This Web page has been archived on the Web.

1994 Report of the Auditor General of Canada

Assistant Auditor General: Richard B. Fadden
Responsible Auditor: Louis J. Lalonde

Foreward

The decision by the government to conduct a Social Security Review came at a time when we were considering the importance of information to Parliament on the results that social programs have achieved over the years, and on the issues involved. This presented the Office with a dilemma.

It is clearly the government's prerogative to review Canada's social policies with a view to seeking changes from Parliament, without interference from us. Indeed, it is obviously the government's responsibility to deal with policy matters, not ours. Furthermore, although we would be dealing with matters of legitimate concern to the Office, we would not wish our work to disrupt unduly the efforts of officials working on the government's policy review.

At the same time, we were convinced that we could make a contribution to Parliament's consideration of this subject, on the basis of both our past audit work and the work we could undertake. Therefore, we decided to publish a study that would be largely based on our past work and on material recently available on the programs under review.

This is a departure for the Office, in that large portions of this study are not based on regular audit work. We believe, nevertheless, that our bringing available information on the value-for-money and financial elements of the programs under review together with background information can be useful to Parliament when it considers the government's eventual proposals. We also believe that our analysis of program results and the identification of information gaps can make a worthwhile contribution to the information to be considered by Parliament.

The study was undertaken in such a way as to minimize disruption of the work being done by departments (Human Resources Development and Finance) in support of the government's policy review. Departmental officials have had an opportunity to read this study report in draft form and were asked to confirm any data that originated from their departments and to indicate any important disagreement with the information provided.

One final point should be noted. Much of the information available on the impact of these programs comes from economic and social studies. These can be subject to a margin of error in individual instances. However, since there is considerable corroboration of basic conclusions from a variety of sources, we believe that the material we cite gives a reasonable, if approximate, perspective on those areas where there is information on what the programs are doing. We also note important areas where further information and clarification are required.

Main Points

6.2 A central concern in our study is that Parliament, as it reviews proposals for reform, lacks important information on the performance and effects of existing social programs. We believe Parliament should know whether these programs are producing the results expected, whether value for money is being obtained, and whether the programs in certain circumstances are having possible negative effects on the economy.

6.3 To assist Parliament, we attempted to put together, using official and public sources, an overview of Canada's social safety net. We focussed on federal government expenditures and the role of federal social spending in the national economy. We also reviewed certain economic and social trends that touch the lives of Canadians and that affect the shape of future social programs.

6.4 We have also looked at the cross-subsidization occurring among industrial sectors and at the difficulties raised by the inclusion of seasonal workers in the Unemployment Insurance Program. We discuss the timing and impact of premium rate increases on the economy, the cost of financing the UI deficit, and the effect of the current funding provisions on the federal deficit.

6.5 We noted that studies of social programs may use different definitions of the activities that could be grouped under the title of social spending, and in the calculations of actual expenditure. We are concerned that these differences in definition may cause difficulties for parliamentarians and others as they review existing and proposed social programs, because such differences complicate the tasks of comparing and analyzing program costs and benefits. This point is important, as there is no consensus on how much our social security system costs and how it compares with those of other countries.

6.6 What the numbers clearly show is that the scale of expenditure is huge, whether measured against the overall federal budget or against the size of the Canadian economy.

6.7 A large portion of this money goes directly back to individual Canadians, although several studies indicate that this spending does not always serve to redistribute funds among the various income groups.

6.8 The studies also indicate that the increased costs of social programs since 1975, combined with other government expenditures, could not be sustained over the years by the tax revenue available and, therefore, significant borrowing became necessary.

6.9 Clearly, the economy is in the throes of change. The changes provide a context for assessing existing and proposed social programs. For example, the information available suggests that:

• the pace of industrial adjustment is increasing, and requiring higher levels of education and skills from the work force;
• labour force growth is slowing, the number of young workers is declining, and there are new pressures on long-time employees to adjust;
• the unemployment levels have been moving upward, particularly since the 1970s; and
• temporary or unstable jobs are on the rise.

• Unemployment Insurance benefits and Canada Assistance Plan transfer payments provide an important social safety net to people in need;
• social programs appear to have made important contributions to reducing regional income disparities;
• Unemployment Insurance has acted as an important economic stabilizer for the economy;
• certain training programs have been effective in increasing earnings and employability of recipients.

• rising social program use and high repeated use suggest the social programs may be creating a long-term dependence among some users;
• disincentives to work when benefits from social programs are compared to earnings from jobs;
• employers and employees may be using Unemployment Insurance to support short-term layoff strategies;
• interactions among social programs may result in the programs working at cross-purposes; and
• Unemployment Insurance may be a factor in Canada's rising level of unemployment and in the lower level of output that results.

6.13 To assist in identifying the needed information, we summarize our findings on key gaps in information available on the workings and impact of the programs studied. We realize that it will not be possible to address all our concerns on information gaps at once. Therefore, it is important for the government to identify the gaps that should be addressed sooner rather than later.

6.14 Parliamentarians need a clear picture of the impacts of social programs on the well-being of Canadians, and their effects on workers' and employers' behaviours toward the labour market. The size and complexity of these programs, and the links between them increase the need for comprehensive and reliable information about results, and particularly about those results that could have a major impact on the redesign of government social programs and policies. We believe that efforts will be worthwhile, notwithstanding the difficulties that will undoubtedly be encountered.

Background

Purpose of the Study

• provides background to the study, tracing the federal government's share of social programs and the impact of social spending on the federal debt;
• looks at some trends in the economic and social environment that are particularly relevant to the analysis of social programs;
• uses available information to determine how programs in three key social policy areas have performed;
• discusses in separate sections matters related to the financing of Unemployment Insurance and to key information gaps; and
• concludes by highlighting several concerns that we believe should be of significance to Parliament.

Program Areas Included in Social Security Review

"Canada's social security system must be responsive to the economic and social realities of the 1990s. The Government will announce an action plan for major reform of the social security system to be completed within two years. The Government will work closely and co-operatively with the provinces and will consult Canadians on the modernization and restructuring of our collective social security. New legislation will then be placed before Parliament."

Also, as announced in the speech, the government plans to sponsor, later this year, a Forum on Health in co-operation with the provinces, to discuss a renewal of Canada's health system.

6.19 The Minister also set two goals for his action plan: to clearly confront the issues facing Canadians and to propose clear options for change. He described the issues and underlying goals in the following terms.

Issues Facing Canadians:

• long-term structural unemployment, even in times of growth;
• the impact of accelerated technological change on labour market and training systems;
• unacceptably high drop-out rates and illiteracy levels, and skills shortages;
• the need to make better use of workers and provide new opportunities for new workers;
• poverty, especially among children;
• lack of training and work for young people;
• tensions between new family structures and the demands of work;
• duplication of government programs; and
• limited financial ability of governments.

• to meet basic labour market adjustments and insurance requirements;
• to restructure parts of the Unemployment Insurance Program and Canada Assistance Plan to create a new form of employment insurance;
• to help people make the crucial transition from school to work by providing a range of options and training, apprenticeship, community service and work;
• to broaden our educational and training assistance to support lifelong learning;
• to enhance support and care of our children in society;
• to redefine the distribution of work and rules of the workplace;
• to ensure that individuals with disabilities can achieve equality, independence and full participation;
• to seek a much better balance between incentives for job creation and payroll tax levels;
• to ensure basic security for those in need;
• to redefine responsibilities between governments and strengthen co-operative arrangements;
• to achieve savings through greater efficiency; and
• to design new, smarter ways to deliver services and avoid duplication.

6.21 The government also stated in the Budget Plan of February 1994 that it will release a paper that will examine the challenges and opportunities posed by Canada's ageing society. It also singled out the Canada Assistance Plan and the Post-Secondary Education transfers as programs that may be modified or replaced pursuant to the Social Security Review. Options to reform the Canada Student Loans programs will be considered separately.

6.22 Several important measures were announced in the Budget Plan. It provided $400 million for 1995-96 and 1996-97, to develop innovative approaches for social security, and $1.7 billion over the next five years for displaced workers of the Atlantic groundfish industry. Immediate measures were introduced to constrain or reduce UI expenditures and social security transfers to provinces and territories. These are discussed further in paragraphs 6.62 to 6.66.

6.23 In March 1994, the Standing Committee on Human Resources Development issued an interim report on the views of Canadians on the reform process. A number of witnesses appearing before the Committee were concerned about:

• the fundamental alterations in the nature and organization of work itself;
• the constant interaction of economic changes with broader social and demographic shifts;
• the persistently high levels of unemployment, and jobless economic growth; and
• the fiscal situation facing Canada.

Transitions Occurring in the Labour Market: "Canadian Realities"

6.26 Social Security in Canada and Canada's Economic Challenges , two booklets published by the Department of Human Resources Development and the Department of Finance respectively, provide further information on major trends affecting Canadians.

The National Picture

6.28 The Caledon Institute of Social Policy (a private non-profit organization) has provided the Office with updated estimates of the social spending figures published in its1993 report, Opening the Books on Social Spending . The figures cover spending on the broad range of income security and labour market programs, social services and health care delivered by the federal, provincial and local governments, enabling us to draw a national picture. Exhibit 6.2 shows both public social spending in Canada by jurisdiction and total spending as a percentage of the Gross Domestic Product (GDP). In constant 1994 dollars, total social spending went from $26.3 billion in 1966-67 to $147.8 billion in 1992-93, representing respectively 8.3 percent and 21 percent of the GDP.

