper presents a new approach to predicting the business cycle in the context of the Canadian economy. This approach uses a range of parametric and nonparametric tests to gauge the ability of various indicators to predict turning points in the business cycle. The paper also presents a model that links the inflation rate to the business cycle and the rates of change in the exchange rate and in unit labor costs.

VII. The Canada Pension Plan: Problems and Policies 1/

1. Overview

In the next century, the Canada Pension Plan (CPP) will face severe fiscal pressures from changing demographics. 2/ Since the CPP’s inception in 1966, life expectancy has increased by 3.1 years and is expected to increase by another 1.5 years by 2030. Fertility rates have approximately halved and are not expected to show much increase in coming years. 3/ As a result, a substantial decline in the ratio of workers to pensioners (the support ratio) is anticipated. In addition, the enrichment of benefits and increased payments under its disability provisions since 1966 have put further pressure on the plan’s finances. 4/ These trends mean that even if contribution rates rise in the future as currently scheduled, the CPP’s assets will be exhausted in 2015. 5/ It is estimated that in order to maintain the current benefit structure, the contribution rate will have to increase from 5.6 percent in 1996 to 14.2 percent by 2030. 6/

The financial strains on the plan have been further aggravated by higher real interest rates and lower real wage growth than anticipated at the plan’s inception. Expectations that wages would grow rapidly and interest rates would remain low meant that it was reasonable to initially finance the CPP on a pay-as-you-go (PAYG) basis with a small asset base. Today, with lower growth in real wages and higher real interest rates than anticipated 30 years ago, the original strategy for the plan’s financing no longer appears sustainable. 7/

The future burden on a pension plan is summarized in its unfunded liability--the present value of the difference between expected future contributions and future outlays, net of current assets. 8/ The CPP’s unfunded liability is estimated to be 71 percent of GDP. Under the current schedule of rates, Canada’s Department of Finance projects that the unfunded liability will grow by about $50 billion per year. The actuarially-fair contribution rate is about 10-11 percent, almost twice the present rate. 1/

These calculations clearly indicate that some changes in the plan’s provisions will be needed. This paper simulates the effects of various reform options, including increasing workers’ contributions, reducing benefits, and raising the age of retirement. 2/ The simulations show that substantial changes in benefits or contributions will be needed to keep the plan from accumulating large debts.

2. How the CPP operates

The Canadian public system of old-age support (including the CPP, the Old-Age Supplement (OAS), the Guaranteed Income Supplement (GIS), and the Spouse Allowance Program (SAP)) is designed to ensure a modest level of income for the elderly. The CPP is an earnings-related plan, whose coverage is mandatory for income earners, including the self-employed. The OAS is a flat-rate, taxable monthly benefit with a clawback of benefits for those earning more than about 1 1/2 times the average industrial wage. The GIS is a means-tested benefit paid to single pensioners with income lower than about $11,000 or married couples with income lower than about $14,500, while the SAP is a benefit to needy widows and widowers. Private pension plans form the third tier of the pension system in Canada. Hence, like the pension systems of a number of other countries, the Canadian system combines elements of forced saving, social welfare, and incentives for private long-term savings.

Since the focus of this chapter is the CPP, some of its details warrant explanation. 2/ Under current policies, workers contribute a fraction of income (the contribution rate) between a maximum (the year’s maximum pensionable earnings or YMPE) and a basic exemption (presently about 10 percent of the YMPE). Benefits on retirement are a fraction (the nominal replacement rate) of average earnings below the YMPE, where the average disregards up to seven years of lowest earnings and periods when the worker is disabled or raising a child. 1/2/ The YMPE is at the level of the average wage in industry and is indexed to average earnings growth. After retirement, benefits are indexed to consumer price inflation. Presently, the contribution rate is 5.6 percent, the YMPE is about $35,000, and the replacement rate is 25 percent. The contribution rate and replacement rate are fairly low compared to those in other industrial countries. 3/

Finally, workers can retire between age 60 and age 70, but those retiring before age 65 face a reduction of benefits of 0.5 percent per month of retirement before age 65. Those retiring after age 65, by contrast, are rewarded with an additional 0.5 percent of benefits per extra month of work. 4/ The CPP also provides survivor’s pensions, disability benefits, and other types of benefits, but those aspects of the CPP are not covered in the simulations presented below.

