Canada: Economic Developments and Policies

This paper examines economic developments and policies in Canada during 1990–95. Spurred by the robust growth in the United States and the easing of monetary conditions between 1991 and 1993, economic growth in Canada continued to strengthen during 1994. Real GDP grew by 4.5 percent in 1994 after growing by 2.2 percent in 1993 and 0.6 percent in 1992. Economic growth in 1994 was led by exports and investment in machinery and equipment. However, growth was more broadly based in 1994; private consumption strengthened, and there was a rebound in residential and nonresidential construction.

Abstract

This paper examines economic developments and policies in Canada during 1990–95. Spurred by the robust growth in the United States and the easing of monetary conditions between 1991 and 1993, economic growth in Canada continued to strengthen during 1994. Real GDP grew by 4.5 percent in 1994 after growing by 2.2 percent in 1993 and 0.6 percent in 1992. Economic growth in 1994 was led by exports and investment in machinery and equipment. However, growth was more broadly based in 1994; private consumption strengthened, and there was a rebound in residential and nonresidential construction.

IX. Unemployment Insurance in Canada 1/

1. History and main program features

The Canadian unemployment insurance (UI) program, as originally defined by the 1940 Unemployment Insurance Act, was intended to act as an insurance program to provide temporary income replacement during periods of unemployment. In order that insurance principles would reasonably apply, coverage was not given initially to individuals who expected to experience recurrent or predictable unemployment or to those with unstable jobs. Exclusions included those workers employed in agriculture and forestry, fishing, transportation, teaching, health and government services, and any part-time or seasonal work. These exclusions meant that only about 40 percent of the labor force was covered under the original program, primarily in industry and commerce.

During the 1950s and 1960s, UI increasingly came to be viewed as an important part of the social safety net, and changes were made toward providing more universal coverage and greater redistribution of income from low-risk to high-risk workers. Key features that were added included supplementary benefits for people who were unable to qualify for or who had exhausted regular benefits, and a special program for self-employed fishers. 2/ In 1971 UI was expanded further to provide coverage to most of the labor force (about 90 percent in 1976); regionally-extended benefits, which link the number of weeks of benefits to the regional unemployment rate, were also introduced. At the same time, eligibility requirements were reduced significantly, benefits were raised, and separate sickness and maternity benefits were introduced.

In response to the escalating costs of UI during the past two decades, recent modifications to the program have partially reversed some of the changes introduced by the 1971 legislation. In particular, the benefit rate has been reduced and entrance requirements have been increased. 3/ At the same time, greater emphasis has been given to more “active” labor market policies, such as job search assistance and training. Despite these changes, payments for regular UI benefits remain the largest program expense; special benefits payments and expenses for employment services (developmental uses) each represent only 10 percent of total benefit payments (see tabulation below).

Canadian Unemployment Insurance Account 1/

(In billions of dollars)

article image

The UI program is self-financing on the basis of employer and employee premiums. In 1994, employees paid $3.07 per $100 of earnings up to a maximum insured weekly pay of $780; employers paid 1.4 times that amount. Although the account can run cumulative surpluses and deficits during the business cycle, premiums are required to be set so that the account balances over a relatively short period of time. 2/ This requirement has meant that premiums tend to be adjusted in a procyclical manner. Moveover, the rise in the average unemployment rate over time has raised the required UI premium rate from 1.35-1.50 percent in the late 1970s, to as high as 2.35 percent in the 1980s, and to 3.07 percent in 1994. 3/ As a result, UI premiums have increased in importance as a source of government revenue; by the early 1990s, UI premiums had become the second largest revenue source, after personal income tax revenue.

2. Trends in the labor market and UI usage

Certain features of the UI program, including generous benefits (for long-term, repeat users as well as occasional claimants), regionally-extended benefits, and high premium rates, are thought to contribute to inappropriate use of the UI system and/or unintended labor market consequences.