6.29 There is, at the present time, no single way of describing Canada's social safety net. Authorities have different views on what its components are, and have come to different conclusions about both its reach and its cost.

6.30 The social safety net is about people. It reaches millions of Canadians, from children to the elderly, from the unemployed to the disabled. One person's notion of social policy will differ to some extent from that of someone else.

6.31 We reviewed many studies and publications in an effort to grasp the nature and extent of the social safety net. "Social programs", "social safety net", "social policy", "social welfare", "social services", and "the welfare state" are used often as synonyms but some believe that these terms have different meanings. In various studies, the programs are brought together under the label of social spending, social security or simply, social programs. A distinction ought to be made in order to allow comparisons among them.

6.32 "Social spending" generally incorporates federal, provincial and local government social expenditures and the costs of social insurance (Unemployment Insurance, Workers' Compensation and Canada/Quebec Pension Plans) that are supported through employer and employee contributions. Social spending also includes the costs of income transfer programs, employment programs, social services and health care. "Social programs" encompasses all programs that are designed to improve the economic or social well-being of persons or groups of individuals and that are delivered by governments. "Social security", according to Statistics Canada, encompasses a subset of social programs, namely, those directed at income security.

6.33 There are different estimates of social spending virtually everywhere one looks, domestically and internationally. Thus, estimates of spending in the social field vary widely, depending on what is included in the package. None of these estimates includes tax expenditures. To illustrate the range of views, we present the positions of the Caledon Institute of Social Policy, the Department of Finance and Statistics Canada.

6.34 The Caledon Institute estimates that, in current dollars, gross spending in the social field in Canada reached $144 billion in 1992-93.

6.35 In January 1994, the Department of Finance published a background paper on federal spending in 1992-93. The document identifies the federal government's involvement in the social sphere and its spending.

6.36 The Department describes social spending as "encompassing a range of programs designed to enhance the quality of life of Canadians, particularly those who have modest incomes or who are disadvantaged." For the Department, this includes income support, mostly for the elderly and unemployed; transfers to the provinces for health, education and social assistance; and programs for aboriginal Canadians. It also includes spending in areas such as social housing, student loans, worker training and adjustment.

6.37 The Finance figures show federal social program spending at $49.6 billion for 1992-93, including $42 billion in direct transfers to individuals. Another $18 billion in cash transfers went to the provinces for education, health care and social assistance, to bring total social spending to more than $67 billion. A further amount of about $12 billion was provided to the provinces in the form of tax transfers; these represent the value of income tax points in respect of specific program areas, transferred to provinces in lieu of direct cash transfers.

6.38 Social security, according to Statistics Canada, covered a total of $78 billion in expenditures at the federal, provincial and municipal levels for 1990-91.

6.39 Differences in definition and coverage make it difficult to obtain consistent data for comparisons and trend analysis. Tax expenditure data are vital for a proper measure of social spending and should be included because the federal and provincial governments make extensive use of the personal income tax system to deliver social benefits. Furthermore, while most social programs in this country are financed and administered by governments, important social benefits are also delivered by employers, the volunteer sector, the private sector and family members. The timely availability of such comprehensive information can only help policy makers in the current review of social programs.

The Constitution

6.41 The Constitution empowers the federal government to operate three major income support programs: Unemployment Insurance, Old Age Security and the Guaranteed Income Supplement.

6.42 Under the Constitution, the provinces have jurisdiction over major aspects of social security, such as health, welfare, and education. The provinces also have jurisdiction, with important exceptions, over labour matters.

6.43 The federal government has some responsibility in the social security fields by virtue of its spending powers. This means it may set general conditions and provide financial support. It may not administer the programs.

6.44 Other activities are a matter of common jurisdiction. Both levels of government may operate in the pension area, where the provinces have primacy, and in the immigration field, where federal law takes precedence.

6.45 The Constitution Act , 1982, committed both levels of government to jointly pursuing national equity goals. The Charter of Rights and Freedoms offered mobility and equality guarantees. Further, the Constitution, as amended in 1982, committed the federal government and the provinces to promoting "equal opportunities for the well-being of Canadians" and "providing essential public services of reasonable quality to all Canadians".

Weaving the Social Safety Net

The Federal Share of Social Programs Spending

6.48 Financing is necessarily a central concern in any serious review of Canada's social programs. They represent very large expenditures. Therefore, we looked at the sources of revenue and where the social spending occurs.

6.49 Most of the funds for the social programs are raised by two levels of government, federal and provincial, and the costs of some major programs are shared between these two levels. In some provinces, municipal governments share in the responsibility for funding welfare payments. Regardless of which government pays, the source of funds is, for the most part, the same Canadian taxpayer.

6.50 Federal revenues are generated by taxes such as personal income tax, sales and excise taxes, corporate income tax, and payroll taxes such as Unemployment Insurance premiums. Exhibit 6.4 shows the 1993-94 federal revenue mix, which represents a total of $116 billion.

6.51 According to data supplied by the Department of Finance, total government revenues (including federal, state/provincial, and local levels) in most industrialized countries have increased, to varying degrees, as a percentage of Gross Domestic Product (GDP) since 1980. Government revenues in Canada rose from 31.6 percent of GDP in 1980 to 37.3 percent by 1991.

6.52 All taxes have increased, except for corporate income taxes. Exhibit 6.5 presents the change in the federal revenue mix between 1980 and 1992, as a percentage of total revenues. The share of total federal revenues from personal income tax increased from 40 percent in 1980 to 48 percent by 1992. Sales and excise taxes grew over the period, from 14 percent to 18 percent, and Unemployment Insurance premiums from 6 percent to 14 percent. Furthermore, Unemployment Insurance and Canada Pension Plan contributions from employees increased from 1.8 percent of personal income in 1980 to 2.7 percent in 1992. Including the employers' share, the growth in these payroll taxes during the same period was from 5 percent to 8.5 percent of personal income. Corporate income tax declined from 15 percent in 1980 to 7 percent in 1992. The Department of Finance attributes this primarily to the impact of the recession on corporate profits in the early 1990s.

6.53 The federal government can pursue its social and economic objectives in many ways. Exhibit 6.6 shows where the federal government spends its money, and the portion allocated to social programs.

6.54 Exhibit 6.7 shows the amount spent for each social program and the number of beneficiaries that are assisted. As the Canada Pension Plan (CPP) is jointly administered by the federal and provincial governments, it is not accounted for through the federal government's financial statements; therefore, it is not included in the federal government's spending totals. CPP spending, which excludes Quebec, was around $13.2 billion in 1993.

6.55 Transfer payments to persons are mainly directed to the unemployed, elderly Canadians, children and veterans. Transfers to aboriginal administrations are made for social development, education, community infrastructure, band management, economic development and health. These cash transfers are made through the departments of Indian and Northern Affairs and Health.

6.56 As stated by the Department of Finance in Federal Spending , federal transfers to provinces "help the provinces in providing services such as medicare, education and social assistance."

6.57 In the case of transfers to other levels of government, as illustrated by Exhibit 6.8 , the government responsible for taxing is not always the one responsible for spending. Thus, taxpayers cannot easily associate the services provided by a government with the taxes paid to that government. We acknowledge that intergovernmental transfers are a necessary part of any decentralized system of government, such as the Canadian system, to ensure efficient allocation of resources across provinces. While they may pursue other laudable policy goals, intergovernmental transfers can obscure the relationship between programs and taxes. Thus, accountability can be blurred when governments spend money they have not raised directly.

6.58 The government may also pursue policy objectives through the income tax system, providing relief from the payment of taxes through such measures as income exclusions and deductions, tax credits and deferrals. In many ways, these measures represent an alternative form of government assistance, with financial implications similar to those of direct expenditures. Therefore, they are referred to as tax expenditures.

6.59 Reasonable differences of opinion exist on what constitutes a tax expenditure related to social spending. Different organizations have produced different estimates of social tax spending. The Caledon Institute of Social Policy estimated the federal and provincial social tax spending to be $28.2 billion (current dollars) for 1991, including provincial revenue losses. The Institute for Research on Public Policy estimated that, for 1989 alone, the cost to the federal government for selected personal income tax expenditures related to social policy objectives was about $34.4 billion. We have not reviewed the bases for these estimates.

6.60 Tax expenditures pose several problems for measuring social spending, beyond the discussion about what is and what is not a social benefit. In general, tax expenditures reduce the resources available to fund other programs or to reduce the deficit, and can result in the need for higher tax rates to obtain the same amount of revenue. In the case of social tax expenditures, there is no consistent time series to complement the historical direct spending figures; so, only a partial picture of trends and patterns in total social spending emerges. A recent example serves to illustrate our point.

6.61 Total benefits for children exceeded $5 billion in 1992-93. Starting 1 January 1993, Family Allowances and the Child Tax Credit and Refundable Child Tax Credit were replaced by a new program, the Child Tax Benefit. Because the new Child Tax Benefit is now delivered through the tax system, it is not included in budgetary spending. Thus, it becomes difficult to make comparisons or analyze trends in program spending in this area.