As with many public pension schemes, the system’s finances are on a PAYG basis--today’s contributions are not ‘saved’ for future retirees, but are used to pay today’s retirees. The CPP maintains assets targeted at twice annual expenditures; these assets are currently about 2 1/2 times annual expenditures. The schedule of future contribution rates is adjusted under a mandated review of the CPP every five years to keep assets at about the target level. The most recent actuarial report shows the contribution rate rising by about 0.2 percentage points per year to peak at about 15 percent in 2030. 5/ In the baseline simulation presented below, however, we assume that the contribution rate remains constant at the current level. The baseline is thus somewhat more pessimistic than the CPP’s own status-quo projections.

3. Projected CPP finances under the baseline and alternatives

This section presents projections of the public pension system’s finances. 1/ Baseline projections are presented first, to give a view of the magnitude of the problem and provide a basis for comparison. The baseline projections in Table VII-1 shows the financial status of the public pension system for the years 2000-2050 under the assumption that the contribution rate remains constant at 5.6 percent, the replacement rate remains at 25 percent, benefits are fully indexed to the CPI, and retirement can be taken without penalty at 65 years of age. 2/ Under these assumptions, the CPP moves quickly into a net liability position. Net liabilities total 150 percent of GDP in 2050, more than double the current ratio of public debt to GDP. The worsening of the net asset position coincides with the decline in the support ratio from 4 workers per pensioner to 1.7 workers per pensioner.

Table VII-1.

Canada: Baseline Projections of CPP Finances

(In percent of GDP)

article image
Source: Fund staff estimates.

One option to stem the decline in the system’s finances would be to increase the contribution rate. The second half of Table VII-1 shows the finances of the plan under the assumption that the contribution rate is increased immediately from 5.6 percent to 12.7 percent, a rate that ensures that net assets are the same ratio of GDP at the end of the projection period as at the beginning (8.9 percent of GDP). The constant contribution rate that restores the net asset ratio at the end of the projection horizon is known as the balancing rate. Though the net asset position is positive in 2050, the plan shows increasing assets as a proportion of GDP until 2023, and a gradual depletion of reserves thereafter. This implies that the higher contribution rate would not be sufficient to keep the net asset position at 8.9 percent of GDP after 2050.

Another option would be to lower the replacement rate. 3/ Table VII-2 contains projections for the system under the assumption that the nominal replacement rate is lowered to 19 percent from the present level of 25 percent. 4/ As under the baseline, the net asset position deteriorates over time, though to a smaller liability position of 95 percent of GDP in 2050. The second half of the table shows that the balancing rate would fall to 10.2 percent, somewhat lower than under the baseline but still well over the current rate. Much larger reductions in benefits would be needed to keep the CPP from accumulating a large debt.

Table VII-2.

Canada: Replacement Rate Lowered from 25 Percent to 19 Percent for New Pensioners

(In percent of GDP)

article image
Source: Fund staff estimates.

The fact that CPI inflation overstates increases in the cost of living has led some to suggest only partial indexation of benefits for public pension plans. 1/ Table VII-3 shows the effect of indexation to 80 percent of consumer price inflation. Given the baseline assumption of inflation equal to 3 percent, the effect is to lower indexation by about 0.6 percentage point per annum, or roughly the amount of bias estimated in the CPI. 2/ Partial indexation reduces outlays by only about one-tenth of one percent of GDP, so that the pension system still shows a large net liability position in the last year (about 137 percent of GDP). As a result, this option would allow only a modest decrease in the balancing contribution rate relative to the baseline (to 12.1 percent from 12.7 percent).

Table VII-3.