From a macroeconomic perspective, the natural rate of unemployment is estimated to have increased significantly during the 1970-80s, and some studies have attributed part of the increase to the greater generosity of UI benefits. The generosity of unemployment insurance is thought to affect labor supply and demand decisions in a number of ways. First, longer entitlement benefit periods are thought to increase the duration of unemployment by reducing the intensity with which the unemployed search for work. Second, unemployment insurance benefits are thought to increase a worker’s reservation wage--the wage at which a job offer will be accepted.

In fact, the rise in average unemployment during the past two decades has been associated with an increase in the average duration of unemployment. Although the incidence of unemployment has risen and fallen with the business cycle, the average duration is almost twice as long as in the mid-1970s. 1/ On the demand side of the labor market, the absence of experience-rated UI employer premia in Canada is thought to encourage firms to rely on temporary layoffs in the face of shocks, instead of adjusting hours worked or wages or taking measures to increase efficiency or to diversify. 2/

Corak (1994) provides a review of the literature on the effects of the Canadian UI program on the aggregate unemployment rate. Recent empirical studies relating UI generosity to the unemployment rate yield a wide range of estimates. Several studies find no effect on unemployment, while several others conclude that the increase in UI generosity in 1971 added between 1 and 3 1/2 percentage points to the unemployment rate. The absence of consistent empirical findings may reflect the fact that labor supply decisions by some individuals may not translate into aggregate effects when the labor market is characterized by excess supply. Under such conditions, the decision by some UI recipients to remain unemployed longer in the face of increased generosity of benefits would in turn increase the potential job offers available to other unemployed individuals, allowing them to experience shorter unemployment spells.

Rising UI premiums, which raise the cost of labor, are a second possible explanation for the increase in Canada’s natural rate of unemployment. The OECD (1994) describes some recent empirical work that would imply a long-run effect on the unemployment rate in Canada in response to a change in payroll taxes, but concedes that other studies have found only short-run effects on unemployment. 1/ Even if one concludes that payroll taxes have no long-run effect on unemployment, the estimated short-run effects of tax changes on unemployment would argue for modifying UI financing requirements in order to avoid premium rate increases during economic downturns.

One need not find evidence of significant macroeconomic effects to conclude that the UI program creates important disincentives in the labor market. In a review of the microeconomic literature, Corak (1994) cites studies that show that the duration of unemployment spells depend positively on the UI benefit rate (for females), the UI benefit entitlement (or potential duration of benefits), and the aggregate rate of unemployment (for males). There is also evidence that employment spells are influenced by UI program design. Specifically, in regions with high unemployment, the length of some employment spells appears to have been shortened to closely correspond to UI eligibility rules (minimum weeks of employment required to become eligible for UI). Finally, empirical studies have shown that the probability of making repeat UI claims is a function of an individual’s past employment and UI history, but not the generosity of UI benefits. 2/

Although the link between the generosity of benefits and repeat use of the UI system has not been shown empirically, it is clear that a significant portion of the population use UI as a form of annual income supplement, and that certain regions and industries are overrepresented in the UI population. If one were to define frequent claimants as those with at least three UI claims in five years, 38 percent of UI recipients would have been classified as frequent claimants in 1991; in the Atlantic provinces, the share of frequent claimants ranged from 53 percent to 68 percent of recipients in 1991.

Regional differences in unemployment rates have also resulted in the overrepresentation of the eastern provinces in the general UI recipient population: Newfoundland and Prince Edward Island both have shares of UI recipients that are twice as high as their shares of the labor force, while the other eastern provinces have UI recipient shares that are 20 percent to 60 percent higher than their labor force shares. To the extent that UI generosity influences the unemployment rate, the existence of regionally-extended benefits, which allow for lower entrance requirements and longer benefit entitlements in regions where unemployment is relatively high, exacerbate these inter-provincial differences.