Recent Budget Proposals

6.63 The Canada Assistance Plan's main objectives are to help provinces provide adequate levels of social assistance and institutional care for persons in need, and welfare services that lessen, remove or prevent the causes and effects of poverty, child neglect or dependence on public assistance. Until recently, CAP transfers were open-ended, conditional, matching grants available uniformly across all provinces. The federal government placed a limit on annual increases in transfers to three provinces, from 1990-91 to 1994-95, for cost sharing of provincial social assistance and social service programs under CAP ("cap on CAP"). For the other provinces, the federal government pays half the cost of provincial social assistance and welfare services. As shown in Exhibit 6.7, the CAP expenditures for the federal government went from $6.7 billion in 1992-93 to $7.2 billion in 1993-94, an increase of over seven percent.

6.64 The Canada Assistance Plan provisions remain unchanged for 1994-95. For 1995-96 and thereafter, transfers to the provinces under CAP will be constrained to 1994-95 levels, subject to whatever change may arise from the Social Security Review. Possible changes are planned for introduction in 1996-97. The cap on CAP remains in place over the period of the review. Then, according to the Budget Plan, entitlements under CAP, or its successor, will be no higher after reform in 1996-97 than in 1993-94.

6.65 For the Established Programs Financing (EPF), the existing freeze on per capita entitlements remains in place for 1994-95 through 1995-96. For 1996-97, the post-secondary component of EPF will be adjusted so that, when it is combined with CAP (or its successor), the total social security transfer will not exceed 1993-94 levels. The freeze does not apply to the health care component. Thus, there will be a clear separation of EPF into its two components and each will be placed on a different track.

6.66 The Budget Plan proposed the following changes to Unemployment Insurance: a reduction of the maximum duration of claims, an increase in the minimum entrance requirement, and a reduction of the benefit rate in most cases. These measures are expected to reduce Unemployment Insurance expenditures by $725 million in 1994-95 and $2.4 billion each year thereafter. Other measures were introduced to reduce premium rates for 1995.

Significance of Federal Social Programs

6.68 The long-term trend in federal Current and Capital Expenditures, as a share of the Gross Domestic Product, has been stable with gradual declines. Government Transfers to Business fluctuated considerably in relation to GDP from 1975 to 1985; since then, they have declined significantly. Business Transfers is the smallest of the expenditure categories we examined, and by 1992, it had roughly the same GDP share it had in 1966.

6.69 In contrast, Transfers to Persons and Transfers to Governments grew very rapidly in the second half of the 1960s. While subject to cyclical fluctuation, Transfers to Persons continued a pronounced, long-term upswing. As a share of GDP, the Transfers to Persons ratio grew from a cyclical high of around six percent in the mid-1970s to nearly eight percent by 1992. Transfers to Governments stabilized as a share of GDP from the mid-1970s onward.

6.70 Overall, Transfers to Persons became the single largest item among federal government expenditures from the early 1970s onward, with long-term expansion exceeding the rate of growth in the economy. If payments under the Canada Pension Plan were also included, the level of expenditures under Transfers to Persons would increase even more. As noted, CPP payments are not included as part of the federal government's financial statements and the spending total on social programs. As we indicated in paragraph 6.54, CPP spending was around $13.2 billion in 1993; a significant addition to the $50 billion we identified for Transfers to Persons in Exhibit 6.7 .

6.71 Since the mid-1970s, social program cost increases have come largely from the growth in federal social security spending, backed by steadily mounting debt charges. These increases have been one of the major contributors to an uninterrupted string of federal government deficits. Other transfers, including those to assist provincial governments' social spending, have been held in check by increasingly stringent measures over the past decade, and have kept in line with the nation's economic growth.

6.72 It is particularly important to consider the cost-effectiveness of each component of Canada's social programs, because of the scale of expenditure and its significance to the deficit and the debt.

6.73 In summary, as we have shown in the Background section, federal social programs enter broadly into the fabric of Canadian life and are a very significant component of public spending at the federal level. The social safety net was established in the four decades between the Great Depression and the late seventies. The federal government's social spending currently ranges up to some $67 billion and, even by modest estimates, is equal to more than half of federal revenues. Some $42 billion of this is put directly into the hands of individual Canadians.

6.74 Canada's level of public spending has important implications for its national credit rating and its standing on international money markets. An analysis of social program expenditures and recipients over time furnishes vital information on trends and patterns that, in turn, should help policy makers better understand the social and economic factors that affect the current and future demands on the social security system. However, Canada still needs better social program information for a complete, accurate and consistent accounting of public expenditures.

6.75 The pattern of long-term trends in federal government expenditures indicates that, from 1975, when deficit concerns first surfaced, the increased costs of the social programs, combined with the other program expenditures, could not be sustained over the years by the tax revenue available and, as a result, significant borrowing became necessary. This is reflected in the very significant growth trend in interest payments.

Socio-economic Trends

6.77 These trends (paragraphs 6.80 to 6.101) offer a framework for assessing information on how the programs are functioning and the results they have achieved over time.

6.78 One striking feature that emerges from the description of the current environment is the extent of the changes under way. The trends themselves can be seen to be changing - in many cases, they appear to reflect significant changes in the structure of the economy and the labour market as well as changes in social conditions. This strongly implies that they are unlikely to be temporary in nature or automatically self-correcting.

6.79 In all, we identified 10 basic trends that are of particular importance to the social programs. These are summarized below.

6.80 Industrial adjustment in the economy has been increasing and can be expected to continue at high levels. The Department of Finance's 1994 publication, Canada's Economic Challenges , identified two types of development that are forcing industrial adjustment:

• changing trade relationships (such as The North American Free Trade Agreement and the wider globalization of the economy); and
• technological change and innovation.

6.82 Major demographic changes under way will affect the use and cost of social programs. The labour force is projected to grow at a much slower rate in the 1990s. It is estimated that, on average, only about 180,000 people will join the labour force each year, compared with over 200,000 during the 1980s and 300,000 in the 1970s ( Exhibit 6.11 ).

6.83 With slower growth in population, young people aged 15 to 24 will make up a smaller portion of the available labour supply ( Exhibit 6.12 ). The "middle age" categories of workers will grow as a portion of the population and labour force ( Exhibit 6.13 ).

6.84 In the past, a large and growing supply of young workers was available to meet the economy's new skill needs. Younger workers change jobs more frequently and have more flexibility to adapt. The ageing of the work force will mean that more and more established workers will be exposed to the industrial adjustment and redeployment process. This will increase the pressure to use social programs to facilitate adjustment and redeployment, rather than simply to provide income support.

6.85 There has been a long-term trend toward increasing unemployment rates. Cyclical unemployment has been an important factor in increasing the numbers of the unemployed since 1990. However, over and above this, the long-term trend in unemployment has been upward ( Exhibit 6.14 ) and underwent a significant increase commencing in the 1970s. By the late 1980s, even when economic growth was strong, the national unemployment rate did not fall below 7.5 percent, compared with a low of approximately 4.5 percent in the 1960s.

6.86 Over the long term, job creation in Canada has been better than in most industrialized countries, but the nature of jobs is changing. Average annual employment growth between 1973 and 1988 was well over two percent for Canada ( Exhibit 6.15 ) and less than one percent for most other industrialized countries.

6.87 It should be noted, however, that while Canada's long-term employment growth record has been impressive, jobs have become less stable. The Economic Council of Canada estimated that 44 percent of Canada's employment growth in the 1980s was accounted for by temporary or unstable jobs.

6.88 Multi-earner families are now the norm. By 1989, about 70 percent of Unemployment Insurance claimants belonged to families with two or more earners ( Exhibit 6.16 ). Thus, while general employment conditions may have worsened, the impact of unemployment on most families will not necessarily be as severe as in past periods of high unemployment. Single-earner families, in contrast, will continue to face severe difficulties when job loss occurs.

6.89 There is a strongly entrenched problem of regional disparities in unemployment. Persistent regional unemployment disparities can indicate a structural problem in the economy, involving among other factors a geographical mismatch between labour demand and supply.

6.90 A comparison of regional unemployment rates against the national average for Canada at the peak of the business cycle, when labour demand is strong ( Exhibit 6.17 ), shows no overall improvement in regional disparities. In some cases, disparities have worsened.

6.91 Workers from abroad meet a significant proportion of labour needs in growth regions. As a percentage of employment growth, immigrants arriving in Canada between 1986 and 1991 amounted to 38 percent of the increase in employment in Ontario during this period. In British Columbia, the other major employment growth area, the figure was nearly 18 percent ( Exhibit 6.18 ).

6.92 The extent of the dependence on workers from abroad in growth areas raises questions about the geographical mobility of Canadian workers. It also raises questions about the possible influence of the Immigration Program on where firms locate. The shortage of jobs in higher unemployment regions affects the need for social programs.

6.93 The trend toward increased unemployment has occurred even when job vacancies were more plentiful. This trend has existed for some time (see Exhibit 6.19 ). The 1991 Organisation for Economic Co-operation and Development (OECD) Economic Survey of Canada indicated a basic doubling of the unemployment rate between the 1960s and the late 1980s for a given rate of job openings generated by the economy. In the mid-1960s the job vacancy rate was accompanied by an unemployment rate of less than four percent. By the late 1980s, essentially the same vacancy rate co-existed with an unemployment rate of over eight percent, suggesting the possibility of a labour supply adjustment problem.

