|
|
     |
||
 |
     |
 |
|
Benefits and Other Features of Canada Pension Plan Final Report
Evaluation and Data Development May 1997
SP-AH010E-05-97
This evaluation study was conducted by ARC Applied Research Consultants, and William M. Mercer Limited of Ottawa, under the direction of an Evaluation Steering Committee comprised of representatives from HRDC, Status of Women Canada, and Finance Canada. The evaluation team would like to thank all those who contributed to the study. We are particularly grateful to Nancy Lawand, Rick Levinsky, Judy Richardson of Income Security Programs, to Pierre Fortier, Rod Hagglund, Rick Morrison, Joe Burpee, Gary Bagley, and Rachel St-Jean of Social Policy Directorate, Strategic Policy, HRDC who provided valuable assistance and feedback to the evaluation effort.
The Canada Pension Plan (CPP), introduced in 1966, was designed to provide pension income to retired and disabled workers, surviving spouses of deceased contributors and orphans. For the first time in Canada the plan provided a public, earnings-related retirement income, together with ancillary benefits, survivor benefits and disability benefits. Central to the survivor benefits are a Surviving Spouse’s Pension, Orphan’s Benefits, and a Death Benefit. Virtually all workers in Canada, whether employees or self-employed, were required to contribute to the plan. Benefits have been indexed annually to offset the effect of inflation. Since 1989, the survivor benefit has stabilized at about 17% of total CPP payments, with the bulk of these paid to women over the age of 65. B. The Scope of This Report This study is part of a larger effort which encompasses the review of retirement benefits and disability benefits under the CPP being conducted by Human Resources Development Canada (HRDC). This phase of the CPP evaluation is aimed at determining whether there is a continuing rationale for providing survivor benefits and for other features of the CPP, such as the dropout provisions and credit splitting. At present, the general dropout provision allows 15% of years with lowest earnings to be dropped out in calculating CPP benefits. The child rearing dropout permits parents to drop additional years for raising children up to 7 years of age. Credit splitting refers to the division of CPP pension credits between members of divorced or separated couples. This report integrates qualitative and quantitative findings from a wide range of sources, including surveys of surviving spouses and of the general public, as well as from simulations using the CPP Actuarial Model, other micro-simulation models, analysis of administrative data, and opinion from experts in the field. The report describes the historical development of CPP survivor benefits and other features of interest. Then the current program coverage and continuing rationale are examined. The report examines the extent to which objectives have been achieved and addresses the key issue of the significance of survivor benefits in relation to all other sources of income, including personal income and, where applicable, that of all members of the household. Impacts and effects of the survivor benefits and ancillary features are examined, along with an analysis of the economic effects of CPP contributions on employers and employees. The last chapter identifies issues that deserve further consideration. C. The Rationale for Survivor Benefits Remains Despite Social Changes Ninety percent of survivors are women; orphans benefits are paid mainly for children who have lost their fathers. Also, the general dropout provision is beneficial to those with erratic earnings, traditionally more characteristic of women. Credit splitting was intended to benefit women. The mid-1960s model of the typical family pattern—a male breadwinner and a non-working wife taking care of the children—has changed, but the need for earnings replacement has not diminished. Some of the key changes that have occurred together with their impact on the rationale for various components of the CPP are summarized below:
We found strong support among experts, the public, and our own analysis of data for the rationale of continuing federally provided, post-retirement survivor benefits in their present form and—by extension—to pre-retirement survivors over 55. However, there was some disagreement among experts on survivor benefits for pre-retirement survivors. The formula for pre-retirement survivor benefits includes a flat-rate component which is not related to the deceased contributor’s earnings record. There is therefore an element of "income support" as well as income replacement. Interviewees as well as the expert panel found the rationale for this benefit design is less clear than in the case of the post-retirement benefits, where only the earnings-related component is present. Insofar as orphan’s benefits and the death benefit are concerned, the expert panel and key informants were less supportive of the rationale than in the case of spousal survivor benefits. However, there was little support for the elimination of these benefits, partly because of the relatively small savings that would result. The continued rationale for the general dropout was supported, even enhanced in view of labour market instability. In spite of changed labour force participation of working mothers, evidence for continued need for the specific child rearing dropout provision was offered. Strong support was shown from experts for the rationale for continuation of credit splitting on marriage breakdown. On the other hand, assignment of pension benefits in the absence of marriage breakdown was questioned. D. The Profile of Beneficiaries has Changed The number of recipients of CPP survivor’s pensions has increased rapidly over the last three decades, from less than 100,000 in the early 1970s to 735,345 in January 1996. In that month, 89% of the beneficiaries were women, and of the women almost three-quarters were over the age of 65. The age of new beneficiaries has steadily increased. There has also been a dramatic increase in female beneficiaries over 75—a 24 percentage point increase from 1984 to 1995. A 1996 survey of beneficiaries showed that less than 10% of female survivors were remarried; 75% lived alone. In addition, only 18% of all female survivors had been employed (full-time or part-time) in 1995. The total dollars expended for Surviving Spouse’s Benefits in the month of January 1996 was $180.7 million, with an average benefit of $244.01. The maximum benefit in 1996 is $436.25 if the spouse is 65 or over. Younger female beneficiaries—those of pre-retirement age—represented only 28% of the number of beneficiaries, but represents 32% of the dollar value of the benefits to female survivors. E. Survivor Benefits are Very Significant for Only a Minority of Women Survivors For 80% of female survivors, the survivor benefit represents less than 20% of household income. However, multiple lines of evidence indicate that for low income women—those with $10,000 in income or less—survivor benefits represent a very significant proportion of total gross household income. Especially among lowest income women, the proportion of income represented by CPP survivor benefits has increased over time. Survey data showed that female survivors who perceive their current income to be less than adequate are disproportionately numerous among women of pre-retirement age and those with little or no education. F. Experts and the Public Differ Somewhat on Eligibility Rules for Survivor Benefits
The general public tends to be both restrictive and generous with respect to eligibility for benefits. Canadians think that survivors who remarry should not receive a benefit, but they would open up eligibility for younger pre-retirement survivors in addition to the disabled and those with children. Current beneficiaries tended to support the status quo with respect to whether survivor benefits should vary with the age of the surviving spouse and whether benefits should be related to the survivor’s income. They also support the current rules with respect to Orphan’s Benefits. Experts and key informants think the current rules are, by and large, appropriate. Even on the more controversial pre-retirement eligibility rules, there is general acceptance of the income support aspects—largely because the principle of income support for families overrides their dislike for a departure from the income replacement principle. If the $3 billion of expenditure on Survivor Benefits were cut from the CPP/QPP, the net effect would mean a compensating increase of $1 billion in other income support programs. The manner in which the lower costs net out is complex, given that OAS, GIS and tax credits are all affected by the presence of a CPP survivor benefit as part of income. G. Retention of Other Survivor Benefits, Dropout Provisions and Credit Splitting is Generally Well Supported
Experts and our key informants supported retaining the death benefit, orphans benefits as well as the general and child rearing dropout provisions and credit splitting. Regarding the death benefit, male and female beneficiaries differed in their perceptions of adequacy—more males than females found the death benefit to be "less than adequate." The opinions of female survivors were consistent with the views of key informants and the expert panel—namely, that the death benefit makes a reasonable contribution following the death of a spouse and, therefore, should not be dropped. Many survivors find the current name insensitive and would like to see it changed. The Surviving Child’s Benefit seems more appropriate. Regarding the Orphan’s Benefit the public supported the status quo, but there was fairly strong support for extending Orphan’s Benefit to at least 22 years of age, even if the child was not in school. On the general dropout provision, the public favoured extending it to cover other forms of family-related care-giving beyond the current rules, but there was no agreement on lengthening or shortening the number of years of general dropout permitted. The experts and the public differ on the mandatory aspect of credit splitting: the public is more disposed to negotiation in the case of divorce and separation; the experts favour mandatory provisions. Simulation results using the CPP Actuarial Model showed that the removal of mandatory credit splitting on divorce would actually increase CPP costs because of the way credit splitting currently interacts with dropout provisions. Credit splitting should transfer credits from men to women and from those with a less interrupted earnings record to those with a more interrupted one. Women also live longer than men, such that equal benefits transferred to women would be more costly. However, the effect is outweighed by its interaction with the splitting of the child rearing dropout—which results in lower total costs to the CPP. This reduces the percentage of maximum CPP pension available to husbands after credit splitting. The result is some savings to the Plan. H. Survivor Benefits Do Not Have a Major Impact on Labour Force Behaviour Women’s labour market behaviour is little affected by the death of a spouse and the receipt of a benefit. Most of the current beneficiaries were not in the labour force at the time of the death of a spouse and, for most, the situation did not change afterwards. Trend analysis showed that the number of weeks worked by women revealed a downward trend both before and after the start of benefits. This is reflected in annual average earnings. We also found that employed survivors differed, but not significantly, from those with no employment income with respect to the proportion of income provided by survivor benefits. The issue has been raised as to whether employed survivors should receive lower benefits. We found little justification in the data for considering changes to the benefit structure to accommodate differences in income related to employment. In future, it is not likely that significantly more widows will be employed at the time of the death of their spouse/partner. In part this is because the longevity of males is greater now than in the past and also because of a trend to earlier retirement. This makes it more likely that neither partner will be working at the time of the death of a spouse, and there is little reason to think that many will seek/find employment afterwards. I. CPP Payroll Taxes currently Account for a Fifth of Employer Payroll Taxes In comparing tax incentives that affect the labour market, a complete perspective requires information on other taxes borne by employers. Currently the CPP does not dominate the cost impacts on employers, since payroll taxes for CPP account for one-fifth of all employer payroll taxes. The employer cost impact has increased in the last decade, but historically the CPP payroll tax has been smaller than that of our major trading partners. On the other hand, the contribution of employers will rise significantly. According to draft legislation to amend the Canada Pension Plan tabled on February 14, 1997, contribution rates will rise over the next six years to 9.9% of contributory earnings and then remain steady. These contributions are split equally between employer and employees, so employer contributions will be slightly below 5% of contributory earnings. The self-employed pay the full amount. J. No Major Unintended Impacts of CPP Were Found, But some Anomalies Exist Evidence on the impacts of the program features we examined offered no major concerns for unintended consequences, although there were some anomalies:
K. Comparison of CPP Costs to Private Insurance Not Conclusive It is difficult to compare actuarial and administrative costs for survivor benefits with comparable benefits provided by the private sector, because of differences in the population covered and the variety of plans available. However, it seems likely that the costs of providing CPP benefits are slightly higher due to its coverage of a broader population. CPP has comparable administrative costs to those of large pension plans and group life plans and significantly lower administrative costs than individual insurance. L. Several Program Changes Warrant Serious Consideration Many suggestions for changes, both major and minor, were elicited from key informants, including major stakeholders, the expert panel, as well as CPP program officers. Input on alternatives was also provided by surveys of survivors and of the general public. Ideas were also provided by reviewing how Canada compares with selected other countries. We concluded that the basic structures examined should be retained. However, based on the input received, we have identified a substantial list of changes that are worth pursuing. The most important are listed below, along with a brief statement of the rationale for each change.