UI claimants are also overrepresented in some occupations, particularly in seasonal industries, implying significant transfers of income between various occupational groups. For example, construction workers make up 11 percent of the UI population, but only about 6 percent of the labor force; they receive in UI benefits almost four times the amount they pay in premia. Other occupations that receive net transfers from UI include agriculture, forestry, and fishing, food and beverage services, mining, and transportation. Many of these occupations also have higher than average shares of frequent users.

3. Proposals for reform

a. General issues

The Canadian UI program has shifted away from being an insurance program toward being a significant social program that redistributes large amounts of income among occupations and regions of the country. In reforming UI, the question to be addressed is the extent to which insurance principles or income transfer features should apply to UI, which in turn depends upon the reasons one believes the public sector should provide unemployment insurance in the first place.

Corak (1994) presents the assumptions of the “idealized insurance model” which, if satisfied, would permit a private insurance market to exist. These assumptions are that: (1) there are no moral hazards (the influence of the insured on the extent and probability of loss is observable); (2) there is no adverse selection (inherent differences in risk are known to the insurer so that premiums can be set at actuarially fair rates) ; (3) there is no correlation in the probabilities of loss (unemployment claims are not strongly correlated among individuals); and (4) the probabilities of loss can be calculated on the basis of past experience (the insurer can form a reasonable prediction of the size of future unemployment claims). The fact that these assumptions are violated to some degree argues for the public provision of some form of unemployment insurance. But the decision to operate a public insurance program as a relatively pure insurance fund, rather than an income transfer scheme, depends in part on the degree to which the various assumptions are violated.

A public program that is closer to a pure insurance scheme may be more appropriate if violations of assumptions (1) and (2) are relatively important. In the context of unemployment insurance, moral hazards and adverse selection are more likely to exist in a labor market characterized by frictional or structural unemployment. The voluntary element of such unemployment implies that workers are able to influence both the probability of becoming unemployed and the duration of an unemployment spell (through search intensity and the choice of a reservation wage). With public insurance, the Government could monitor the behavior of individuals sufficiently to minimize moral hazards, while addressing the problem of adverse selection by making compulsory the purchase of unemployment insurance at actuarially fair rates. Violations of assumptions (3) and 4) would be less important if unemployment were primarily frictional since a relatively small fraction of the population would tend to suffer a loss at any given time, and past experience would provide a relatively sound basis on which to predict future losses. 1/

Alternatively, a public insurance program that contains some elements of an income transfer scheme may be appropriate if unemployment is primarily cyclical and involuntary. Under these circumstances, violations of the latter two assumptions would be relatively important since unemployment claims would be highly correlated and the ability of the insurer to predict the probability of loss would be based on the ability to predict the severity and depth of the business cycle. At the same time, the moral hazard problem would be less serious under conditions of involuntary unemployment since the probability and duration of unemployment is less under the control of the individual.

Under these conditions the Government program would serve as an insurance program, rather than merely a forced saving program for high-risk groups, only if it were compulsory for low-risk groups. But, unless the Government could observe risk differences among individuals and set premiums at actuarially fair rates, such a program would necessarily involve some redistribution of income between individuals with different risks of unemployment. Under circumstances of involuntary unemployment and relatively weak labor market disincentives, the case for deviating from actuarially fair premiums and permitting some redistribution of income is increased.

b. Specific proposals

(l) General program changes

The Government’s paper on unemployment insurance reform (Government of Canada (1994a)) discusses two approaches to reducing the generosity of UI with the objective of improving labor market incentives. One approach would be simply to make the program less generous for all claimants through stricter entrance requirements, lower benefit levels, or restrictions on the duration of benefits. This approach would not differentiate between those claimants who use UI as an insurance program during temporary, involuntary spells of unemployment and those claimants who repeatedly rely on UI as an income supplement. As noted above, those experiencing temporary, involuntary spells of unemployment are less likely (and are less able) to alter their behavior in response to changes in the generosity of the program. Thus, while such changes could affect the search behavior of some claimants, this approach would result in reduced benefits for all claimants, regardless of their income or UI claim history.