6.94 The central question is, "Why would job vacancies go unfilled while unemployed workers are available?" As we noted in paragraph 6.85, there is clear evidence of a long-term trend toward increased unemployment in the economy. Several factors are likely to be involved, including the effect of structural changes in the economy. Another possible reason is that the levels of income support provided by the social programs, (see paragraph 6.96) may be generating work disincentives and causing a form of voluntary unemployment. This would contribute to the growing unemployment trend and the eight percent unemployment figure cited above.

6.95 With voluntary unemployment, even if jobs were available, workers would prefer the benefits of the social safety net. Where significant levels of job vacancies and unemployment co-exist, this can constrain economic growth. Further, it can greatly increase the burden on social program expenditures.

6.96 The Department of Finance, in its January 1994 report on Canada's Economic Challenges, concluded that the system of Unemployment Insurance in Canada is, considerably more generous than the safety net for workers in the United States and the average for other advanced industrialized countries. Exhibit 6.20 compares the levels of income support for unemployed workers in the G-7 countries. These levels of income support involve coverage and accessibility as well as levels of payments. The Department of Finance links the issue of generosity to the problem of work disincentives. While the level of support of the Canadian UI system has been reduced as a result of program changes since 1990, it still remains an important area of concern.

6.97 Growth in per capita wage income has slowed and earnings disparities are widening. In Canada, average wage income grew significantly until the early 1970s. In the 1980s, there was very little overall improvement in wages ( Exhibit 6.21 ).

6.98 At the same time, income disparities widened. However, this was offset to a considerable extent by total government transfer payments ( Exhibit 6.22 ).

6.99 Government income transfers are now more important in maintaining income levels. The effect of income transfers on individual income categories is shown in Exhibit 6.23 .

6.100 The role of government income transfers in maintaining income levels has grown from 1981 onward. Except in the lowest 20 percent income group, where transfer payments were already the predominant income source, support from income transfer programs has more than doubled since the early 1970s. This raises questions about the appropriateness of universality instead of targeting and about the levels of support of the various income transfer programs.

6.101 We were unable to locate comprehensive information on the extent of "tax-back" of income transfers for the income categories shown in Exhibit 6.23. Such information would be important in identifying more clearly the net effect of the income transfer programs and in assessing whether they were being delivered in the most cost-effective way.

6.102 In summary, the picture that emerges from our examination of prevailing socio-economic trends is one of significant industrial adjustment in the economy, involving an equally significant need for labour force adjustment and adaptation. This economic adjustment coincides with a significant lessening of growth in wage income and an increased importance of government income transfers across all income groups.

6.103 Information on the growth in such income transfers, combined with international comparisons of levels of support provided through unemployment insurance, raises important questions about the cost-effectiveness of targeting, about work disincentives, and about the relative generosity of programs.

6.104 In this context, we now turn to information available on the actual results individual social programs are achieving vis-à-vis the objectives involved.

Information on Program Results

6.106 Federal departments are required, under Treasury Board policy, to undertake periodic evaluations of their programs, "where material and cost-effective to do so".

6.107 Government departments collect assorted ongoing data on their programs, some of which may give a partial perspective on program results. These, and outside data such as Statistics Canada surveys, are sometimes examined in wider program reviews by departments.

6.108 Researchers also periodically undertake technical studies of general program effects. Furthermore, commissions of inquiry have analyzed some aspects of programs. International institutions such as the Organisation for Economic Co-operation and Development and the International Monetary Fund carry out studies touching on program performance of member countries, including Canada, as part of wider country reviews or as part of comparative analyses between countries.

6.109 With all these sources, it is possible to obtain some perspective on the results achieved by federal social programs over the years.

6.110 However, as will be shown, the perspective is a very partial one in some respects. One of our major findings identifies information gaps on existing programs, which need to be considered in the context of the Social Security Review.

Program Activities Reviewed

• Unemployment Insurance.
• Social assistance (welfare) programs supported by the federal government through the Canada Assistance Plan.
• Worker training: for UI claimants and social assistance recipients (SARs).

6.113 The program activities are key components of the federal government's social safety net, with a strong focus on members of the working-age population and their dependants.

Recent Program Evaluation Reports

6.115 Program evaluation has been recognized in government policy over the years as a key form of results measurement. The deputy heads of departments are accountable for the planning, conduct and reporting of program evaluation studies.

6.116 As outlined below, the Department of Human Resources Development made available to us program evaluation reports examining program effects in three of the four program areas. The Department also provided us with a pre-evaluation assessment report for Unemployment Insurance regular benefits.

6.117 With respect to Unemployment Insurance, we noted in our 1993 Report that this area had received only limited evaluation coverage to date.

6.118 In its pre-evaluation assessment of Unemployment Insurance regular benefits, the Department acknowledged that this important program component had not been evaluated previously, although it is the federal government's largest labour market adjustment program, and despite the revived concern over the persistent and dramatic growth of the expenditures in past years.

6.119 The Department is at present in the process of completing a major evaluation of the impacts and effects of regular UI benefits on the clientele and on the economy as a whole. Some preliminary information on the findings can be drawn from a series of studies examining such program effects. Other studies are planned for this fall and for 1995. Such information can contribute to the current review of Canada's social security system. It is important that information from program evaluation studies be made available to Parliament when major decisions are being taken. We encourage the Department of Human Resources Development to continue its evaluation activities as planned.

Report Title Date Status

Evaluation of the Canada
Assistance Plan 1991 Final Report

Pre-Evaluation assessment
of the Unemployment Insurance
Regular Benefits Program 1992 Draft Report

An Evaluation of UI-
Sponsored Training 1993 Draft Report

Evaluation of Employability
Initiatives for Social
Assistance Recipients 1993 Draft Report

6.120 While the present evaluation work is a positive development, its timing may be regarded in some ways as "bolting the stable door after the horse has gone". This is particularly true in view of the significant questions raised over the years about possible negative effects from Unemployment Insurance, its size and its costs. These are discussed in paragraphs 6.184 to 6.200. Earlier program reviews, including two major commissions of inquiry, could have benefited from the availability of results measurement information.

Intended Effects

6.122 In the case of Unemployment Insurance, the objectives include the redeployment of unemployed workers (through income support, counselling, job search and training assistance) and the provision of a social safety net (financial assistance) for the jobless. However, as discussed below, the scope of Canada's UI Program has not been well defined. This is evidenced by the fact that observers have different views of its role. Many critics of the program insist that its proper role is a particular form of insurance, while others see it as a vehicle for income redistribution.

6.123 In the case of the Canada Assistance Plan and training for social assistance recipients (SARs), the objectives are more straightforward. CAP's focus is on income support "as a measure of last resort" provided through provincial welfare systems: clearly a social objective. For SAR training, the principal intended effect is to increase employability and reduce dependence on government assistance.

6.124 The UI Program lacks clear objectives. In keeping with enabling legislation in general, Parliament has set out the purpose of the Unemployment Insurance Program in broad terms. This broad statement of intent should be translated, to the greatest extent possible, into clear, consistent and measurable objectives as the basis for program design, management and accountability. This has not been done adequately, as evidenced by the various statements related to UI objectives ( Exhibit 6.25 ).

6.125 The most recent description of the objectives for UI is:

"to promote economic growth and flexibility by providing temporary income support to unemployed workers who qualify for benefits under the UI Act without placing an unnecessary burden on individuals, groups or regions."

6.127 Similarly, the intended meaning of the expressions "economic growth" and "flexibility" is not clear in UI's function either as an insurance program or as a social income redistribution measure. The absence of well-structured and well-defined objectives is evidenced by the fact that UI has "branched out" over the years to areas such as sickness benefits, maternity and parental leave, and special provisions for self-employed workers in the fishing industry. Similar questions must be asked about "developmental uses" of UI, where training represents the largest share but which also include direct job creation, self-employment and work-sharing expenditures.

6.128 This is not to say that such activities are not legitimate. However, accountability for public funds requires that the program objectives be clearly defined and the scope of the activities be clearly identified.

6.129 The situation is further confused by the fact that there is no single, consistent government statement of program objectives for Unemployment Insurance. Instead, as indicated in Exhibit 6.25 , authoritative documents offer different views of what is expected from the program. Moreover, each of these sources - Part III of the Estimates, a commission of inquiry, departmental review documents and the like - is authoritative in its own right. In short, the objectives for UI vary, depending on where one looks. Further, as the exhibit indicates, in the Department's annual report - an area where it might be expected that the objectives of UI would be clearly and consistently stated - there was no statement at all of the program's objectives.

6.130 Why is the lack of clear objectives for the Unemployment Insurance Program so important? There are several reasons.

6.131 First, the size of the overall program budget, which has ranged up to more than $19 billion in recent years, means that large sums of public money are being collected and distributed through the program. Contributors are entitled to have a clear statement of what the money is intended for and actually used for, and what results have been achieved.

6.132 Further, if clear objectives have not been formulated and communicated by senior management, there is a high risk that general governance and operational management will be impaired. This is so since there cannot then be a coherent focus on what the program is specifically trying to achieve in relation to Parliament's stated intentions when it passed enabling legislation.