Canada: Indexation of Pension Benefits Limited to 80 Percent of CPI Inflation

(In percent of GDP)

article image
Source: Fund staff estimates.

Another option currently under consideration is to increase the age for retirement with no penalty by two years (to 67). Table VII-4 shows the result of this change. The net asset position is still a large liability in the last year with no change in the contribution rate, and the balancing contribution rate is still high (11.4 percent) relative to the baseline. Hence, this measure would not ensure the plan’s viability. 3/

Table VII-4.

Canada: Retirement Age Increased to 67

(In percent of GDP)

article image
Source: Fund staff estimates.

4. Sensitivity to interest rates and productivity growth

The outlook for the pension system is sensitive to assumptions regarding the real interest rate (which affects the return on assets) and the growth rate of productivity (which affects both increases in the wage base for contributions and the growth rate of GDP). Table VII-5 shows the effects of lower productivity growth (1 percent per annum rather than 1.5 percent in the baseline). GDP growth is assumed to be linked to productivity growth, while benefits are indexed to the CPI, and so lower productivity growth increases expenditures as a percentage of GDP relative to the baseline. As a result, in the last year of the projection the net liability position is almost 30 percent of GDP larger than under the baseline. 4/ However, the balancing contribution rate is little changed (the rate increases less than 0.1 percentage point). GDP is smaller in the last year of the projection than under the baseline, so less of an increase in the level of net assets is needed to restore the asset position as a percentage of GDP.

Table VII-5.

Canada: Lower Productivity Gains (1 Percent Instead of 1.5 Percent)

(In percent of GDP)

article image
Source: Fund staff estimates.

Table VII-6 shows the effects of lower real interest rates (2 percent instead of 3.5 percent). The assumption of a lower real interest rate actually improves the financial situation of the plan, because it implies a lower cost of servicing the net liability position: With the assumption of a lower interest rate, the net liability position in 2050 is about 35 percentage points of GDP lower than under the baseline. The balancing contribution rate is higher than under the baseline (14.2 percent rather than 12.7 percent), however. To ensure that assets in 2050 are the same as in 1995, the plan must accumulate assets early in the projection period. A lower real interest rate means that the return on those accumulated assets contributes less to the overall balance, so that primary revenues must contribute more.

Table VII-6.

Canada: Lower Real Interest Rate (2 Percent Instead of 3.5 Percent)

(In percent of GDP)

article image
Source: Fund staff estimates.

5. Conclusions

The Canadian pension system faces serious challenges in the coming years. As with the pension systems of many other countries, these challenges largely stem from demographic shifts that will make it impossible for the system to maintain benefits at current contribution rates. If the present contribution rate of 5.6 percent were maintained, the system would move to a net liability position starting in 2013 and reach a net liability position of 150 percent of GDP in 2050. The simulations suggest that contribution rates would have to more than double to keep benefits at current levels. Maintaining the viability of the plan would require a large increase in contributions, a substantial cutback in benefits, or both.

APPENDIX Details of the Projections

The projections required macroeconomic and demographic assumptions. These are described below, and information on these assumptions is displayed in Table VII-7.

Table VII-7.

Canada: Macroeconomic and Demographic Assumptions under Baseline Scenario

(In percentage change, average over five years)

article image
Source: World Bank; Statistics Canada; and Fund staff projections.

Nominal wage growth is assumed equal to inflation plus productivity growth.

The projections employed macroeconomic assumptions about the labor market, growth, inflation, and real interest rates. Labor-force participation rates and unemployment rates for each gender and age cohort were assumed to remain the same as their current rates. They were applied to demographic projections (described below) to yield labor force and employment projections. Figures on the distribution of wage income by sex and age group were taken from Beach and Slotsve (1991), who provide data on 10-year age cohorts. Projections of growth in nominal wages from productivity growth and inflation were then used to generate average aggregate wage income, which in turn determined contributions and benefits. CPI inflation was projected at 3 percent per annum, with inflation as measured by the GDP deflator assumed equal to inflation as measured by the CPI. 1/ The growth rate of real GDP was determined by the growth rates of productivity (1.5 percent under the baseline) and the labor force. 2/ Real interest rates were projected at 3.5 percent, with nominal interest rates then determined by the projected inflation rate.