The tabulation below contains estimates of the savings that could be generated by a number of alternative program changes. One reform option would be to introduce stricter entrance requirements (the number of weeks of work that are required before one can become eligible for UI). Current entrance requirements are between 12 and 20 weeks of employment during the past year depending on the regional unemployment rate. The justification for variable entrance requirements (VER) is that there is less opportunity for longer work periods in regions with high rates of unemployment; however VERs also discourage workers from moving to regions with better employment opportunities.

Reductions in UI Benefit Payments In Response to General Program Changes 1/

(In billions of dollars)

article image

A second way in which to reduce the generosity of UI is to reduce benefit levels through the replacement rate. Current benefit levels are 55 percent of insurable earnings (up to earnings of $780 a week) for most claimants and 60 percent of insurable earnings for claimants with low earnings and dependents. Only 15 percent of the UI population is eligible for the higher rate, so the incremental savings associated with reducing this rate are not very large (see tabulation).

Benefit levels can also be reduced by shortening the benefit entitlement, or maximum duration of benefits. Claimants are currently entitled to benefits for 14-50 weeks depending on the number of weeks of work and the regional unemployment rate, although most UI recipients actually stop receiving benefits before exhausting their entitlement. The elimination of regionally-extended benefits would make the benefit entitlement solely a function of weeks worked, up to a maximum entitlement of 32 weeks of benefits. Reductions in the maximum entitlement would retain some of the regional variation in benefits since variable entrance requirements would remain.

With the exception of changes in the basic replacement rate which would affect all provinces equally, the Atlantic provinces would-be hardest hit by the changes described above. For example, while the flat 26 week entrance requirement would disqualify 20 percent of UI claimants nationwide, 40 percent of claimants in the Atlantic provinces would be disqualified. Similarly, the elimination of regionally-extended benefits would reduce total UI benefit payments by 30 percent, but would reduce payments to the Atlantic provinces by 48 percent. Simultaneously introducing income testing (see below) would have a mitigating effect on the impact of reform in the Atlantic provinces.

Reducing UI benefits may yield lower savings to the general government than the above tabulation would imply if some UI recipients become eligible for welfare. In that case, costs would merely be shifted to provincial budgets in the form of increased welfare payments. However, a preliminary study by the Canadian Human Resources Development Department found that reductions in UI benefits would likely result in less off-loading onto the welfare system than previously thought. Of UI recipients who exhaust their benefits, only 10-15 percent go on welfare; 80 percent have family incomes above the welfare income limits. 1/

(2) “Adjustment insurance” for frequent claimants

As an alternative to reducing the overall generosity of the UI program, Government of Canada (1994a) proposes the creation of an “Employment Insurance Program” that would attempt to differentiate between claimants subject to longer-term, structural unemployment (who would tend to use UI as a form of social assistance) and those subject to short-term, involuntary unemployment (who would tend to rely on UI as a form of insurance). It would introduce “adjustment insurance” for frequent users, which could involve a lower benefit rate or income testing of benefits, and increase training and job search assistance. Benefits for short-term users (“basic insurance”) would be maintained in its current form.

The tabulation below shows the estimated savings associated with reducing the benefit rate for frequent claimants (those with at least three claims in five years). Another reform option would be to subject benefits for frequent claimants to an income test, since frequent claimants are viewed as relying on UI benefits as a source of supplemental income rather than a source of earnings replacement.

Reductions in UI Benefit Payments In Response to Different Treatment of Frequent Claimants 1/

(In billions of dollars)

article image

Making UI benefits subject to an income test would address the concern that, while income transfers between those at high risk of unemployment and those at low risk of unemployment may be justified, these transfers should take place through means-tested programs. However, to the extent that significant structural unemployment (and therefore UI disincentives) also exist, one could argue that other programs, such as social assistance programs and the Child Tax Benefit, would remain better suited than UI to provide longer-term income support and to reduce income inequality.