6.133 As a result, program managers, parliamentarians, clients and other stakeholders will likely have different perceptions of what the program is trying to do. Cost-effectiveness will likely be reduced, since it is difficult to be clear on intended results and to take whatever corrective action may be necessary. It also becomes more difficult to hold government accountable for results.

6.134 When the absence of clear objectives is combined with the absence of any comprehensive feedback on results achieved for moneys spent, the spectacle emerges of a major program that has been functioning for some time without adequate accountability.

6.135 Most observers recognize that UI cannot operate entirely the way a private sector insurance scheme would. However, it is necessary that it function within a clearly identified framework of consistent objectives. We acknowledge the difficulty of setting clear objectives and recognize that, in some cases, there will be conflict between them. When this is the case, the obligation to be transparent and to measure the results achieved is all the greater. In addition, because of the complexity of the program, Parliament may wish to indicate more precisely in future legislation what its intent is with respect to the program's objectives.

Positive Effects of the Programs

Unemployment Insurance

6.138 UI has in place a series of control measures to detect and deter abuse. Evaluation studies by the former Department of Employment and Immigration in the late 1980s indicated that the controls have generated program savings. However, the issue remains on whether, overall, the most cost-effective approach for an unemployment insurance system is with features that include ease of accessibility (see paragraph 6.96) combined with a range of administrative controls and increased complexity to prevent abuse.

6.139 The need for recourse to provincial welfare systems and the complexity of program delivery require further attention by the Department so that operational effectiveness can be improved and value-for-money results strengthened.

6.140 Program data indicate that UI has functioned well as an income transfer mechanism related to regional disparities in unemployment. An important effect of UI is a significant degree of income redistribution across provinces. Exhibit 6.26 illustrates the interprovincial distributive effects of UI, confirming that it is an important instrument of income transfer to provinces with higher unemployment and slower employment growth. In 1992, approximately $3.2 billion was transferred in this fashion. Historically, a major share of the transfers has gone from Ontario and the western provinces to Quebec and the Atlantic provinces. This trend arises from both the differences in incidence of unemployment across provinces and the important design feature of the program that results in a uniform premium regardless of province of residence.

6.141 However, the program may be much less effective as a redistribution mechanism among income groups. For example, a departmental analysis indicates that, in 1990, only 7 percent of total UI benefits were paid to families with less than $15,000 annual income, while about 50 percent of the benefits went to families with annual incomes greater than $40,000 and over 33 percent to families with incomes greater than $50,000.

6.142 The above analysis does not provide information on premium revenues compared with program expenditures by family income group. This, along with information on the more general income tax provisions, would allow development of a perspective on the net income redistribution effect of the program. There is also evidence from a departmental evaluation study that the introduction of more restrictive UI provisions has further reduced the income redistribution role of the program.

6.143 It is as an economic stabilizer that UI appears to have had its greatest positive impact on the economy. A departmental evaluation study concludes that net UI spending has a significant effect on reducing the employment impact of an economic slowdown. This is especially true of UI benefit payments but less so with respect to UI premiums. The evaluation points out that the premium rate has moved "pro-cyclically"; that is, the rate tends to be raised in downturns.

6.144 The study concludes that premium rate increases, which have occurred during economic downturns, have reduced the program's potential for economic stabilization. However, if the federal government were to keep these premium rates constant, it would incur higher deficits and have to bear the cost of financing them. This is further discussed in paragraphs 6.239 to 6.243.

6.145 The study's estimate of the average stabilizing effect of UI is about 10 to 12 percent. This means that the UI Program has prevented about 10 to 12 percent of the decline in output that would otherwise have occurred at the lowest point of a cyclical downturn. In employment terms, this translates into 11 to 14 percent of job losses being averted through the income maintenance provisions of the program. A parallel study commissioned by the Department arrives at a similar basic conclusion.

6.146 While UI income support can help workers find more suitable jobs, the extent of the effect is not large . A preliminary departmental study on intensity of job search by the unemployed concludes that the availability of UI can help workers engage in a more productive job search, resulting in a better match between labour demand and supply.

6.147 The study estimates that, with this improved matching, UI claimants' wages increased as much as seven to nine percent. However, it concludes that these benefits may not exceed the additional costs of claimants remaining on longer UI claim periods to find a more suitable job.

Canada Assistance Plan

6.149 The evaluation found that disparities in provincial social assistance levels have been reduced since CAP was introduced. The evaluation also noted that, in some provinces and particularly the poorer ones, authorities claimed that social services would not have developed as rapidly without CAP support.

6.150 However, there is a significant need for more information on the programs that CAP supports. What are the characteristics of those being assisted, such as age, sex, family status, education and previous job experiences? Why do individuals require assistance; how often; for how long? What programs have proved the most or least effective in reducing reliance on welfare? Improved information on the causes, extent and effects of poverty and the impact of existing programs is necessary to evaluate the impact of federal cost sharing on Canada's safety net programs. Information on the overall effects in redistributing income to benefit needy Canadians is not available.

6.151 While the federal government shares the financing, the provinces are responsible for the design and administration of social assistance in their respective jurisdictions. The flexibility of CAP when combined with regional diversity has resulted in a unique welfare system in each of the 12 jurisdictions (the 10 provinces and two territories) in Canada. Understandably, federal-provincial program definitions are not harmonized and categories of recipients and beneficiaries are not always consistent between provinces. As a result, the data received from provinces are fragmented and difficult to roll up on a national basis.

6.152 The problem is further exacerbated by federal government decisions that have a direct impact on provincial welfare systems. Following the "cap on CAP" announced in 1990, the three provinces concerned (Ontario, Alberta and British Columbia) no longer provide information to the federal government once they reach their maximum level of funding. Also, recent cuts to the Unemployment Insurance program could result in increased provincial welfare rolls but the two levels of government do not agree on the extent of this consequence. Such recent events have made provincial governments reluctant to provide more information on their welfare program activities.

6.153 The lack of information about social programs funded by CAP has existed since its inception. The complexity of federal-provincial relations, combined with the problems of aggregating data from various sources, makes obtaining information on program results difficult at all levels. However, as we have pointed out in previous reports, we urge the federal government to work with the provinces on this long-standing issue.

6.154 The fundamental question that needs to be answered is what is being achieved with the funds actually spent. This includes the resources the federal government spends annually on CAP ($7.2 billion in 1993-94). In our opinion, feedback information of this type is essential before decisions are made on redesigning safety net programs. The absence of such information means that Parliament cannot assess the appropriateness of the program's targeting through provincial welfare systems or the degree to which the assistance promotes dependence among welfare clients. Such issues are further considered below, in the context of training initiatives for social assistance recipients.

Training

6.156 On the positive side, while both trainees and workers who had not received assistance were equally likely to find jobs, trainees on average required 11 to 17 fewer weeks than did their non-trainee counterparts to find employment. Individual earning gains for trainees reached more than $6,000 a year for some groups.

6.157 However, results varied greatly by type of training. In certain cases, trainees registered a modest but significant rise in applications for welfare assistance, compared with the experience of non-trainees. In other words, they were, if anything, worse off than similar workers who had not had government-sponsored skill development assistance. The reasons for this are not clear at present.

6.158 Training for social assistance recipients (SARs) can produce positive results. The available information indicates that SARs enjoyed substantial gains in both earnings and employability after undertaking training. They exhibited reduced dependence on social assistance, an increase in weeks employed and greater labour force attachment. However, there was also a subsequent increase in the use of Unemployment Insurance at the end of SARs' employment spells, as a result of greater UI eligibility and assistance from employment gains and wages.

6.159 Increased employment of SARs, when it leads to increased use of Unemployment Insurance, will involve a transfer from one income support program to another. The Department claims that the program showed a net benefit for the national economy. It is not clear, however, to what extent the newly trained workers displace others from employment or push them into the social safety net. This should be investigated by the Department.

6.160 Cost-effectiveness of resource distribution needs study. There is a need to reassess the present resource distribution between "passive" labour market measures such as income support and "active" measures such as training and other types of worker redeployment programs.

6.161 Data on seven leading industrial countries, prepared by the Organisation for Economic Co-operation and Development, show that Canada's expenditure on unemployment compensation, which amounted to nearly 2.3 percent of its Gross Domestic Product in 1992, was second only to Sweden's. Even when differences in unemployment rates are taken into account, Canadian expenditure on unemployment compensation remains high in comparison with that of countries such as the United States, Japan, France and the United Kingdom ( Exhibit 6.27 ).

6.162 In contrast, the Canadian percentage of Gross Domestic Product spent on "active" labour market measures such as training is more in line with the mainstream of advanced industrialized countries ( Exhibit 6.28 ).

6.163 Overall, Canada has the lowest ratio of expenditures on active to passive measures. As indicated earlier, available information on the UI Program's positive effects underlines the limited effectiveness of passive income assistance alone in enhancing the redeployment of workers to other jobs. This raises an important value-for-money question about levels of expenditure on passive income support under UI.

Negative Effects of the Programs

6.165 Negative effects raise important value-for-money issues because they can undermine or counteract positive results produced by the programs, increase program costs, and decrease program effectiveness.