Several possible changes were seen as not worth pursuing, most notably alternatives relating to pre-retirement survivor benefits that would include eliminating the age and family status distinctions; establishing a separate program; and eliminating the flat rate component. Nor do the findings support instituting a dynamic dropout based on future expected labour market conditions or linking the dropout to specific causes, e.g., further education, unemployment. The alternatives identified as worth pursuing create benefits from "modernizing" provisions of the CPP to respond to the changed context. They appear not to create countervailing negative effects, especially increased cost and administrative complexity.
A. The Scope of This Report This study is part of a larger effort which encompasses the review of the retirement benefits and disability benefits under the Canada Pension Plan (CPP) being conducted by Human Resources Development Canada (HRDC). It is expected to contribute to the reshaping of all the elements of the CPP to meet the needs of the Canadian population into the future in an efficient and cost effective manner. Prior to the introduction of the CPP in 1966, the government’s direct role in providing retirement income to retired workers was limited to the flat rate Old Age Security program. There were no specific provisions for payments to survivors of workers, although surviving spouses over the age of 65 would have received an OAS benefit in their own name. Provision for survivors was available either through private pension plans, or through insurance policies purchased in the private insurance market. The Canada Pension Plan was designed to make income available to retired and disabled workers, surviving spouses of deceased contributors and orphans. Virtually all workers in Canada, whether employees or self-employed, were required to contribute to the system. Quebec has its own scheme in lieu of CPP. CPP benefits are earnings-related and indexed annually to offset the effect of inflation. The subjects for study in this report fall into two broad categories:
This phase of the CPP evaluation is aimed at establishing whether there is a continuing rationale for the provision of survivor benefits and of other features of the CPP, to explore the success of this component in achieving its objectives, to examine how survivor benefits fit into the whole retirement and income security system and, finally, to examine possible alternatives. B. The Sources of Evaluation Findings This evaluation combines information derived from many sources. A literature review was conducted to provide the context for the evaluation by surveying issues surrounding survivor benefits and the ancillary features of the CPP. Among other topics we reviewed the relevance of survivor benefits in the current labour market and different commentator’s views on the continuing need for a range of program features. This was followed by a special literature review of the labour market and competitiveness impacts of the employer/employee payroll tax that finances the CPP. In addition, we conducted an international comparison of six countries, selected to offer a broad spectrum of generosity in the treatment of survivor benefits and other relevant features of the CPP. Interviews with federal program officials, provincial officials, and a range of stakeholder groups were conducted with respect to the full range of evaluation issues. Opinions were analyzed to identify those issue areas where there appeared to be significant consensus among knowledgeable informants and those areas where there was either uncertainty on the subject matter or differences in viewpoint.
Quantitative analysis conducted for this evaluation offered other findings:
A panel of experts reviewed our principal findings and discussed with us their own views relating particularly to the continued relevance of survivor benefits and the pros and cons of certain alternative approaches to both survivor benefits and other components of the CPP. This report integrates the qualitative and quantitative findings and indicates when findings are supported by multiple lines of evidence and where, also, there is either less conclusiveness in public and expert opinion or less certainty with respect to the appropriate interpretation of available data. C. The Structure of This Report The next chapter describes the historical development of CPP survivor benefits and selected CPP features of interest. Then the current program coverage and continuing rationale are examined. Chapter 4 examines the extent to which objectives have been achieved and addresses the key issue of the significance of survivor benefits in relation to all other sources of income, including personal income and, where applicable, that of other members of the household. Impacts and effects of the survivor benefits and other features of the CPP are examined in Chapter 5, including an analysis of the economic effects of CPP contributions on employers and employees. The last chapter considers possible alternatives to the current policy, distinguishing clearly between aspects that affect Canadians of "pre-retirement" age and those in "post-retirement." It ends with a summary of changes worth pursuing and those not worth pursuing.
A. The CPP in 1966 The Canada Pension Plan was introduced in 1966, in parallel with the Quebec Pension Plan (QPP). The Plan provided for the first time a public, earnings-related retirement income, together with ancillary benefits such as disability benefits and survivor benefits. The purpose of the CPP was to make reasonable minimum earnings replacement available to all workers at retirement, if they were disabled, and to their dependants in case of death, up to an earnings ceiling. The Year’s Maximum Pensionable Earnings (YMPE), which is the earnings ceiling, has been approximately equal to the average, annual industrial wage. CPP survivor benefits were implemented as part of the original CPP in 1966, when most married women usually had little experience working outside the home in the paid labour force. Many employed Canadians did not have access to employer-sponsored pension plans, and many of those who did were subject to quite varying benefits. It was believed that most widows and dependent children would require assistance, as they could not adequately support themselves with employment earnings or income from other sources, such as investment income. Therefore, the federal and provincial governments agreed that survivor benefits under the CPP and QPP should ensure that the female spouse and dependent children of a male contributor would have a measure of earnings protection in the event of the contributor’s death. CPP survivor benefits, like the entire CPP, have always been financed through compulsory contributions of all employees and employers (including self-employed workers) and from the investment earnings of the CPP Investment Fund. The system was designed to be on a pay-as-you-go basis, with a small reserve fund in the long run. Initially, however, income exceeded outgo, and the excess funds were maintained in the CPP Investment Fund. The CPP is a defined benefit plan in that CPP survivor benefits paid out, as well as CPP retirement and CPP disability benefits, are based on a formula driven by the level of earnings and length of contributions made by the worker. A ten year transition period was established for retirement benefits to deal with the first generation of recipients for whom the CPP was in place for only part of their working lives. CPP survivor benefits have always had three main components: Death Benefits, widow’s/widower’s benefits (now called Surviving Spouse’s Pension Benefits), and Orphan’s Benefits. Death Benefits provide a lump sum benefit equal to the lesser of six months of the deceased’s CPP pension or 10% of the Year’s Maximum Pensionable Earnings. The death benefit is subject to a minimum qualifying period of contributions of three to ten years. Widows and widowers now receive benefits based on the same criteria. Originally, eligibility for widower’s pensions was restricted to men who were disabled and substantially dependent on their wife for financial support at the time of her death. Widows qualified for a survivor pension if they had dependent/disabled children or were over the age of 35, as long as sufficient contributions had been made by the husband. Surviving spouses over age 65 receive up to 60% of their deceased spouse’s retirement pension. Surviving spouses under 65 receive a flat rate portion plus 37.5% of the contributor’s retirement pension. Prior to 1987, both widows and disabled widowers were not eligible to continue to receive CPP survivor benefits when they remarried. Orphan’s benefits are now paid on a flat-rate monthly basis for each dependent child of the contributor at the time of death. However, the original CPP allowed only four children to receive the full orphan’s benefit pension ($25.50 in 1966); the fifth and subsequent children could receive only half the benefits of the first four ($12.75 in 1966). Since it was introduced, the CPP has also included a general dropout provision, whereby the contributor’s lowest earnings years are omitted from the calculation of lifetime earnings. This calculation is the basis of all the earnings-related pensions and, consequently, determines Surviving Spouse’s Pension benefits. Up to 15% of the years in the total contributory period (ages 18 to 65, or to the retirement age, if the contributor retired between age 60 and 65) can be dropped out. B. Reforms to CPP Survivor Benefits and Other Features - The Main Effects Since the introduction of the CPP and QPP, several changes have been made to both systems. The key changes are described in this section. In 1973, the QPP increased survivors benefits substantially, almost tripling them, to make up for reduced purchasing power of the pension benefits resulting from the high inflation of the late 1960s and 1970s. The CPP did not adjust survivor benefits that year, resulting in substantial differences in benefits between the two systems. The CPP did index all benefits to the Consumer Price Index in 1974, but in some cases major discrepancies remained between the QPP and CPP flat rate payments for orphan’s benefits. Orphan’s benefits under the CPP were somewhat higher and fully indexed for inflation. QPP benefits for orphan’s were lower and not indexed for inflation. In January 1975, CPP survivor benefits changed such that survivors of female contributors received the same benefits as survivors of male contributors, without the need to prove substantial dependence on the female contributor. Eligibility for being deemed a surviving spouse was also made easier. Credit splitting was introduced in 1978 under the CPP, one year later than in Quebec. CPP pension credits earned by either spouse during the years of cohabitation are split equally between husband and wife upon divorce or annulment of marriage. Each spouse receives half of the couple’s total pension credits, regardless of their individual contribution. This amendment to the CPP affects retirement benefits as well as survivor benefits in that credits transferred from an earlier marriage may be part of the calculation for benefits paid to surviving spouses and children from a subsequent marriage. Another dropout provision to recognize child rearing was added to CPP in 1978 (although not implemented until 1983 when sufficient provincial approval was received). Spouses who leave the paid labour force to raise children under the age of seven (or who had below average earnings during such periods) can "drop out" those years from the calculation of life-time pensionable earnings, if it is advantageous to do so. This effectively raises the average pensionable earnings, which in turn determine benefit levels. Orphan’s benefits were also adjusted in 1978 to eliminate the restriction on benefits paid for orphaned children who were the fifth or subsequent child in the family. The period from 1976 through 1986 witnessed a major pension reform debate on all aspects of the CPP, as well as other elements of the pension system. The pension reform process attempted to deal with virtually all aspects of the CPP including funding, equality, Charter of Rights issues, and financial stability. The process of review led to a range of task forces, conferences, reports and federal-provincial agreements, which examined such ideas as substantially increasing benefits paid to survivors over 65, and transition payments for survivors under age 65. Eligibility rules relating to common-law spouses, remarried spouses, same-sex couples, ex-spouses and adult dependants were also examined. Only some of these issues were actually addressed by Bill C-116, which was passed by Parliament in 1986 to come into effect the next year. As a result of the Bill, survivor benefits were no longer terminated upon re-marriage. Combined benefits (paid to a survivor with CPP retirement or disability pensions) were also made more generous. The earnings-related portions of combined retirement and survivor benefits were stacked, subject to a ceiling of one maximum retirement pension, for recipients over age 65. For those between 60 and 65, the flat rate component of the survivor’s benefit was added to their actuarially-adjusted retirement pension and the earnings-related portion of the survivor’s benefit. For combinations with disability benefits, the higher of the survivor or disability flat rate was paid. The combined survivor and disability payments were allowed to exceed the maximum retirement benefit ceiling. Another survivor benefit reform introduced in 1986 was that orphan’s benefits would be paid regardless of the status of the surviving child. In particular, the benefit continued to be paid even if the surviving child married, whereas prior to this amendment, marriage of the child would disentitle him or her to further benefits. Changes were also made in regard to entitlement of adopted children on the death of a natural parent. Reconsideration of survivor benefits policy continued even as the new rules were being given effect. A Parliamentary Task Force on Pension Reform noted that demographic changes—in particular, the increased participation of women in the workplace—suggested the need for a specific study on survivor benefits. A consultation paper, published in 1987, pointed out:
The report went on to make proposals for changes to four main elements: implementation provisions; a transitional benefit structure for current surviving spouses; a new benefit structure for future surviving spouses; and increased children’s benefits. The proposals were subject to Parliamentary Committee review, public consultations, federal-provincial consultations and consideration of Finance ministers, but were not implemented. In 1991 an amending bill to the CPP was passed by Parliament that contained increases of 30% to orphan’s benefits and benefits for children of disabled contributors. However, other changes to the Surviving Spouse’s Pension were not agreed upon and were not included in the 1991 amendment. Reconsideration of survivor benefits has continued since that time, but no further amendments have been made. Retirement benefit payments under CPP increased sharply between 1973 and 1986, rising from a low of 43% of total net payments to 66% in 1986. This period corresponds to increases in the retirement beneficiary population as the plan matured, to the liberalizing of pension eligibility, and to the improvement of benefits. Since 1986, retirement benefits have consistently accounted for almost two thirds of the total payments. Recently, the share of total payments accounted for by retirement has fluctuated between 65% and 67% from 1991 to 1996. Survivor benefits combined with orphan’s and death benefits made up 16% of total net CPP payments in 1994. This represents the lowest percentage in the period from 1970 through 1994 attributable to these components of the CPP. As shown in Exhibit II-1, the portion of total payments allocated to survivor benefits (including the Surviving Spouse’s Pension, orphan's and death benefits) has fallen steadily since 1970. Since 1989, the survivor benefit category of CPP payments has stabilized at between 16% and 17% of total payments. In this category, the Surviving Spouse’s Pension comprised 13.6%, orphan's benefits were 1.3%, and death benefits made up 1.5% of total net payments in 1994. Orphan’s benefits have accounted for a smaller portion of total payments than death benefits since 1986. Prior to 1986, orphan's benefits were always a greater portion of total payments than death benefits. These trends are accounted for by the maturing of the over-65 beneficiary population and the lower birth rate since the baby-boom.