The demographic projections were taken from the World Bank’s World ‘ovulation Projections, 1994-95 Edition. These projections, by gender for 5-year cohorts, allow detailed modeling of the evolution of the population and labor force. As described above, in combination with labor-market assumptions they yield projections of the labor force and the retired population. Based on information provided in the most recent CPP actuarial report, half of the population is assumed to retire early at the age of 60, and half are assumed to retire at the penalty-free retirement age. The last line of Table VII-7 shows the increase in the elderly dependency ratio (defined as the number of persons age 65 or older divided by the population between 20 and 64 years of age) from 20.1 in 2000 to over 46 percent in the last year of the simulation.

References

  • Bayoumi, Tamim, Aging Population and Canadian Public Pension Plans,IMF Working Paper WP/94/89 (August 1994).

  • Beach, Charles M. and George A. Slotsve, “Polarization of Earnings in the Canadian Labour Market,” in Stabilization. Growth and Distribution: Linkages in the Knowledge Era, ed. by Thomas J. Courchene, John Deutsch Institute for the Study of Economic Policy (Kingston: 1991).

    • Search Google Scholar
    • Export Citation
  • Burbidge, John, Social Security in Canada: An Economic Appraisal, Canadian Tax Paper 79, Canadian Tax Foundation (Toronto: 1987).

  • Crawford, Allan, Measurement Biases in the Canadian CPI: A Technical Note,Bank of Canada Review (Summer 1993), pp. 2136.

  • Department of Finance, Canada, An Information Paper for Consultations on the Canada Pension Plan (Ottawa: February 1996).

  • Fortin, Pierre, “Do We Measure Inflation Correctly?”, in Zero Inflation: The Goal of Price Stability, ed. by Richard G. Lipsey, C.D. Howe Institute (Ottawa: March 1990).

    • Search Google Scholar
    • Export Citation
  • James, Steven, Chris Matier, Human Sakhnini, and Munir Sheik, The Economics of Canada Pension Plan Reforms,Working Paper 95-09, Department of Finance Canada (Ottawa: November 1995).

    • Search Google Scholar
    • Export Citation
  • Jobin, Guy A., Raymond Koskie, Patrick Longhurst, and Mark Zigler, eds., Employes Benefits in Canada, International Foundation of Employee Benefit Plans (Brookfield: 1991).

    • Search Google Scholar
    • Export Citation
  • Leibfritz, Willi, Deborah Roseveare, Douglas Fore, and Eckhard Wurzel, Ageing Populations, Pension Systems and Government Budgets: How Do They Affect Saving?”, Working Paper No. 156, Economics Department, Organization for Economic Cooperation and Development (Paris: 1995).

    • Search Google Scholar
    • Export Citation
  • Masson, Paul R., and Michael Mussa, Long-Term Tendencies in Budget Deficits and Debts,IMF Working Paper WP/95/128 (December 1995).

    • Search Google Scholar
    • Export Citation
  • Office of the Superintendent of Financial Institutions, Canada Pension Plan: Fifteenth Actuarial Report as at 31 December 1993, (Ottawa: 1995).

    • Search Google Scholar
    • Export Citation
  • Thirsk, Wayne, and Jeff Moore, The Social Cost of Canadian Labour Taxes,Canadian Tax Journal, Vol. 39 No. 3 (1991), pp. 55466.

    • Search Google Scholar
    • Export Citation
  • U.S. Social Security Administration, Social Security Programs Throughout the World-1995, Research Report No. 64 (Washington: July 1995).

    • Search Google Scholar
    • Export Citation
1/

Prepared by Etienne de Callatay and Charles Kramer.