6.166 Our assessment, based on the limited information available, is that there may be possible negative effects from federal social programs in the following areas:

• chronic dependence on government programs;
• built-in work disincentives for clients in the welfare system;
• increased worker layoffs by employers as a result of UI;
• the social programs may be causing increased unemployment levels; and
• where UI-induced unemployment occurs, there may be significant costs beyond a waste of program resources.

6.168 Chronic dependence on government programs. Recent evidence raises concern that there is an extensive and growing dependence on government social programs.

6.169 While use of social programs naturally increases during economic slowdowns, the programs do not return entirely to earlier levels when economic and employment growth recovers. Rather, the programs continue to operate at higher levels that become the new "floor" for subsequent activities ( Exhibit 6.29 ).

6.170 The extensive repeated use of UI has led some observers to characterize a significant part of the assistance as a form of regular income supplement. Departmental information on program "repeaters" for 1991 indicates that some 1.3 million regular claimants (about half) had used UI at least one additional time in the previous two years. One-third of regular claimants had at least three other claims in the preceding seven years. On the basis of these figures alone, it must be expected that the costs arising from program dependence are considerable.

6.171 A departmental evaluation study confirms the trend of increased program use indicated by the data. The study concludes that a "first time exposure to UI permanently increases the probability of receiving UI again in the future." The study's results indicate that the first use of UI increases the probability of future use by 12 percent for older workers and 6 percent for younger workers.

6.172 Built-in work disincentives for clients in the welfare system. Back in the 1960s, most people on assistance were aged people and persons with disabilities who were not at that time expected to earn income through employment. Since the 1980s, problems related to disincentives to work have started to emerge, with an increasing number of employable individuals claiming welfare.

6.173 Every province has a set of guidelines known as "earning exemptions". These are the amounts that welfare recipients may earn without any reduction to their welfare assistance. Earnings from employment that exceed designated levels typically reduce welfare benefits, dollar for dollar. Welfare recipients are allowed to keep only a small portion of the earnings; the net gain from working is minimal.

6.174 Furthermore, the earnings of welfare recipients are subject to income taxes and payroll taxes; welfare income in itself is not taxable. Thus, the combined effects of earning exemptions and taxation on earned income constitute a barrier to moving off welfare.

6.175 Social assistance recipients who enter or re-enter the labour market may lose access to income in kind, for example, health-related benefits, dental care, medication plans and assistance for work-related expenses. Although the value of income in kind is not known, it can be worth substantial amounts of money, especially for households with any form of special need, such as a medical or disabling condition.

6.176 According to a study by the Economic Council of Canada, a growing proportion of the new employment opportunities is in non-standard jobs (that is, work arrangements that differ from the traditional model of full-year, full-time employment). About 44 percent of all new jobs created between 1981 and 1989 were short-term or part-time jobs. This type of employment represented about 28 percent of all jobs in 1989. Statistics Canada data show that part-time employment rose by 211,000 between 1990 and 1993, while full-time employment fell by 400,000, indicating that the trend observed in the 1980s is continuing in the 1990s.

6.177 Non-standard employment typically, although not always, offers low wages, few benefits and little security. According to the Economic Council study, the difference between government financial assistance provided to people in need and the income earned from jobs that are less than full-time may not be large enough to encourage some people to work; the extent of such a disincentive effect was not estimated. There is also a risk that any increase in income support levels may even encourage some low-wage earners to withdraw from the labour force (see paragraphs 6.214 to 6.217). More and better research on these issues is needed.

6.178 Increased worker layoffs by employers as a result of UI. From the individual employer's point of view, the wage bill and short-term production costs can be reduced by the use of a combination of temporary layoffs and recalls of workers eligible for Unemployment Insurance. The more generous the UI provisions, the greater the acceptability to workers of such layoffs as a response to a firm's business fluctuations. As noted above, the Economic Council of Canada has estimated that as many as 44 percent of the jobs created in the 1980s were unstable.

6.179 Preliminary information on UI Program effects, from departmental evaluation studies, gives a further perspective on the problem.

6.180 This information indicates that layoffs exceed the level of job quits as part of the movement of workers out of employment each month. In some cases, extensive repeated use of the UI Program has been occurring in association with the same group of employers; over 40 percent of workers with five or more UI claims had each been employed by three or fewer employers. This implies a significant layoff-recall pattern among certain employer groups.

6.181 Analysis sponsored by the Department also indicates that, when qualifying periods for UI are increased, it is employers who tend to adjust short-term employment periods to ensure UI eligibility before laying off workers. Thus, the adjustment to changing UI provisions involves employers as well as workers.

6.182 As indicated in paragraph 6.277, the use of UI varies significantly among industry sectors. This means that premium revenue from some sectors is used to finance heavier program use in others. This raises the issue of experience rating (discussed further in paragraphs 6.229 to 6.231). Moreover, such use of UI resources may cause certain sectors to retain too much labour, at the expense of contributors in other industry sectors.

6.183 From the available information, there are grounds for concern that the UI Program's provisions may be contributing to increased short-term job spells in the economy. This possible effect needs further examination by the Department.

6.184 The social programs may be causing increased unemployment levels. The two income transfer programs examined as part of our study raise concerns. In both Unemployment Insurance and the Canada Assistance Plan, work disincentive effects may accompany the programs.

6.185 In the case of income transfers and other social assistance benefits supported through the Canada Assistance Plan, only very limited information is available (see paragraphs 6.148 to 6.154). It is not clear how extensive the work disincentive effects may be (see paragraphs 6.172 to 6.177). Yet, it is extremely important to have this information if value for money is to be obtained from scarce program resources.

6.186 In the case of the Unemployment Insurance Program, there is a clearer perspective. Here too, however, the information base is seriously incomplete.

6.187 There is at present no clear consensus on what the precise effects of UI are on overall unemployment in Canada.

6.188 However, the available evidence indicates that the UI Program's activities are accompanied by certain work disincentives and there is a significant risk that these may be increasing Canada's unemployment rate ( Exhibit 6.30 ). The Department should clarify this matter as quickly as possible.

6.189 As noted, the 1981 study, Unemployment Insurance in the 1980s , identified concerns that the increased availability and level of support of UI after 1971 might have brought an upward shift in the unemployment rate, ranging from 0.5 to 1.3 percentage points.

6.190 Subsequent studies, most of which were undertaken outside government, produced a range of estimates; in some cases, they failed to find any effect, and in others, they estimated upward shifts in unemployment ranging up to two percentage points and beyond. Consequently, it is clear that this is a difficult area to measure.

6.191 The Organisation for Economic Co-operation and Development (OECD) concluded, in a 1991 study, that the provision and availability of UI in Canada had added two percentage points to the level of unemployment after the program changes of the 1970s.

6.192 Subsequent revisions to UI are likely to have lowered the level. However, given the continued generosity of Canada's UI system, compared with many in other advanced industrialized countries, and the longer-term trend for high frequency users to account for an increasing share of UI spells, the net effect of UI-induced unemployment is likely to remain significant. The program would thus have a structural effect on the labour market, through worker and employer behaviour in response to its provisions.

6.193 The UI Program is so large that even a relatively small induced unemployment effect can have significant cost implications. For example, adding a one-half percentage point to the unemployment rate through increased program use due to work disincentives could involve more than $500 million in current program expenditures. This is an important, negative and unintended effect. The Department should ensure that its assessment of the impact of UI on the unemployment rate is reported to Parliament and that the implications for program expenditures are identified.

6.194 Where UI-induced unemployment occurs, there may be significant costs beyond a waste of program resources. As discussed in the section on socio-economic trends, one striking characteristic of the changed environment has been the trend toward increased unemployment in the economy even when economic growth is occurring. While there are probably several factors involved, our examination has raised the issue of UI work disincentives generating voluntary unemployment. With voluntary unemployment, even if job openings were available, workers would prefer the benefits of the social safety net, and the level of unemployment in the economy would rise.

6.195 We have also noted that job vacancies in the economy appear to have been co-existing with much higher unemployment levels than in the past. OECD estimates for Canada indicate a basic doubling of the unemployment rate between the mid-1960s and the late 1980s occurring while the rate of job vacancies was basically similar during these periods.

6.196 Where significant levels of job openings and unemployment occur at the same time, it raises the issue that constraints to economic growth may be a result. The 1991 OECD analysis of vacancies and unemployment in the Canadian labour market demonstrates a growing long-term trend toward such conditions during periods of economic growth.

6.197 According to available benchmark information, if UI-induced unemployment constrained economic growth as a result of voluntary unemployment, increasing the unemployment rate by 0.5 percent, then the cost to the economy could involve up to $7.5 billion in lost output - a very high cost indeed. This is not a pressing issue in present economic conditions. However, it is still important to identify as fully as possible any effects from the program in this respect.

6.198 A 0.5 percentage point level of UI-induced unemployment would not necessarily have this full effect since, in some cases, UI may be inducing people who would not otherwise be participating in the labour force to register as unemployed. Also, in the case of UI clients, there may be further factors (such as the occupational skills and industry sectors involved) that will reduce the impact of the loss of production.

6.199 However, our analysis offers strong evidence (paragraphs 6.189 to 6.193) that a loss-of-production effect in the Canadian economy should be a source of concern. The actual impact should be examined and measured by the Department as part of its program evaluation efforts.