EXHIBIT II-1CPP Benefit Components, Percentage of Payments Source: Data provided by Planning and Strategic Studies, Human Resources Development Canada, January 1995.
A. Introduction This chapter reviews the rationale for survivor benefits and other features of the CPP. The social, demographic and economic conditions in Canada have evolved dramatically since the mid 1960’s, when the basic provisions being studied were implemented. Conditions have even changed since the 1970’s and 1980’s when new features to the CPP (e.g. flexible retirement, child rearing dropout, credit splitting on marriage breakdown) were implemented. The chapter is organized as follows:
B. Original Rationale for Survivor Benefits and Other Ancillary Benefits in General 1. Survivor Benefits The stated reason for the introduction of the CPP as a whole was to provide a measure of protection to all Canadian workers and their families against the loss of earnings due to the death, disability or retirement of the worker. The CPP survivor benefits were developed to address the potential financial difficulty faced by the surviving spouse and dependent children in the event of the death of the main income provider in the family. The primary role of the CPP is earnings replacement, as indicated by the Minister of National Health and Welfare at the time of the CPP’s introduction. The Minister summarized the issue as follows: "What people need, if either retirement or the death or disablement of the head of the family removes their regular income, is related in part to the level of earnings to which they have been accustomed." Income replacement provided by the CPP was tied to earnings, not solely to the contributions made to the program. There are both elements of "income distribution" and "insurance" that weaken the strict relationship between contributions made by the contributor and benefits received by a beneficiary. For example, contributors pay contributions only on earnings in excess of the Year’s Basic Exemption (YBE),1 while benefits are based on the entire earnings rate, up to the Year’s Maximum Pensionable Earnings (YMPE). The table below illustrates this effect for two sample salaries:
The insurance element in the case of disability and survivor benefits is reflected in the flat rate component. This is independent of the contributor’s earnings record and is payable once the contributor has met the eligibility criteria for the benefit. Therefore, low income contributors, disabled persons, and survivors in many cases receive a much larger benefit, as a proportion of contributions paid, than other beneficiaries. The provision of CPP survivor benefits may also act as a substitute for life insurance for the family’s main income earner. For young families who have relatively higher demand for life insurance, survivor benefits may be viewed as having a life insurance component as well as income support provisions. Substitution of CPP for life insurance would be most prevalent in lower income families. Prior to the introduction of the CPP in 1966, private pension plans were becoming an increasingly important part of the negotiated compensation packages provided in the private sector. Private plans were growing in importance, particularly in Ontario, and would have become more prevalent in the absence of the CPP according to the Report of the Ontario Task Force on Inflation Protection for Employment Pension Plans (1988). By 1960, 34% of Canadian workers participated in private pension plans; coverage increased to 38% of workers by 1965. More recent data from Statistics Canada on private pension plans indicate that 44.6% of paid workers were covered by private pension plans in 1993. Of the total labour force, only 35.4% were covered by private plans.2 Statistics Canada’s Ageing and Independence survey, 1991, shows that men were far more likely than women to have a job-related pension. Data from the Survey of Surviving Spouses conducted for this evaluation shows that 45% of the female survivors reported having some income from private pension plans, although the amount (proportion of total household income) was not established. Private registered retirement savings plans (RRSPs) do provide for surviving spouses in that the RRSP can be transferred to the surviving spouse when the owner of the RRSP dies. RRSPs are not used by the majority of the Canadian population. The 1993 Canadian Institute of Actuaries report cites a survey that found 35% of Canadians contribute to RRSPs and Statistics Canada reports that only 26% of all taxfilers contributed to an RRSP in 1993. 3Revenue Canada data for 1992 show that, at all age groups, a smaller percentage of women than men contributed to RRSPs. Data on withdrawals from RRSPs indicate that many Canadians use RRSP funds before they retire, thus weakening the effectiveness of RRSPs as a retirement income protection instrument. With respect to income from RRSPs, our survey of CPP beneficiaries indicated that in 1995, 25% of the female survivors had some income from an RRSP. Workers’ Compensation programs were an earlier form of social security that offered coverage for survivors and children. These programs have existed in all provinces since 1950, well before the Canada Pension Plan. However, Workers’ Compensation provides benefits for survivors only in cases of work-related causes of death. 2. Surviving Spouse’s Pensions "In the social context in which CPP was introduced, a man was considered to be the family breadwinner, and was expected to provide financial security for his wife and children. Married women, for the most part, were expected to be homemakers, without earnings of their own. For this reason, survivors’ benefits were designed to help widows and orphans. However, widows under age 45 were presumed capable of finding gainful employment, unless they were disabled or had dependent children in their care."4 In the case of widows age 65 or over, the survivor pension was 60% of the husband’s pension. In addition, she would have received $75 per month in 1966 from OAS. This plan design was consistent with major public sector employee-spousal pension plans at the time. Widows under age 65 would receive 37.5% of their husband’s entitlement, plus $25, provided they were over the age of 45 and with dependent children. Younger widows without dependent children or not disabled would receive a pro-rated amount (no benefit if they were under age 35). This design is not consistent with that for other pension plans—the rationale seems to be that the Surviving Spouse’s Pension for survivors under the age of 65 is partially "earnings replacement" and partially "income support". In fact, the flat rate component clearly simulates a proportion of the OAS that survivors over the age of 65 would have received. Subsequent changes to the CPP all recognized that the notion of all married women as homemakers was obsolete and removed the distinction between male and female survivors. However, the age, presence of a disability, and family status criteria remain. 3. Orphan’s Benefits The Orphan’s Benefit was a flat $25 per month and was seen as insurance in the event of the death of the father. Originally only one benefit per child was payable, even if both parents were CPP contributors. This was amended in 1987, indicating a subtle change from "insurance" to "entitlement". A limit of two benefits per child was substituted. It is likely that this benefit was introduced based on provisions in existing social security programs in other countries and other pension plans in Canada. For example, the Public Service Superannuation Act (PSSA) provides children’s benefits to deceased contributors to age 21, or age 25 if they are in full time education. This design was also consistent with the view of family benefits and the need to provide for children on the death of the breadwinner, prevalent at the time. 4. Death Benefit Health and Welfare Canada’s 1992 review of the CPP offers a rare glimpse into the rationale for the Death Benefit component. "A benefit payable on the death of a contributor was considered desirable for reasons of fairness and practicality. Even if there were no survivors, it would be only fair to provide some return of the contributions that had been made to the Plan which might be used to cover funeral expenses. If a pension were already in pay, the application for a death benefit would be the notification that it should be terminated."5 This quote indicates that the rationale was two-fold:
Reference is also made to "funeral expenses", but this does not seem to be its main rationale. 5. Dropout Provisions Since its inception in 1966, the CPP has permitted contributors to exclude (or "drop out") the years of lowest earnings from their contributory period. Since inception, this general dropout provision has allowed 15% of years with lowest earnings to be dropped out in calculating C/QPP benefits. Given the maximum contributory period will ultimately be 47 years for those retiring at age 65 or more (65-18), the general dropout provision could mean that full benefits would be paid if 40 years are worked. The child rearing dropout provision was added to the CPP in 1978 (although implementation was delayed until 1983 due to provincial opposition). This provision permits parents to drop out additional years for raising children up to seven years old. The number of years dropped out for child rearing is not limited to 15%; it is determined only by the number years out of the workforce, or in which earnings are below the contributor’s average while caring for children under seven. The rationale for the child rearing dropout provision was to avoid penalizing contributors (mostly women) who left the workforce, or who had years of low earnings while caring for children under the age of seven. It is not clear whether there was a more specific objective of ensuring that such contributors’ benefits would approximate those of contributors who had not left the workforce. In this case, this specific objective is not being met in a number of cases, as will be explained later. The Canadian Institute of Actuaries comments that the general dropout provision makes sense for many contributors who often do not start working until their early twenties, or who miss some period of employment, or who must retire early. They also recognize that the child rearing dropout is valuable in that it addresses concerns about lack of coverage for homemakers. Dropout provisions effectively raise pension (and therefore survivor) benefits. Pensionable earnings are calculated based on average lifetime earnings. Omitting low earnings years raises the average on which CPP pensions are calculated. The dropout provisions are to permit contributors to disregard certain periods of low or zero earnings, thereby preserving the value of the pension earned outside these periods. Flexible retirement in the CPP has effectively shortened the potential contributory period. For example, a full pension (although reduced by the early retirement reduction) is available after 35.7 years for a person who retires at age 60. This and other changes (child rearing dropout and credit splitting) have had some unexpected impacts on the dropout provision. These are further discussed in Chapter 5. 6. Credit Splitting Credit splitting refers to the division of CPP pension credits 6between members of divorced or separated couples. Credit splitting is done by adding together all pension credits of both spouses for each year they cohabited, and dividing the credits equally between them. Credit splitting is mandatory for most divorcing couples. It applies to legally married and common-law couples that have lived together for one year or more. The split credits are not actually paid to either spouse but are credited to the individual’s Record of Earnings which determines retirement, survivor, disability or children’s benefits. Credit splitting was first introduced in 1978 to address the increased frequency of marriage breakdowns. Credit splitting recognizes the reality that pensions are marital property that should be divided equally on marriage breakdown. 7. Assignment of Benefits The rationale for the assignment of a pension in pay would appear to be based on an argument of equity — non-separated spouses should not have fewer options than separated spouses. In any case, this provision seems mainly to have the effect of allowing income splitting for income tax purposes. However, it does provide a spouse who had little or no attachment to the workforce a pension in his or her own name, as is currently the case with OAS and GIS and in the proposed income-tested Seniors Benefit. C. Changes in Social, Demographic and Economic Conditions It should be recognized that most of the benefits being studied in the evaluation are primarily of interest to women: 90% of survivors are women; orphan’s benefits are mostly paid to children who have lost their fathers; and credit splitting is mainly intended to benefit women. The general dropout provision is of greater benefit to those with erratic earnings, as compared to those with a more level earnings record. Women tend to have more erratic earnings than men, generally speaking. In the mid-1960’s the typical Canadian family pattern consisted of a male breadwinner and a non-working wife taking care of children. The death of the breadwinner would not only put an immediate financial strain on the family unit, but would also compromise the retirement savings of the couple. Exhibit III-1 identifies some of the key changes that have occurred, in recent years, together with impacts such changes might have on the rationale for various components. In the next section, we examine the rationale for the program given the social changes that have occurred.