The case for substituting training and employment services for cash payments to the long-term unemployed appears strong, particularly when a major source of long-term unemployment is thought to be technological change. However, a detailed review by the OECD of such programs in industrial countries has shown that the results, in terms of unemployment effects, are mixed. 2/

With regard to the aggregate employment effects of active labor market policies (ALMPs), no causal relationship has been identified between the level of public expenditure on ALMPs and employment or unemployment. 3/ This is not surprising since such spending tends itself to respond to developments in employment and unemployment. However, there is some evidence that in countries where spending on ALMPs is relatively high, a given amount of output growth generates more employment growth. From a microeconomic perspective, evaluations of a broad range of programs show significant effects on employment and earnings for some, but insignificant effects for others. The OECD concludes that the more successful labor market programs are those which are designed to serve a well-targeted recipient population and include a high level of involvement of public employment agency caseworkers.

c. Financing issues

The argument could be made that the Canadian provinces would be better suited than the Federal Government to assess local needs and design well-targeted employment and training programs. 1/ At a minimum, Government of Canada (1994b) argues for greater coordination and evaluation of programs at the federal, provincial, and local levels of government. 2/ For example, at the federal level, $1.9 billion was spent in 1994 on UI Developmental Uses (financed by UI premiums), while an additional $1.4 billion was spent on training and employment services for the general population (financed by general revenues). The different sources of funding and program authorities have resulted in a lack of coordination among programs, despite the similar needs of UI and non-UI recipients. Government of Canada (1994a) recommends creating a consolidated fund for federal employment development services financed either in part (with UI premiums) or in whole by general revenues.

The speed with which UI reform could be implemented and savings realized varies considerably among alternative reform proposals. Changes in general benefit levels could be made quickly, but changes to entrance requirements or decisions to treat frequent users differently would likely need to be phased in. Government of Canada (1994a) envisages defining frequent users in terms of the pattern of usage following the implementation of reforms. Thus, if frequent claimants were defined as those with at least three claims in five years, it would take 6 years following reform to identify all frequent claimants. 3/

Government of Canada (1994a) argues that any savings from UI reform should be used to finance premium reductions and/or increase funds for training and employment services. Underlying this argument is the view that lower payroll taxes and/or greater expenditure on employment development services would contribute to lower unemployment rates (see above), but the argument also reflects the current requirement that the UI account be self-financing. On this basis, the savings that might arise from UI reform would not contribute to deficit reduction over the longer run.

Whether the self-financing feature of the UI program should be retained is closely linked to the nature of unemployment and the purpose of UI. To the extent that UI is intended to cushion temporary, short-term spells of unemployment and its insurance features are restored, one could argue that the account should remain self-financing, but that it should balance on average over a long enough period of time so that premiums can remain relatively stable, particularly during economic downturns. On the other hand, to the extent that the economy experiences periods of large-scale, involuntary unemployment, and UI provides income support or training and employment services to the long-term unemployed, the rationale for retaining the self-financing nature of UI is less clear. Under these conditions, UI would contain few features of a pure insurance program, which provided the initial rationale for maintaining a self-financing account. Moreover, the Government would need to be able to predict the depth and length of business cycles in order to finance UI as a fund with relatively stable premiums.

References

  • Coe, David, “Structural Determinants of the Natural Rate of Unemployment in Canada,” Staff Papers, International Monetary Fund, Vol. 37 No. 1 (1990), pp. 94-115.

    • Search Google Scholar
    • Export Citation
  • Corak, Miles, “Unemployment Insurance, Work Disincentives, and the Canadian Labor Market: An Overview,” in Christopher Green, Fred Lazar, Miles Corak, and Dominique Gross (eds) Unemployment Insurance: How to Make It Work. (Ottawa: C.D. Howe Institute 1994).

    • Search Google Scholar
    • Export Citation
  • Cozier, B. and K. Mang, “Explaining the Jobless Recovery,” Canadian Department of Finance, (unpublished manuscript, 1994).