6.200 In our opinion, such possible negative effects should have been investigated and measured years ago, for reasons of sound governance and to fulfil the requirement for accountability to Parliament for moneys voted.

Inter-program Linkage

6.202 Program linkage can, at times, expand or compound particular program effects on the economy or the clientele. Programs can be designed to complement each other to produce jointly intended program results. In other instances, program interactions can work at cross-purposes and detract from the achievement of program objectives.

6.203 We examined the limited available information on inter-program linkage and its effects. Exhibit 6.31 assesses some possible key program links. The most important findings are summarized below.

6.204 There is evidence that, in some contexts, welfare claimants move between welfare and UI. A significant proportion of welfare recipients have also received Unemployment Insurance. Statistics Canada's Labour Market Activity Survey indicates that approximately 32 percent of people in Canada's working-age population who received welfare sometime between 1988 and 1990 also received Unemployment Insurance. This group made up roughly nine percent of UI claimants during the same period. The implication from this is that any changes to the provisions of the UI Program are likely to have an impact on provincial welfare systems.

6.205 A preliminary report of a departmental evaluation covering 1993 indicates that, in the first five months following job separation, nearly six percent of UI claimants also received social assistance. However, in over half the cases, welfare assistance was claimed for a two-week period while claimants waited for UI benefits.

6.206 As a measure of last resort, it would be expected that, in the case of the long-term unemployed, workers exhausting UI would then move to welfare, where eligible. What the departmental evaluation shows is that a significant number of UI claimants ask for welfare prior to receiving UI, because of the two-week waiting period for the UI program to respond to their claim.

6.207 A 1992 departmental study looked at the possibility of provincial government measures moving welfare claimants onto the UI rolls. It calculated that the resulting cost to the UI Program may have reached $2.6 billion in 1991. This raises concerns that an ongoing "recycling" is under way between the social programs, rather than cost-effective action to enhance long-term employability.

6.208 The Immigration and Unemployment Insurance programs may at times be operating at cross-purposes. Workers from abroad played a significant role as a supplementary source of labour between 1986 and 1991 in Ontario and British Columbia, two regions of employment growth (Exhibit 6.18). In Ontario, immigrants were considerably more significant than net internal labour movements from other provinces, while in British Columbia, they were roughly equal.

6.209 Both the Macdonald Commission and the OECD have voiced concerns that an unintended effect of social programs such as UI has been the interruption of the geographical adjustment of labour supply and demand in response to the jobs the Canadian economy creates.

6.210 Thus, a paradox exists: serious regional disparities in unemployment (labour surpluses) co-existing with a strong dependence on a supplementary labour supply channel from abroad (immigration) to fill Canada's occupational needs. Workers coming to Canada through this channel, which leads mainly to regions of employment growth, contribute to programs such as UI and greatly assist in paying for the interprovincial transfer of funds to regions of higher unemployment. In other words, the supply of labour from abroad assists in underwriting the costs of the social programs that the Macdonald Commission and the OECD felt were making more difficult the matching of labour demand and existing labour supply in Canada. This can result in the need for further supplementary labour supplies from outside the country to fill labour requirements in areas where there are employment opportunities, while other areas continue to experience labour surpluses.

6.211 The problem of matching labour demand and labour surpluses within Canada can involve several areas of concern. One is the movement of workers to other regions where jobs are more plentiful. A further concern is the possible role played by supplementary supplies of labour from abroad in influencing the concentration of Canadian industry in regions of existing economic growth, while other regions experience high unemployment. We were unable to find adequate information touching on this issue of possible program effects. Yet, the possible implications clearly indicate that this is a very important issue that should be considered in the context of the range of effects accompanying Canada's immigration system.

6.212 Federal training programs for UI claimants and social assistance recipients (SARs) have an impact on their earnings, employability and dependence on social programs. A departmental study shows that some types of training for UI claimants increase their earnings and employability. As mentioned in paragraph 6.157, in certain cases, trainees registered a modest but significant rise in applications for welfare assistance, compared with the experience of comparable workers who had not taken training.

6.213 As indicated in paragraph 6.158, training for SARs increases their earnings and employability. The Department indicates that, as a result, dependence on social assistance is reduced. However, provincial government measures appear to be moving welfare claimants onto Unemployment Insurance. It is also not clear to what extent newly trained workers displace others or cause them to move to social assistance. This should be investigated further by the Department.

6.214 There is often little financial incentive for a social assistance recipient to accept minimum-wage work. Welfare rates generally are calculated on the basis of family size. Minimum wage is a standard amount that does not vary with the circumstances of the household. Depending on the province, the size of a family and its needs, a household can be better off on welfare than on a minimum-wage income.

6.215 Exhibit 6.32 , prepared by the National Council of Welfare, compares welfare benefits, which are not taxable, to working at minimum wage. An incentive to work is calculated when earnings from minimum-wage jobs (gross income minus deductions from income taxes, CPP contributions and UI premiums) exceed the net welfare income (basic welfare income plus exempt earnings permitted without any reduction in welfare benefits). Disincentives are noted where welfare benefits are higher than earnings from minimum-wage jobs.

6.216 According to the National Council of Welfare, there are strong incentives for employable single people and two-earner couples to work in most provinces, but strong disincentives for other types of recipients. The table shows that earning less or more on welfare than on minimum wage is an incentive or a disincentive (shown in parentheses in the exhibit). This may be the case but there is no evidence on what these differences should be, in dollar terms, to really act as a motivation for taking work.

6.217 In many cases, the support levels for welfare substantially exceed feasible earning levels for those in minimum-wage jobs. This could be inequitable to those who support themselves and their families through low-wage employment. This possible inequity is exacerbated by the fact that many welfare clients are also eligible for income in kind (non-cash income) such as subsidized housing, training, family dental care, and so on. Since these are strong disincentives for some welfare recipients, measures to help them enter the labour market should be explored.

6.218 In summary, we have looked in this section at three large spending programs. These are Unemployment Insurance, social assistance programs supported by CAP, and training for UI claimants and social assistance recipients. In each case, we were attempting to determine what they were intended to achieve, and to discover from departmental information and other sources the actual effects of the money spent.

6.219 Recent program evaluation initiatives by the Department of Human Resources Development are significantly increasing the available information on UI effects. Its coverage needs to be extended to additional effects resulting from inter-program linkage in the socio-economic field. There is simply no shortcut to building up a body of evaluation information on a program over a period of years.

6.220 In our opinion, the results measurement now being undertaken by the Department can make a real and positive difference to the public debate. Making these studies available on a timely basis will shed more light on the results achieved by federal social programs.

Matters Related to the Financing of UI

6.222 Changes have been made in the method of financing UI and in its benefits. The federal government, which had borne more than a third of the costs of UI in the 1970s, progressively transferred all costs to employers and employees between 1979 and 1990.

6.223 UI premium rates increased by 54 percent between 1989 and 1993, while the UI account went from a cumulative surplus of $1.1 billion to a cumulative deficit of $5.9 billion at the end of 1993. For 1994, the Employment and Immigration Commission forecast an annual surplus of some $2 billion that should reduce the cumulative deficit to less than $4 billion.

6.224 UI financing raises several important questions that should be dealt with in the context of a Social Security Review, but which we cannot fully address in this study because they relate to policy. In particular:

• what are the appropriate roles of the private sector (through employers and employees) and of government in the financing and governance of UI?
• is the current premium allocation between employers and employees appropriate?
• how should premium rates and maximum insurable earnings be determined?
• should the treatment of earnings while clients claim benefits be changed to reduce work disincentives?
• should the benefit repayment provision for high-income claimants be changed to alter the threshold for repayment?

• UI contributors and beneficiaries;
• the cost of financing the UI deficit; and
• the impact of the UI deficit on the deficit of the Government of Canada.

UI Contributors and Beneficiaries

6.227 Some industries are financing less than their share of benefits. Industry-related data, obtained from the Department, show that unemployed workers in the primary and construction industries collected from UI more than three times the amount those industries contributed in 1992 (see Exhibit 6.34 ). These two sectors accounted for more than 20 percent of benefits paid.

6.228 As recognized in many studies, large variations in the use of UI between industrial sectors have led to a situation where industries providing relatively stable, year-round employment are regarded as subsidizing less stable industries.

6.229 Experience rating is often proposed as a means of eliminating or reducing cross-subsidies between industries. It is argued that, in keeping with "insurance principles", firms or industries that generate a disproportionate amount of unemployment through their layoff practices should pay higher premiums. The 1971 Act allowed the Canada Employment and Immigration Commission to make regulations for experience rating of individual employers. This provision was repealed in 1977 without ever having been used.

6.230 The Commission is on record as rejecting experience rating in its own reviews (1977, 1980 and 1986). The government argued, among other things, that it was paying a substantial portion of UI costs. It maintained that some costs were beyond the employers' control and, therefore, should not be financed according to experience rating. It also argued that experience rating could potentially reduce employment opportunities in small firms and for youth. Administrative costs were also cited as an issue.

6.231 In the present context of scarce program resources, user pay is being more and more studied. This raises questions about the extent to which cross-sector subsidies pay for repeated use of UI by employees in seasonal industries.