EXHIBIT III-1 Key Changes in Canadian Family Patterns Affecting Program Rationale
D. Analysis of Evidence for Continued Rationale 1. Surviving Spouse’s Benefits
2. Orphan's Benefits Orphan’s Benefits are payable as a flat rate to children of a deceased contributor. The rationale is clearly to provide support for children of deceased contributors. They are payable unconditionally to age 18, and as late as age 25, contingent on the child being in full-time education. Far fewer respondents expressed opinions about the rationale for orphan’s benefits, as compared to surviving spouse’s benefits. On the other hand, those who did saw them in the context of "family benefits", which are particularly needful in case of death where the contributor has a young family. As noted above, the level of survivor benefits was originally established with the entire range of social benefits in mind. Childrens’ benefits have undergone the most radical change to date, from a "universal" benefit to a benefit aimed at lower income families and no benefits at all for higher income families. In the light of the changes, panelists felt that the rationale for an orphan’s benefits in the original form should be reviewed. The rationale for contingent payment after age 18, based on continuation of education, was questioned by some. Although the rationale is clear (to support students of a deceased contributor), uniformity of age of cessation was cited as a reason for eliminating this provision. However, the counter argument would stress the value of investment in human capital. While there was not much discussion of these benefits, it was generally felt that they are not costly; so there was not a strong consensus to eliminate them either. 3. Death Benefits The original rationale, namely, to provide a basic, minimum payment and to encourage the reporting of a death, has not changed. However, very little discussion of this benefit was engendered, either with the expert panel or the interviewees. Evidence from the international study was mixed—with some countries having a lesser benefit than Canada, while others have substantially greater benefits. In summary, a number of respondents questioned the need for this benefit in today’s society. Those who think of it solely as a subsidy of funeral expenses think that few, besides the totally indigent, would not have enough money available for a simple funeral. Nonetheless, as with orphan’s benefits, the death benefit was not considered a significant cost within the total CPP either and few recommended its elimination. 4. General Dropout Provisions The general dropout of 15% of the years of lowest earnings during the contributory period has been a feature from the inception of the CPP. While its intention was to eliminate the negative impact on the CPP pension of years of absence from, or weak attachment to, the workforce due to illness, unemployment, and continued education, these specific grounds were not incorporated into the Plan. The dropout was designed on a "general" basis. The specific dropout was for months while on CPP disability pension. (The corollary of this is that virtually the only contributors receiving a 100% CPP retirement pension are those who have experienced considerable time on CPP disability. In other words, it is extremely difficult to qualify for a 100% pension even with the general dropout.) Changes to the CPP since inception, principally flexible retirement ages, and introduction of the child rearing dropout and credit splitting have produced some unexpected impacts on the general dropout. More discussion of these issues, together with examples, are found in Chapter 5. Some of the stakeholder interviewees seemed to be aware of these issues, although others minimized their importance. Among those who responded, there was questioning of the basic rationale for the dropout provision—of its place in a contributory social insurance plan. Others, however, supported the original rationale. The expert panel was very supportive of the initial rationale, and indeed felt it was even more required in the current and evolving economic conditions of a less secure employment, more non-standard work and difficulty experienced by young people in entering the workforce in the first place. There was no support for linking the dropout to specific causes of absence, other than child rearing dropout, as at present. Also the panel did not seem to be concerned by the anomalies mentioned above. However, it was suggested that the 15% figure be reviewed in the light of changing economic and social conditions (see Chapter 7 for more discussion on this point). In relation to comparator countries, a dropout expressed as a proportion of a fixed contributory period would appear to resemble the systems indicated in our international review more closely. As the report on the international review points out, direct comparison with the CPP dropout system is quite difficult. The U.K. probably has the closest design, with a fixed 20 year contributory period. The U.S. would also be categorized in a similar manner, although the benefit design in terms of earnings levels and years of contribution is considerably more complex than in the CPP. 5. Child Rearing Dropout Provision The provision of a child rearing dropout was not part of the original CPP design. It came into effect in 1983 retroactive to 1978 as a response to higher participation of women in the workforce and the need to protect contributors who left the workforce to raise young children against a negative impact on their CPP pension due to low or zero-earning years. It should be noted that there is no limit to the number of years that can be dropped out for this purpose, as compared to the limit on the general dropout provision. There was much support for the continuation of this provision, as it is a means to encourage women to remain in the workforce in the longer run, while taking time out to raise a family. Others highlighted some concerns with this provision, for example:
In spite of these shortcomings, support for the continued rationale was particularly strong among the expert panel, somewhat more mixed among the key informants. 6. Credit Splitting Credit splitting on marriage breakdown was introduced in 1978, 12 years after the inception of the CPP. This provision recognized the evolving social reality of higher divorce rates and lesser probability of life-time unions, as compared to earlier epochs. Subsequent amendments to this provision had the effect of giving recognition to common-law unions, and separation as well as divorce or annulment. Other amendments effectively made this provision mandatory instead of voluntary, at least in principle. On separation, credit splitting is only by application to the CPP. Provinces may decide whether this provision can be applied on a voluntary basis, allowing couples to trade-off CPP entitlements in the same way as other marital assets. Currently, only Saskatchewan and British Columbia have taken advantage of this provision. The situation in Québec would be similar to that in these two provinces, although the Civil Code differs somewhat from the common law in effect in the other nine provinces. There was strong support for the rationale for this provision, given the continued high rates of marriage breakdown and general acceptance of sharing of marital property. There is also strong support for a mandatory approach. It should be noted that a significant proportion of the interviewees were not aware that it was already mandatory, and recommended that the provision be "changed" to make it mandatory. This reflects the reality that, to date, awareness of the provision is low and effective enforcement/communication mechanisms have not been found. It is interesting to note a split in opinion regarding the mandatory nature of the credit split between experts and key informants on the one hand and the general public (as indicated in the various surveys) on the other. The general public appears to support "choice" in general and opposes mandatory provisions which limit choice. The experts, on the other hand, recognize that parties (especially women) may be poorly informed about the value of their entitlement and might be inclined to trade it away too easily. The experts feel that limitation of choice was warranted in this case due to the asymmetric knowledge and bargaining power of the two sides of marriage breakdown. The international study showed that no other country has implemented a credit splitting approach. However, some countries do pro-rate survivor benefits, while others are considering credit splitting provisions. 7. Assignment of Benefits Even in the absence of a marriage breakdown, spouses in receipt of a CPP pension can assign each other half of their pension. There was not much discussion of this particular provision. Where it was discussed it gave rise to mixed feelings. On the one hand, it was seen as purely an income splitting device to reduce income taxes. The rationale for this was questionable, especially as "special deals" for seniors seem to be coming to an end. On the other hand, it is viewed as giving a spouse who may not have his or her own income a right to a pension in his or her own name. This is similar to the current OAS and GIS provisions as well as the proposed Seniors Benefits. E. Summary In spite of changing conditions there is strong evidence for a need for surviving spouse’s benefits. There is strong evidence for the rationale of post-retirement survivor benefits in its present form. There is consensus for continuation of pre-retirement survivor benefits in their current form, but it is weaker than the consensus in regard to post-retirement benefits. Flat-rate benefits and age/family status criteria are questioned by some experts. Current beneficiaries tended to support the status quo when surveyed with respect to whether benefits should vary with the age of the surviving spouse and whether benefits should be related to the survivor’s income. Insofar as orphan’s benefits and the death benefit are concerned, far less discussion was elicited from the expert panel and key informants. Current beneficiaries support the current rules. By and large, experts were less supportive of the rationale than in the case of spousal survivor benefits. Conversely, there was little support for the elimination of these benefits, partly because of the insignificant savings that this would bring about. Adequacy of the benefit was validated in a qualitative sense, in relation to original, implicit replacement target, but interaction with other evolving social programs was questioned. Finally, strong support was shown for continued provision of survivor benefits from the federal source (i.e. CPP). The continued rationale for the general dropout provision was supported, even enhanced in view of labour market instability, and in spite of changed labour force participation of working mothers, evidence for continued need for the specific child rearing dropout provision was offered. Strong support was shown among experts for the rationale for continuation of credit splitting on marriage breakdown. Assignment of pensions in the absence of marriage breakdown was questioned.