  • Government of Canada, From Unemployment Insurance to Employment Insurance: A Supplementary Paper, (Ottawa, 1994a).

  • Government of Canada, Improving Social Security in Canada: A Discussion Paper, (Ottawa, 1994b).

  • Green, Christopher, “What Should We Do With the UI System?,” in Christopher Green, Fred Lazar, Miles Corak, and Dominique Gross (eds) Unemployment Insurance: How to Make It Work, (Ottawa: C.D. Howe Institute 1994).

    • Search Google Scholar
    • Export Citation
  • Lazar, Fred, “UI as a Redistributive Scheme and Automatic Stabilizer,” Christopher Green, Fred Lazar, Miles Corak, and Dominique Gross (eds) Unemployment Insurance: How to Make It Work, (Ottawa: C.D. How Institute 1994).

    • Search Google Scholar
    • Export Citation
  • Mang, K., “Sticky Wages and Price Stability,” Canadian Department of Finance, (unpublished manuscript, 1994).

  • Organization For Economic Cooperation and Development, Employment Outlook, (Paris: OECD, July 1993).

  • Organization For Economic Cooperation and Development, The OECD Jobs Study: EvidenOce and Explanations, (Paris: OECD 1994).

1/

Prepared by Ellen M. Nedde.

2/

Government of Canada (1994a) notes that UI embodied characteristics of a social insurance program from the beginning, for example by (initially) providing higher benefit levels to those claimants with dependents. A pure insurance program would provide benefits only on the basis of prior earnings.

3/

The benefit rate relates UI benefits to previous earnings (up to an earnings ceiling); entrance requirements specify the number of weeks of work that are required before one can become eligible for UI.

1/

Source: Government of Canada (1994a), p. 101. The 1994 data are estimates.

2/

Specifically, the premium rate is set annually using a statutory premium rate as a reference point. The statutory rate is determined by a formula which calculates a three-year average of UI costs. If the account is projected to have a cumulative deficit (surplus), the actual premium rate would be set higher (lower) than the statutory premium rate.

3/

Applying the formula for setting the UI premium in 1995 would have raised it further to 3.3 percent, but this increase was overridden in the 1994 budget (Government of Canada (1994a), p. 81).

2/

The proposal, discussed below, to provide a lower level of benefits to frequent UI claimants is one form of experience rating. Another possibility is to raise premiums charged to employees and/or employers who rely heavily on the UI system

3/

OECD (1994), Part II, pp. 244-49. Coe (1990) also found evidence that the level of payroll taxes affects the natural rate of unemployment. OECD (1994) cites Mang (1994) and Cozier and Mang (1994) as examples of studies that have found only short-run effects of payroll taxes on unemployment.

1/

Corak (1994), pp. 112-135. Although Corak does not find a significant link between repeat use and the generosity of benefits, he is supportive of the type of proposal, discussed below, which would substitute employment and training services for cash benefits for frequent UI claimants.

1/

Corak (1994) notes that this argument for the public provision of UI is inadequate since the government could disseminate to the private sector any superior information it has regarding moral hazards and adverse selection, while making compulsory the purchase of private insurance (as in the case of automobile insurance).

1/

Source: Government of Canada (1994a). Savings are estimated relative to projected UI outlays of $15.2 billion in 1996.

1/

This analysis may not be relevant for assessing the effects of changes in entrance requirements, which may deny UI benefits to people with lower family incomes.

1/

Source: Government of Canada (1994a). Savings are estimated relative to projected UI outlays of $15.2 billion in 1996.

2/

OECD (1993), pp. 39-69 and OECD (1994), pp. 63-112.

3/

Lazar (1994) argues against the public provision of training based largely on the absence of such aggregate effects (pp. 60-73).

1/

Although there are examples of successful, well-targeted programs designed by higher levels of government, such as the U.S. federal Job Corps program for disadvantaged youths.