6.232 A 1993 Statistics Canada study reported that seasonal employment accounted for a large portion of the repeated use of UI in the short term (defined as individuals claiming UI twice within 14 weeks). A recent departmental analysis indicates that UI claimants in seasonal employment impose a high ongoing cost. For example, in 1989, claimants in seasonal industries made up about a third of all claimants of regular benefits. About 80 percent of these made repeated claims within a two-year period.

6.233 Seasonal regulations were introduced in 1946. In 1956, coverage was extended to self-employed workers in the fishing industry. Seasonal benefits were integrated with the regular benefits, commencing in 1957-58, by the extension of benefit duration to about five-and-a-half months.

6.234 In 1962, a Committee of Inquiry into the UI Act proposed that seasonal benefits be replaced by an extended benefit plan reflecting regional variations in unemployment and financed by general revenues. Seasonal benefits were viewed as a major departure from the principles of social insurance. The objection was not so much that persons working only part of the year in insured employment received a subsidy during the off-season, but rather that the subsidy was drawn from an insurance plan and financed by insurance contributions.

6.235 In 1971, with relaxed entrance requirements for regular benefits and an extension of their duration, seasonal benefits were no longer required. The government assumed more responsibility for financing the extended benefits that were attributable to national and regional unemployment rates over four percent.

6.236 The 1990 legislative changes simplified the structure of regular benefits and it is no longer possible to identify the amount of extended benefits attributable to regional unemployment rates. The duration of regular benefit payments remains tied to regional unemployment rates, with the cost, as noted, now shifted to employers and employees.

The Cost of Financing the UI Deficit

6.238 In the absence of a reserve, UI has had a cumulative deficit for 14 of the 22 years since 1972 and incurred over $3 billion in interest costs. An additional $1 billion has been projected for interest charges to finance its current deficit. It took about six years to pay off the cumulative deficit in the 1980s (see Exhibit 6.35) .

6.239 One concern about the way premium rates are set relates to their year-to-year instability and, in particular, to the timing of rate adjustments with respect to business cycles. The timing appears to be disruptive to the economy.

6.240 The effect of rate increases on the economy is illustrated by the following examples. At the start of 1983, with the economy near its recessionary trough, the combined employer and employee premium rate was increased from 3.96 percent to 5.52 percent of insurable earnings. This represented an increase of some $2 billion in UI premium payments from employers and workers over a year.

6.241 A similar situation occurred in the recession that started in 1990. The combined rate had already been raised to 5.4 percent in 1990. Again, with the economy near its recessionary trough in late 1992, the combined rate was moved up to 7.2 percent, representing an increase in UI premium payments of some $3 billion for the year ahead.

6.242 Increases of 54 percent in the rate since 1989 are attributable, for the most part, to an expansion in the developmental uses of UI, to the end of the government's participation in financing UI, and to the recession. A large deficit and a subsequent rise in premiums in the early stage of an economic recovery can generate pressures for benefit reductions.

6.243 As we mentioned in our 1988 Report, a reasonable reserve in the UI Account can be desirable. It can alleviate the need for borrowing, reduce interest costs thereof, and provide some relative stability to the premium rate as well as protection against unexpected cost increases. The accumulation of a reserve will, however, create pressure for extending benefit coverage or decreasing premium rates. This raises the question of what constitutes a reasonable reserve and what time is required to build it.

The Impact of the UI Deficit on the Government of Canada Deficit

6.245 In the short term, this interaction can influence the government's management of UI. While UI is self-financing over time, however, its impact on the federal budget is neutral in the long term.

Information Gaps

6.247 In our opinion, these information gaps are important because they represent critical areas where parliamentarians need information on the performance and effects of the existing programs, to be able to consider more fully and realistically any future proposals for program development submitted for Parliament's approval. Exhibit 6.36 summarizes our findings on such information gaps. We encourage the federal government to work with the provinces in order to fill some of the identified gaps.

6.248 Statistics Canada calculated in 1994 that the underground economy was likely to range up to three percent of Gross Domestic Product, a share that translated into some $21 billion annually and included unrecorded employment activity. It is not known to what extent participation in the underground economy co-exists with ongoing dependence on government social programs.

6.249 Poverty is the subject of considerable debate in Canada. To understand the nature of poverty, useful and credible indicators that are acceptable to all stakeholders would help. The estimated number of poor and the poverty rate are two examples of key social indicators. At the present time, there is no general consensus on what poverty is, although numerous studies and research have been done on the subject. Like most nations of the world, Canada has no "official" measure of poverty.

6.250 Statistics Canada's low income cut-offs (LICOs) have been widely used as poverty measures, despite the fact that it has clearly and publicly stated that it does not regard LICOs as indicators or measures of poverty.

6.251 Besides LICOs, several other measures have been suggested as poverty measures. Differences in methods result in significant differences in the estimates of the number of people living in poverty. Thus, with several definitions used, it is currently difficult to determine precisely the number of poor and "near poor", and the extent of poverty in Canada.

6.252 To get a very clear picture of the extent of poverty in Canada, researchers must "all march to the same tune". The use of an approach that allows credible comparisons is of paramount importance. Parliament should have consistent and trustworthy information on poverty before considering any proposals for program redesign.

Conclusion

6.254 We acknowledge that the Social Security Review covers other programs that we have not reviewed in this study. We also understand that the issues we have raised cannot all be addressed at once and it is the prerogative of Parliament to decide what should or should not be considered.

6.255 Over the past few years, we have noted special efforts to study many of the socio-economic issues facing Canada. Such efforts by both the private and the public sectors are very encouraging. More studies are planned for release this fall and in early 1995. The results of these studies should provide more information, close certain gaps and, perhaps, raise more questions that will need answers.

6.256 As we have seen, estimates of spending in the social field vary widely, depending on which programs are included. The available social program statistics could be improved: there are problems of definition in estimating spending and in relation to the availability and quality of data. With the spotlight on social programs, shortcomings in public social expenditure reporting are neither academic nor merely domestic.

6.257 We have attempted to situate federal social spending in the context of overall social spending in Canada. This will, we hope, help everyone see the national picture as well as the federal, provincial and municipal shares of social spending in Canada. After all, the same taxpayers are the source of funds for all three levels.

6.258 Our discussions with key stakeholders suggest that it is not clear if UI is to function as an insurance program or as a special income redistribution measure. There is evidence of repeated use of UI by employees in seasonal work, of increased layoffs and of using UI to reduce costs in social assistance programs. Thus, clarifying the objectives of the UI Program is important; contributors should have as clear a statement as possible of what the money is intended for and actually used for, and what results have been achieved.

6.259 There is substantial evidence of benefits obtained from social programs. These should not be overlooked. UI payments provide an important safety net to unemployed workers, put funds into poorer regions of the country and, at low points in the business cycle, play a significant stabilizing role in the economy, reducing the severity of the downturn. The Canada Assistance Plan, delivered through provincial programs, has successfully directed federal money to people in need and may have helped launch some social services in poorer regions. Training funds for both UI claimants and social assistance recipients have led to increased earnings and employability.

6.260 However, using a variety of sources, we have identified evidence of possible negative effects that have received inadequate attention from the government. These possible effects include long-term dependence on government programs; disincentives to work built into both UI and welfare systems; increased resort to layoffs and use of UI to carry firms through low production periods; and an overall rise in the national unemployment rate, combined with chronic labour surpluses in some regions of the country co-existing with dependence on labour supplies from abroad in others. There are no easy, fast solutions for some of these problems.

6.261 A further concern is the possibility that the UI Program, through its effects on unemployment, may be resulting in a poor employment performance by the Canadian economy in relation to its productive potential during periods of economic growth.

6.262 There are also concerns related to the financing of UI: the timing of rate adjustments, interest costs incurred because of the absence of a reasonable reserve in the UI Account, and temporary layoffs as a result of a uniform premium rate.

6.263 It must be emphasized that, at the present time, there are more questions than answers concerning possible negative effects. Yet, enough information exists to suggest that these questions should be taken seriously and evaluated by the government. Answers - even approximate answers with identified margins of error - should be sought and brought forward for debate.

6.264 From the available information, it is clear that UI and provincial social assistance measures provide financial support to the same clients that experience employment disruptions. The impacts of inter-program linkage should be clarified as Parliament considers social policy changes.

6.265 The complexity, size, design and potential linkage of the social programs, their effects on workers' behaviour in the labour market, and the fact that the programs involve two or more levels of government call for improved accountability. To a large extent, this can be assisted by filling the information gaps on what the programs are doing and the results they are achieving. Given the debt and deficit situation, parliamentarians need a much clearer idea of what works and what does not.

6.266 The government is faced with the difficult task of formulating policies that will give broadly consistent - not conflicting - signals across economic, fiscal, social and education fields, the labour market, and the public and private sectors. Economic constraints should not be allowed to obscure the importance of social objectives, and social policies should not impede economic progress and should be framed in light of the fact that resources are limited.

6.267 The issues and questions discussed in this study are unlikely to be temporary or self-correcting and they are certainly difficult to resolve. What is at stake is the well-being of millions of Canadians who remain unemployed, who live in poverty, and who seek to improve their skills. We hope that this study will be useful to parliamentarians as they consider the government's proposals.