In this chapter we profile current beneficiaries of a Surviving Spouse’s Pension (SSPs) and examine the extent to which the benefits contribute to the program objective of offering a measure of earnings protection to the survivors of a deceased CPP contributor. We then examine the appropriateness of different eligibility rules by examining the opinions of current survivors and the general public, and by simulating the effects of possible changes to the rules. A. Profile of Current Beneficiaries
1. Administrative Data Sources The number of recipients of CPP survivor’s pensions has increased rapidly over the last three decades, from less than 100,000 in the early 1970s to 735,345 in January 1996. In that month, 88.9% of the beneficiaries were women, and of the women 72.4% were over the age of 65. Exhibit IV-1 shows that the number of beneficiaries under the age of 45 is very small indeed. The total dollars expended for Survivor’s Pensions in 1995, including those to survivors living abroad, was $2.2 billion. In the month of January 1996 the domestic benefit expenditure was $180.7 million, with an average benefit of $244.01. Younger female beneficiaries—those of pre-retirement age—represented only 27.6% of the number of beneficiaries, but represented 32% of the dollar value of the benefits to female survivors. Details of the surviving spouses’ monthly benefit are shown below: Surviving Spouse’s monthly benefit:
EXHIBIT IV-1 Distribution of Survivor’s Pensions,* by Age and by Gender—January 1996
*Does not include QPP or benefits paid by the supplementary cheques system nor under international agreements on social security. Source: Canada Pension Plan, Statistical Bulletin, January 1996. Examination of annual benefits by age at start of benefits showed the same pattern for males and females: a tendency evident in the data from 1977 (the first year for which these data were examined) for men, and since 1982 for women, was for those who started benefits after age 65 to receive lower benefits than for those who started before age 65. The discontinuity of average benefits between the 60-64 and 65-69 age groups shown in the exhibit is caused by the change from the pre-retirement benefit, which consists of an earnings-related plus flat-rate benefit, to the post-retirement benefit, which is earnings related only. The drop in average benefits is particularly noticeable for males, where widower’s pensions are based on low earnings of their deceased wife. This effect is masked in the pre-retirement period by the presence of the flat-rate benefit. Exhibit IV-2 shows the trend in the number of new Surviving Spouse’s Pensions since 1984. Each year, until very recently, the number has grown quite steadily, for both males and females. This trend reflects demographic conditions among CPP contributors and their families. As of 1995, however, the data9suggest that the number of new benefits might be starting to stabilize, even to decline slightly. EXHIBIT IV-2 Number of New Surviving Spouse’s Pensions, by Gender and Year Number
Source:Canada Pension Plan Statistics. Analysis of other data shows:
EXHIBIT IV-3 Average Age at Start of Surviving Spouse’s Pension, by Gender
Source:Canada Pension Plan Master Benefit File extract.
2. Characteristics of Beneficiaries: The Survey Results The Survey of Beneficiaries of CPP Surviving Spouse’s Pensions provides characteristics of beneficiaries in terms of attributes not available through CPP administrative data. The respondents to this 1996 survey 11are derived from participants in Canadian Facts’ Canadian Family Opinion (CFO) Panel who indicated in an earlier survey that they were in receipt of a CPP Surviving Spouse’s Pension. The CFO Panel is a continuously-maintained panel, broadly representative of Canadians across the country. CFO panelists participate from time to time in responding to self-completion questionnaires on a variety of subjects. The response rate from qualifiers, i.e., those who said they were beneficiaries of a spouse’s benefit, was 81%. The results reported here have been weighted to reflect the known proportions of male and female beneficiaries—by province. We have no reason to believe that the results are biased in any respect that is likely to affect the usefulness of the profile or the interpretation of opinions with respect to CPP policy issues. We note, however, that the achieved sample may be somewhat better educated than the true population of SBs and may somewhat over-represent females with previous employment income.12 Reflecting the numbers presented in Exhibit IV-1, the majority of respondents to the survey are women: about a third are between the ages of 65 and 74 and fully 41% are 75 years of age and over.
Following is a profile of these female survivors:
Younger survivors (those under 45 years of age) were somewhat more likely to report receiving mortgage insurance or to say that the question was not applicable because they did not have a mortgage. B. Significance of Survivor Benefits in Total Family Income
Whether viewed by gender (Exhibit IV-4) or by age group, annual total incomes of recipients of Surviving Spouse’s Pensions, when expressed in 1996 dollars, have remained quite stable since 1977. EXHIBIT IV-4 Average Annual Total Income (1996$), by Gender
Source:HRDC Longitudinal Labour Force Data Base. Overall, the contribution of the Surviving Spouse’s Pension to beneficiaries’ incomes has also remained fairly stable over the past several years. Exhibit IV-5 shows the average percentage of total personal income accounted for by this benefit, for women and men, over the period 1979 to 1994. One exception to the overall trend occurred because legislative changes that took effect in 1987 resulted in increased benefit levels and a corresponding jump in the proportion of income attributable to the benefit. For women, the proportion before 1987 had held steady at around 16 per cent, then jumped to about 22 per cent in 1987. For men, the jump in 1987 was from about seven per cent to ten per cent. The percentage overall for men has been about half that for women because surviving male spouses tend to receive lower benefits and have higher incomes. The proportion for women dropped in the period 1992 to 1994, to about 18 per cent. The cause of this drop is unclear, but it is notable in that a similar pattern has not occurred for male beneficiaries. Without further analysis, it is not clear why these changes occurred. Combined benefit restrictions were made less stringent in 1987, but it is not clear that this fully accounts for the trends. EXHIBIT IV- 5 Average Percentage of Income Accounted for by SSP, by Gender
Source:HRDC Longitudinal Labour Force Data Base and CPP Master Benefits File. Legislative changes to the Canada Pension Plan increased the levels of Orphans’ Benefits in 1992. To the extent that such benefits were claimed as income by the surviving spouse, income would increase relative to the Surviving Spouse’s Pension and lead to a corresponding fall in the proportion of income represented by SSP. The observed drop in proportion in 1992 occurred mainly in the younger age groups, lending support to the hypothesis that changes to the level of Orphans’ Benefits were responsible for the drop. A trend perspective on the relationship of SSP to personal income is provided by looking at cohorts of survivors beginning to receive benefits. This analysis shows:
Legislative changes to CPP in 1987 caused a temporary increase in the proportion of personal income accounted for by SSP. The proportion has since decreased, slightly for males but quite markedly for females. For females, the peak occurred at 23 or 24 per cent, but the proportion has since dropped to about 18 or 19 per cent.
The data for each cohort of beneficiaries showed very slight fluctuations (usually only a few percentage points), but no consistent trend. Some cohorts appear to have reduced their dependency on SSP benefits over time while for others the dependency increases.
Considering the contribution of all CPP benefits to income, the patterns for male and female beneficiaries are not so dissimilar. In the mid-1980s, women depended more than males on all CPP benefits combined. But both before and since that time, the percentage of income accounted for by CPP benefits has not differed between men and women by more than two or three percentage points and has often been closer than that. Given the relatively older age of most recipients of SSP, this similarity suggests that men rely more on Retirement Pensions (as opposed to Disability Benefits) than do women, likely due to men’s traditionally higher rate of contributions. For the same reason, female surviving spouses receive greater SSP by virtue of their deceased husbands’ higher contributions.
As part of this evaluation we examined average annual total incomes for recipients of a Surviving Spouse’s Pension. The analysis reveals a different pattern when compared to the other labour force indicators. Although the trend over time is not as consistent as was the case for the other indicators, there is clear evidence that total annual personal incomes were generally higher after the start of survivor benefits, especially so for female beneficiaries. Exhibit IV-6 shows the average income for female beneficiaries who started to receive Surviving Spouses Pensions in 1988. This exhibit shows that their incomes have remained fairly stable over time, with perhaps a slight downward trend. The interest here lies in the jump in income levels associated with the death of a spouse and starting to receive benefits from the Surviving Spouse’s Pension, and perhaps from other sources as well, such as a survivor’s benefit related to a spouse’s employer-provided pension.
EXHIBIT IV-6 Average Income Per Year (1996 $), for Women Starting Benefits in 1988
Source:HRDC Longitudinal Labour Force Data Base and CPP Master Benefits File. Examination of the effect of coming into receipt of an SSP or other pension-related income is only a part of the picture, however. To appreciate the full effect, one needs data on the extent of income replacement of family income prior to receipt of SSP that, of course, includes the earnings of the now deceased spouse. This is examined later. The question can still be asked whether there is a need for spousal survivor benefits when the surviving spouse is working? To answer this question, we rely on assessment of the proportion of income accounted for by CPP survivor benefits as an indicator of need for benefits and compare its values by level of employment of the surviving spouse. Since the question is most salient with respect to women, our analysis focuses on females only. The analysis examined, by year, both the stock of current beneficiaries and cohorts of beneficiaries who started receiving Surviving Spouse’s Pensions that year. 13The first step was a simple comparison of the average proportions of income accounted for the Surviving Spouse’s Pension among female recipients who had no employment and those who were employed at least one week in the year. Not having employment is composed of two factors: being unemployed and not being in the labour force. For simplicity, we refer to these persons as unemployed. For the most part, we found the differences between the proportions of income contributed by SSP benefits for employed and unemployed female beneficiaries were between three and five percentage points, with less frequent differences both higher and lower than this range. But the average proportion of income accounted for by SSP benefits was less for employed beneficiaries throughout. This finding suggests that employed female beneficiaries rely less on SSP benefits as a source of income, although the differences are slight. For those who were employed, we pursued this analysis further by using regression models to assess the effect of additional weeks of employment on the proportion of income accounted for by SSP benefits. The analysis involved six models, one for the current stock of beneficiaries, and one for data from each of the five years after starting benefits.14 These results indicate that employed female beneficiaries generally receive a distinctly higher proportion of their income from SSP benefits and that each additional week of employment reduces the proportion by one-fifth to one-quarter of a percentage point. The model would lead one to expect that a beneficiary who was employed 30 weeks, for example would rely on SSP benefits for somewhere between 2.1 and 2.4 per cent less of their income than would be the case for a similar beneficiary who was employed for only 20 weeks. (The difference of ten weeks of employment is multiplied by the estimated coefficient for the variable weeks employed.) In a sense, these findings come as no surprise. Beneficiaries who are employed might be expected to have higher incomes than those who are not, whether the latter are unemployed, retired, or on welfare. Therefore, they should derive a relatively smaller proportion of their total incomes from the Surviving Spouse’s Pension. In any case, the observed differences in this proportion that are related to employment, while often consistent and statistically significant, seem not very large in relation to the average proportion. Evidence suggests that employed beneficiaries differ from those with no employment income with respect to the significance of benefits, but not greatly. There is, therefore, little justification for considering changes to the benefit structure or other changes to the program to accommodate such differences.