2/

The CPP covers residents of all provinces but Quebec. Quebec has a separate but very similar pension plan (the Quebec Pension Plan or QPP). The simulations here cover the combined finances of the QPP and CPP; the same policies are assumed to apply to both of them.

5/

Ibid., p. 49.

6/

Ibid., p. ii.

7/

Ibid., p. 18.

8/

Since the Canada Pension Plan is operated on a PAYG basis, the calculation assumes that future contributions are made at the actuarially-fair rate. In this case, the actuarially fair rate is defined as the contribution rate that would equalize the present value of future contributions and benefits for individuals presently aged 18 years. For a more detailed explanation, see Office of the Superintendent of Financial Institutions (1995), p. 99.

1/

Ibid., Annex D.

2/

A number of other studies have discussed the effects of reforms on the financial health of the CPP. Leibfritz et al. (1995) find that the CPP balance is negative in net present value terms even with the nominal replacement rate 10 percentage points lower, the contribution rate 3 percentage points higher, or the retirement age 5 years later; without adjustments, they estimate that the unfunded liability is between 58 and 188 percent of GDP (depending on the discount rate employed). James et. al. (1995) find that even with contribution rates rising steeply to peak at about 15 percent in 2006, it would take 35 years to eliminate the unfunded liability. Bayoumi (1994) finds that contribution rates for the CPP would have to approximately double in order to maintain the current structure of benefits.

3/

Details on the other aspects of the plan and discussion are available in Jobin et al. (1991), Burbidge (1987), and Department of Finance Canada (1996).

1/

Note that the replacement rate as a fraction of earnings below the YMPE is not necessarily the same as the replacement rate as a fraction of total earnings. To distinguish the two, we refer to the former as the nominal replacement rate and the latter as the effective replacement rate.

2/

The stipulation to disregard certain periods in the working years is known as the dropout provision.

3/

For a comparison of the social security systems of a number of countries, see U.S. Social Security Administration (1995).

4/

Participants can retire after age 70, but can no longer make contributions after that age (see Office of the Superintendent of Financial Institutions (1995), p. 23).

1/

As mentioned earlier, the projections cover both the CPP and the QPP.

2/

Table VIII-1 also shows employment, the number of pensioners, the average pension, and average wage under the baseline. See the Appendix for details of the computations and assumptions underlying the simulations.

3/

Whether the adjustment burden ought to fall on workers (through a higher contribution rate) or on pensioners (through a lower replacement rate) is not obvious. However, some argue that the distortionary effects of payroll taxes and the already-high tax burden in large industrial countries stand in favor of lowering benefits (see Masson and Mussa (1995)); the deadweight loss of payroll taxes in Canada has been estimated at about 30-40 cents per dollar of revenue collected (see Thirsk and Moore (1991)).

4/

The policy change simulated here affects new pensioners only--present pensioners are assumed not to be affected.

2/

Crawford (1993) puts the bias at somewhat less than 0.5 percentage point; Fortin (1990) gives a range of 0.5-1 percentage point.

3/

Moreover, the effectiveness of such measures is limited by the fact that workers can retire early. In Canada, the proportion of workers retiring after 65 is negligible (see Office of the Superintendent of Financial Institutions (1995), pp. 53-5). Early retirement is factored into the projections, as half of workers are assumed to retire at age 60, and half at age 65. However, no increase in the fraction of workers who choose early retirement is assumed under the increase in the penalty-free retirement age.

4/

The analysis in SM/96/7, p. 28, finds similar effects of slower GDP growth on other public pension schemes that index benefits to the CPI.

1/

The results are inflation-neutral, except when limited indexation of benefits is simulated.

2/

It should be noted that with slowing growth in the labor force, this implies some slowing in real GDP growth over the long term from present levels. However, this is a natural consequence of anticipated demographic shifts. Indeed, the effect of such demographic shifts on growth is well-recognized and hardly limited to Canada (see for example “Is the Economy That Weak? First Factor in the Demographics,” Business Week, February 26, 1996, p. 26).