A major objective of the survey of beneficiaries using the Canadian Family Opinion panel was to establish the relative importance of the Surviving Spouse’s Pension in relation to the entire household income of the recipient. We have noted above that relatively few survivors were married or living common-law (8% of females) and relatively few were employed in 1995 (18% of females). We may expect, therefore, that employment income in 1995 from a new spouse or from their own employment will be a factor for only a minority of female beneficiaries. Other sources of income can be expected to vary. Exhibit IV-7 shows the sources of household income in 1995 reported by female panelists. In this exhibit the data are presented for female survivors less than 65 years of age and survivors 65 years of age and older. The exhibit shows that, in order of frequency reported, the five top ranking sources of household income for all female survivors other than Surviving Spouse’s Pensions (which by definition all respondents reported) are:
The Refundable Tax Credit and Employment Income (not necessarily their own) were the most frequently reported other sources of household income for pre-retirement women. For post-retirement women, OAS/ SPA and the Refundable Tax Credit, and CPP Retirement benefits were the most frequently reported. Orphan’s Benefits were mentioned only in multiple person households with survivor beneficiaries under age 65.
EXHIBIT IV-7 Sources of Household Income by Age—Females Only
Source: Survey of Beneficiaries, CFO Panel. Note: Percentages and totals based on respondents; five cases did not respond to any of the questions on sources of income. Exhibit IV-8 shows the distribution of a key analysis variable—Survivor Benefit as a Percentage of Total Household Income—which we will call "Proportion"15for female beneficiaries only. The average (mean) proportion is 19.3% of Total Household Income.
EXHIBIT IV-8 Distribution of "Proportion" for Female Survivor Beneficiaries Only
Source: Survey of Beneficiaries, CFO Panel. Analysis of detailed tables (not shown here) on the distribution of Proportion indicates that the Survivor Benefit represents less than 25% of total household income for as much as 74% of female SBs16. Exhibits IV-9 and IV-10 show the distribution of "Proportion" separately for females under and over age 65, respectively, displaying the relationship with total household income. This presentation focuses attention on the "pre-retirement" and "post-retirement" survivors. Each exhibit shows, for various levels of total household income, the average value of "Proportion" as well as the 95% confidence limits for each such average.17 Both exhibits show the expected relationship (since the Surviving Spouse’s Pension can never exceed a fixed annual limit), that "Proportion" is lower where total household income is higher. The two graphs are quite similar with respect to households with over $15,000 in total income. Below that level, however, female survivor beneficiaries under age 65 report higher average values of "Proportion". This is likely the result of the presence of an Orphan’s Benefit, which is of greater significance for those with lower levels of income. EXHIBIT IV-9 Survivor Benefits as a Percentage of Household Income—Females Under 65 Years of Age
Source: Survey of Beneficiaries, CFO Panel. EXHIBIT IV-10 Survivor Benefits as a Percentage of Household Income—Females 65 Years of Age and Over
Source: Survey of Beneficiaries, CFO Panel. For women with low household incomes—for example, less than $20,000 per annum—Survivor Benefits represent a very substantial contribution to the total. This is true for both pre- and post-retirement survivors.
Statistics Canada has linked taxation data for individuals into groups that approximate true census families. The resulting series of annual files is referred to as T1 Family Files (T1FF). We provided Statistics Canada with a set of specifications for linking data of recipients of survivor benefits with total family income from T1FF. Other CPP benefits were also taken into account. The chief contribution of the T1FF analysis to this evaluation is to confirm the proportion of family income represented by SB derived from the survey results, and to show that the proportion represented by SB has not changed much since 1987, although it has increased somewhat for the lowest income groups. In addition, it provides an answer to the question of the extent to which either the Survivor Benefit or SB plus other CPP income replace family income of the year prior to the death of the spouse. The data for the three years selected—1987, 1990 and 1993 (the last for which full data were available—yield a very large number of tables for analysis. Breaks were calculated by gender, age group, duration (years in receipt) of CPP benefits; family size and family composition. For purposes of this report, we focus on the results for women only and pay particular attention (as we did for the survey data) to women 65 years of age and older. Exhibit IV-11 shows the distribution of Proportion 1 (Survivor Benefits/Family Income) and Proportion 2 (Survivor Benefits plus other CPP Income/Family Income) for all female beneficiaries in 1993. In Proportion 1, Orphan’s Benefits are included where applicable—thus making Proportion similar to that reported in Exhibits IV-8 to IV-10. Other CPP-based sources such as Disability Benefits and CPP Retirement Pension are included in Proportion 2. The 1993 results, which show that "Proportion 1" for women with family incomes of $5,000-$7,499 averages 53%, are quite comparable to those shown in Exhibits IV-8 to IV-10 (from the survey results). The exhibit also parallels the data shown there for 1996 for higher levels of family income. Proportion 1 declines rapidly from 14% for family income of $20,000 to $24,999 to a low of 5% for family income in excess of $50,000 per annum. Proportion 2, that includes more CPP based income, is by definition consistently higher than Proportion 1. However, it too tails off to a low figure (7%) for those whose family income per annum is $50,000 or more. Other observations from Exhibit IV-11 are of relevance:
Exhibit IV-12 shows trends in Proportion 1 for females 65 years of age and older for 1987, 1990 and 1993. The downward level in proportion as family income increases is similar for all years to the pattern in Exhibit IV-10. The percentage of income represented by SB is slightly lower among post-retirement women than for all women. This is likely the effect of the flat rate component for pre-retirement SBs—as was noticed for other data as well. The trend over time is for the ratio of SB to Total Family Income to increase. This is especially evident for those in the lowest income groups. Exhibit IV-12 also presents data on Proportion 2, and here the pattern over time exhibits some differences from that observed for Proportion 1:
Exhibit IV-13 offers a slightly different perspective on the significance of SB. To this point we have examined the "stock" of beneficiaries of SB in each of the three years. Next we present data on the flow—or the new recipients of survivor benefits. The exhibit shows the proportion of family income reported in 1987, 1990 and 1993 that was "replaced" by SB (Surviving Spouse’s Benefit and Orphan’s Benefit) and by SB plus other CPP sources where payment of SB began in 1988, 1991 and 1994. EXHIBIT IV-13 Replacement of Family Income by SB or SB Plus CPP—All Females Who Started Receiving SB in 1988, 1991, 1994
Source: T1FF and CPP Master Benefit File. *Because total income is net of losses, it can be very small or even negative. Values less than $1 were converted to $1 for purposes of calculating the ratio, which led to uniformatively large values for the $0-$4,999 income group. The conclusions we draw are the same as for Proportion 1: above annual incomes of $15,000 SB by itself replaces less than 20% of income and the result is similar in all three years studied. Below $15,000, the replacement value of SB has grown over the three time periods examined. 6. Simulation Results Using MAPSIT a.Introduction As an additional perspective on the issue of the significance of SB in total family income, we used MAPSIT18 to produce a number of simulations to show how "Proportion" varies under a variety of hypothetical but realistic conditions. The difference between these results and those reported earlier from the survey is that survey respondents told us what proportion SB represented of gross family income; MAPSIT results describe the contribution of SB net of taxes and other transfers. To assess a wide range of possible results, we canvassed a number of scenarios: single seniors, young widows with children, and pre-retirement widows with no children—each with no benefits; one-half the maximum Surviving Spouse’s Pension; and at the maximum survivor’s pension. The analysis examined the effect of SB on disposable income while allowing income from other sources to vary from $0 to $100,000. A second set of scenarios included eligibility for social assistance or GIS to see what the net effect would be on family income. We concentrate on the results for the single seniors because, as we have seen, this is by far the predominant group of beneficiaries. We also examine closely the results for the young widow with children on social assistance. In general, the contribution of CPP Survivor Benefits to a household’s disposable income seldom exceeded 20%. b. The Single Senior Scenario The single senior scenario examines the results for a hypothetical individual with the following set of characteristics:
The analysis examines the effect on disposable income of the addition of two different levels of CPP survivor benefits while allowing income from private pensions and investments to vary. The first $1,000 of the latter is assumed to be private pension income, and therefore not subject to tax. With a maximum benefit of $5,235, the greatest increase in disposable income is $3,802 (or 73 per cent of the gross survivor benefit), when private pension and investment income is between $11,305 and $12,014. At half the maximum benefit ($2,617.50), the greatest increase is $1,901 (also 73 per cent of the gross survivor benefit), which occurs over a broader range of private pension income: between $11,305 and $14,631. Exhibit IV-14 shows the change in disposable income as a percentage of the gross benefit, by level of benefit for the single senior scenario where the beneficiary is eligible for GIS. This exhibit addresses the main issue surrounding the analysis: the value of the CPP survivor benefit to the household’s income. It considers the net CPP survivor benefit, after adjusting for the effects of transfers, and displays it as a percentage of the household’s disposable income.
EXHIBIT IV-14 Percentage Net Change in Disposable Income By Level of Benefit, Single Senior Scenario
Source: MAPSIT Simulation. We see that the CPP survivor benefit (that is the change in disposable income due to SB) accounts for at most about 20 per cent of the disposable income of this hypothetical type of household. This level is approached if private pension income is in the area of $12,000 and is available only if the household receives the maximum CPP survivor benefit. Elsewhere in the distribution of private pension income, the contribution to disposable income remains well below this level. If the household receives only half the maximum CPP survivor benefit, the greatest level of contribution to household income is 11 per cent, which occurs at a private pension income of about $11,500. For both benefit levels, the increase in disposable income attributable to CPP survivor benefits declines gradually above $15,500 of other income, due to higher rates of taxation. The exhibit also shows that over a certain range of relatively low incomes, the increase in disposable income was (perhaps) unexpectedly low. If the household receives private pension income of about $6,170, however, its disposable income increases by much less (only 8%). Within the range of private pension income surrounding this value, and in the presence of the survivor benefit, the household is not eligible for the provincial (GAINS) top-up to GIS, the GIS benefit is greatly reduced, and income is subject to both federal and provincial income taxes to an extent that is not the case in the absence of the CPP survivor benefit. A similar pattern occurs for households receiving half the maximum benefit (an increase in disposable income of only 3%). The above findings must be considered in relation to the likelihood that the conditions underlying this example would occur in practice. To assist this determination, the distribution of average gross family income of female survivors over the age of 65, drawn from the T1FF data file, is shown in Exhibit IV-15. EXHIBIT IV-15 Distribution of Total Family Income for Females 65 Years of Age and Older—1993
Source: T1FF. The point of maximum change in disposable income as a result of SB occurs at approximately $12,500 of personal private pension income, as shown in Exhibit IV-14. Taking into account OAS and CPP, this amount likely corresponds to about $20,000 in total family income. From Exhibit IV-15 we see that about one quarter of the 65+ female survivors fell into the family income range of $15,000 to $24,999 in 1993. This information leads to a conclusion that for most post-retirement women SB is not of major significance. Even for the one in four at low incomes for whom it is important, it does not exceed 20% of total income. c. Scenario Involving Single Parent With Two Children on Social Assistance In this scenario we examine the effect of SB on incomes of pre-retirement women with children. The household used in the scenario consists of a widowed female, age 33, with two children, ages 7 and 3, living in Ontario in a three-bedroom apartment. Total income, in current dollars includes:
The analysis then examined the effect of no SB; one half the maximum SSP and two Orphan’s Benefits; and the maximum SSP and two Orphan’s Benefit, assuming different levels of earnings. The presence of children in the household has a dramatic effect on the extent to which the SB increases what we term consumable income (disposable income less child care expenses). Analysis that compares scenarios with and without social assistance helped to isolate an unexpected result: when family earnings fall below $33,000—and especially between $14,000 and $20,000—the contribution of SB to household income is mostly negated by the presence of social assistance. In extreme cases, because of federal and provincial taxes paid, consumable income actually decreases as a result of the receipt of SB when the family is eligible for social assistance. In other words, the family would receive more total income through social assistance than through SB. How many cases in reality fit this modelled scenario? First, consider that earnings of $14,000 to $20,000 likely imply total income of perhaps $20,000 to $30,000, when tax credits, U.I. benefits, and other forms of income are considered. Exhibit IV-16 shows the distribution of single parent households under the age of 65. Among such SB recipients, 42% have incomes of less than $30,000 and about 17% fall into the critical range, where the presence of social assistance has its greatest impact. As a means of ensuring basic economic support levels for pre-retirement survivors of deceased CPP contributors who are eligible for social assistance, SB has little or no value over a range of household incomes that included many, although not a majority of, such households. EXHIBIT IV-16 Distribution of Total Family Income for Single Parent Households—SBs Under the Age of 65
Source: T1FF. d. Conclusion The contribution of CPP survivor benefits to household disposable income seldom exceeds 20%. One explanation for this may be that the effect of the CPP survivor benefit on disposable income is also very low, and sometimes negative, among survivors eligible for social assistance (or GIS) and income tax. If a household has a total income and assets such that it is eligible for social assistance (or GIS), the net change in disposable income that results from the CPP survivor benefit can be quite low because, without the CPP benefit, the household would receive an almost equivalent amount of social assistance. At the higher end of the range of such low incomes it can also happen that total income is large enough to be taxable. In this case, the household not only loses social assistance benefits equal to the CPP survivor benefit, but also pays taxes on the latter, resulting in a net negative contribution to disposable income. We recognize that the interaction of CPP survivor benefits and social assistance programs may be oversimplified in the scenarios we have run. Although CPP survivor benefits contain an element of social insurance, they are nonetheless based on an earned entitlement, while social assistance is a source of last resort for income support and is tested by the value of assets as well as income. Also, the scenarios presented here include only two extremes: those fully eligible and those not eligible for social assistance. Substitutability between CPP survivor benefits and social assistance is likely less fluid than it appears here, and will be further affected by various "self sufficiency experiments" for SARs that are in effect. Nevertheless, the phenomenon described is of interest and is described here as an anomaly of the program which is noteworthy. 7. Beneficiaries’ Perceptions of the Adequacy of Current Household Income In the Survey of Beneficiaries of a Surviving Spouse’s Pension, respondents were asked about the adequacy of household income both before and after the death of a spouse. For many survivors, a considerable number of years would have passed between the date of death of their spouse and their current recollections of the adequacy of the Death Benefit at that time. With respect to adequacy prior to the spouse’s death, women responded as follows: "more than adequate"—14%; "adequate"—68%; "less than adequate"—17%; and 2% did not respond. With respect to the adequacy of current income—i.e., after the death of a spouse—the pattern of response was: "more than adequate"—4%; "adequate"—59%; "less than adequate"—36%; and not stated—2%. Exhibit IV-17 shows, for female survivors only, the relationship between perceptions of household income before and after the death of the spouse. Examination of the columns in the table show where change of status has occurred: 29% of those with "more than adequate" pre-death income now say that their household income is less than adequate; 33% of those with previously "adequate" income now regard their income as "less than adequate." Predictably, the largest group of respondents among those with previously less-than-adequate income fall into the less-than-adequate category (54%) after the death of a spouse. Numerically, however, the largest portion of the female survey respondents with current household incomes perceived to be less-than-adequate are women whose income prior to the death of the spouse was regarded as "adequate." This group represents 22% of the sample of female beneficiaries.
EXHIBIT IV-17 Perceived Adequacy of Household Income, Before and After the Death of a Spouse—Female Survivors Only
Source: Question 7 and Question 10, Survey of Beneficiaries CFO Panel, 1996; 13 respondents did not reply to the questions. Who are the women who say their current household income is less than adequate? Data from the survey indicate that they are disproportionately numerous among:
C. Appropriateness of Eligibility Criteria for Survivor Benefits The question is often asked whether changes in the participation rate of women in the labour force today (and expected into the future) compared with that prevailing at the time of the introduction of CPP Surviving Spouse’s Benefits invalidates the current eligibility criteria. To qualify for a benefit, the surviving spouse must:
The questioning of the current system directs attention to a few specific issues:
The evaluation has assembled data and opinions on most of these eligibility issues from a variety of sources: interviews with key informants and a panel of experts; the survey of survivors; a survey of the general public; and analysis of a series of simulations using the CPP Actuarial model. 1. Key Informant Interviews and the Expert Panel Among key informants we interviewed, there was a strong disposition to state that the current eligibility rules for survivor benefits are inappropriate and that they have to be changed to keep pace with the changes in Canadian society that have taken place since the CPP was introduced. But beyond this consensus, there was none on what changes should be made. A majority believed that survivors rely on the benefit for income and that benefits should not be related to the employment/income of the survivor. The expert panel observed that if the flat rate component of SSP were to be eliminated, all payments would be related to earnings of the deceased contributor and the eligibility criteria that have been established would not be required. That would, in itself, not deal entirely with the eligibility issue, however, as the current law defines eligible beneficiaries as: "a spouse of a deceased contributor or a person of the opposite sex who lived in a marital relationship with the contributor before his or her death." Some experts saw the recognition of same sex partnerships to be inevitable. However, the broadening of the definition of spouse led to the observation that once eligibility is based on a "dependency relationship" it will be difficult to limit eligibility. As for eliminating the flat-rate component, there was no strong support, particularly as it is paid mainly in support of families. Experts also generally agreed that whether or not the surviving spouse is working, receipt of the survivor benefit is appropriate, since there is a need to support a period of adjustment to one income. Panelists differed, however, on the extent of the benefit, as they did on whether amounts of the benefit should vary according to the age of the surviving spouse. Some panelists suggested that the benefit should be uniform regardless of age, unless it was based on the presence of children; in which case it might be preferable to increase Orphan’s Benefits and decrease or make uniform the spousal benefit for pre-retirement spouses. The expert group was unanimous that any move to cease paying survivor benefits on remarriage of the surviving spouse would be retrograde. This provision was introduced into the CPP to protect women’s economic autonomy, and calls to cease benefits played into stereotypes that the number of remarried beneficiaries of the Surviving Spouse’s Pension is significant—a contention which the 1987 and 1996 Surveys of Survivor Beneficiaries show is not so. The latter survey found less than 10% were remarried. 2. Public Opinion on Eligibility Both current beneficiaries of the Surviving Spouse’s Pension and the general public were asked a series of questions on eligibility for SB. We first examine the five opinion items that were common to both surveys.20 In interpreting the results, we consider that a position lacks support if the largest proportion in the answer categories is less than 55%. Because we considered some of the questions to be difficult and perhaps not of interest to many respondents, we offered a "No Opinion" response category. This allowed respondents, who otherwise might have felt constrained to choose among reasons when in fact they had no view, to indicate that they had no opinion. In analysing the results we do not repercentage the responses omitting the "Don’t Knows." Rather, for a view to be considered to enjoy majority support, it must be held by at least 55% of the respondents, including those with no opinion. Survey respondents were clearly divided on three eligibility-related questions:
In general, the level of no opinion is higher in the survey of beneficiaries than it is in the general population survey. We conclude from Exhibit IV-18 that there is no clear support for eligibility changes on the three aspects noted from the perspective, at least, of general public opinion and that of current survivors. With respect to relating benefits to the duration of union, a complex gender/age result occurs: older males and females in the general population and male and female survivors think benefits should be linked to duration of the union, whereas both females and males in the 25-44 age group are more inclined to oppose the linkage with duration. Exhibits IV-19 and IV-20 compare the responses of male and female survivors with those of the 25-44 sample and the 45 and older sample of the general public, on two other eligibility issues—relating SB to the level of the survivor’s income and the impact of remarriage on eligibility. There is slightly more support for relating SB to survivors’ income in the general population than among the current beneficiaries. However, respondents in the general population surveys are quite divided. No age or gender differences appeared in the pattern of responses in the 25 to 44 sample, the 45 and older sample or the survivor’s survey. On the latter opinion question, a strong majority emerged in the general population surveys for a change in eligibility from the status quo. A significant majority of both the 25-44 sample and the 45 and older sample think that if a surviving spouses remarries he/she should not receive a survivor benefit. Again, there were no age or gender differences among the general population respondents. A gender difference arises among survivors, however. A majority of males favoured the status quo whereas female survivors were divided.
EXHIBIT IV-18 Views of Respondents on Three Eligibility Issues—Survivor Beneficiaries and General Population—By Gender and Age
a)Age of Surviving Spouse Some people say that younger surviving spouses should receive the same benefit as older survivors. Others prefer the current eligibility criteria. What is your opinion? (Circle One Answer.)
Source: Q.6 General Population Survey; Q.19 Survey of Beneficiaries. b)Duration of Union Under the current eligibility criteria for a CPP Surviving Spouse’s Pension, benefits are payable to the surviving spouse or to a common-law partner if he or she had lived with the deceased contributor for at last one year immediately prior to death. Some people say that the benefit should be related to the duration of marriage (or period of co-habitation) to the contributor. What is your opinion? (Circle One Answer Only.)
Source: Q.7 General Population Survey; Q.20 Survey of Beneficiaries. *Due to rounding. EXHIBIT IV-18 Views of Female Respondents on Three Eligibility Issues—Survivor Beneficiaries and General Population (cont’d)
Q.21Treatment of Separated Spouses Under the current eligibility criteria for a CPP Surviving Spouse’s Pension, separated spouses (formerly married to the contributor) are not eligible for a survivor benefit if there is an eligible common-law partner who had been living with the contributor for at least one year prior to the contributor’s death. Some have suggested that the common-law partner and the former spouse should both be eligible for a portion of the one survivor benefit. What is your opinion? (Circle One Answer Only.)
Source: Q.8 General Population Survey; Q.21 Survey of Beneficiaries.
EXHIBIT IV-19 Relating Eligibility to the Level of the Survivor’s Income—Comparison of Survey Results
Source: Q12, General Population Survey; Q. 23, Survey of Beneficiaries, CFO Panel. EXHIBIT IV-20 The Impact of Remarriage on Eligibility—Comparison of Survey Results
Source: Q10, General Population Survey; Q. 22, Survey of Beneficiaries, CFO Panel. A few questions on SB eligibility were included in the general population surveys that were not included in the survey of beneficiaries. These yield the following findings:
D. Summary Nearly 90% of current survivors are women, most of whom are neither married nor employed. The average age of female survivors starting benefits has risen considerably in the last decade. There has been a dramatic increase in female beneficiaries over age 75. The survivor benefit represents a significant proportion of household income for not more than one-fifth of the female survivors. Multiple lines of evidence confirm this conclusion. For low income women, those with $10,000 income or less, the benefit represents from 35% to 60% of total gross household income. Female survivors who perceive their current income to be less than adequate are disproportionately numerous among women of pre-retirement age and with little or no education. The general public tends to be both restrictive and generous with respect to eligibility for benefits. They think that survivors who remarry should not receive a benefit, but they would open up eligibility for younger pre-retirement survivors beyond the disabled and those with children. Experts and key informants think the current rules are, by and large, appropriate. Even on the more controversial pre-retirement eligibility rules, there is general acceptance of the income support aspects—largely because the principle of income support for families overrides their dislike for a departure from the income replacement principle.
A. Introduction In Chapter 3 we considered the rationale for the Surviving Spouse’s Pension and several other features of the CPP, with evidence from the literature review, international comparisons and the opinions of the expert panel and key informants. Here we add a perspective on adequacy or appropriateness of these features with the benefit of simulations of their effects and/or the opinion of survivors and the general public over 25 years of age on the two other components of survivor benefits—the Death Benefit and the Orphan’s Benefit. We then examine the evidence with respect to three other features of the CPP:
B. Adequacy of the Death Benefit In 1995, death benefit payments totalled $223.2 million. In January 1996, 8,239 death benefits were paid at an average of $2,521 per benefit. The maximum benefit in 1996 is $3,540. The death benefit is "a lump sum payment equal to six times the monthly retirement pension of the deceased contribution or roughly 10% of the YMPE, whichever is less."22 The death benefit is meant to act as a trigger for survivors to inform the government that CPP retirement beneficiaries are no longer living. The application for a survivor benefit also activates the application for the death benefit. This evaluation has not established whether all survivors who are entitled to a death benefit for a deceased spouse receive one. Determining the "take up rate" is complex because mortality data would have to be aggregated across provinces, and the number of eligible would need to be determined for the same period by identifying only those who were, in fact, contributors to the CPP. A pilot study in one province might be useful to establish whether there is a take-up issue which merits extending the study of the take-up of Death Benefits to the CPP coverage as a whole. The evaluation has examined perceptions of the adequacy of the Death Benefit. To obtain from current beneficiaries of a survivor pension perceptions of the adequacy of the death benefit received at the time of the death of a spouse, the survey of beneficiaries asked:
Exhibit V-1 shows that male and female survivors differ in their perception of the adequacy: 58% of males indicated that the Death Benefit (DB) was less than adequate. For female survivors, the perception that the DB was less than adequate is associated with age (post-retirement age survivors were more likely to find the DB adequate). Recent beneficiaries and those who had re-married were also more inclined than others to see the DB as inadequate. EXHIBIT V-1 Perceived Adequacy of Lump Sum Death Benefit, Male and Female Survivors
Source: Survey of Beneficiaries, CFO Panel. *Due to rounding. The "can’t recall adequacy" response was provided in the questionnaire in recognition of the fact that the way in which the Death Benefit is administered means that if the surviving spouse was not executor of the estate, he/she might not know the amount of the Death Benefit. Also, given the number of years since the death of a spouse for some of the respondents, difficulty in recalling the adequacy was likely to be significant, as indeed proved to be the case. Of the female survivors who could recall and who ventured an opinion, a majority found the DB adequate. These responses appear to be consistent with the view of key informants and experts that the DB makes a reasonable contribution following the death of a spouse and should not be dropped. C. The Orphan’s Benefit Opinions on the appropriateness of the Orphan’s Benefit (OB) were obtained in surveys from both current recipients23 and the general population over 25 years of age. Respondents proved to be divided on the "insensitivity" of the current name, but almost all those who think the name is insensitive want it changed. Since the vast majority of "orphans" will still have one parent alive and are not therefore truly orphans, "surviving child/children" might be more appropriate. In the beneficiaries sample, opinions on other aspects of OB were as follows:
With respect to the age limit for orphans in school and payment of the benefit to children even if they leave the care or custody of a parent or agency, the majority view of current beneficiaries strongly supports the status quo. On the age limit of the OB for those not in school, there is fairly strong support for extending the eligible years to at least 22 years of age. There is also substantial minority support (35%) for relating the OB to the income of those responsible for the support of orphans, although the majority position among those currently receiving OB was for the status quo. Within the general population 25 years of age and older, a majority also supported the status quo on the age limit for benefits and did not want to see changes relating to the Orphan’s Benefit either to the income of the parent/custodian or the contribution level of the deceased contributor. In the general population survey, a question was included that had not been part of the survey of beneficiaries. It explored whether the public supports the idea of paying a benefit only to the contributor’s dependent children—not the surviving spouse:
The result was overwhelming support for the status quo: the support for no change was 84% in the combined samples, with no differences evident in the responses by age or gender. D. DropOut Provisions The calculation of benefits under the Canada Pension Plan is affected by how much and how long people contribute. More contributions generally result in a greater benefit. The CPP provides that a contributor’s low earning months can be dropped out so that reduced earnings may be removed from the calculation of benefits, and, therefore, will not result in lower future pension benefits. In addition, other dropout provisions can be applied under the following conditions:
Taken together these provisions have the effect of increasing benefits, and therefore Plan expenditures and required contributions. Does the general public think they are appropriate provisions? Respondents to the general population surveys were asked whether current criteria on dropout should be continued or:
Exhibit V-2 shows that many respondents had no opinion on the subject—indeed, failure to express a view was higher than on any other issue. However, there was substantial support, especially in the 25-44 age group, among those who expressed a view, for extending the dropout provisions. EXHIBIT V-2 Extending the Dropout CPP Provision—Views of a Survey of General Public, by Age
Source: General Population Survey, Q.4. Respondents were not given an opportunity to say whether they thought the dropout provisions should be eliminated altogether, so that it is possible that some who expressed no opinion may not have found their "answer category" available. We did, however, simulate the effect of eliminating both the general and child rearing dropout provisions. The elimination of these dropout provisions would be expected to reduce expenditures substantially, but it would make achievement of the objective of replacing 25% of income up to the YMPE through the CPP even harder. Exhibit V-3 shows the simulation results from the CPP Actuarial Model and indicates the extent of the reduction in expenditures. There is little difference between the two expenditure curves for new benefits compared to new and existing, but all earnings-related benefits are affected by the change. Therefore, the retirement benefit accounts for most of the change, especially in the later years shown and for the change applying to both new and existing beneficiaries. As will be seen later, the effect of this change would be much greater than that of removing credit-splitting. EXHIBIT V-3 Percentage Reduction in Expenditures Resulting from Removal of the Dropout Provision Source: Simulation using CPP Actuarial Model. A further item in the survey of the general public gave respondents a chance to express a view on the appropriateness of increasing or reducing the dropout benefit—even if it did not ask directly about elimination.
The results in Exhibit V-4 show essentially no clear consensus for change. However, taking the "no change" with the "increase" position there is substantial support among those with a stated opinion for at least retaining the general dropout provision. In this respect, the public and the experts are in accord. In our simulations we examined the effect of a set of options extending the dropout provision to cover 20 or 25 per cent of the contributory period, rather than the current 15 per cent. Exhibit V-5 shows the results for the increase to 25. Under both simulations, the effect on total CPP expenditure is an increase, as claimants will be able to base their benefits more on higher-earning years. The amount of increase is approximately twice as great for the increase to 25 per cent as it is for the increase to 20 per cent. EXHIBIT V-4 Changing the Number of Years of Dropout Allowed—Views of a Survey of the General Public, by Age
Source: General Population Survey, Q.5. EXHIBIT V-5 Percentage Increase in Expenditures Resulting from Extending the Basic Dropout to 25 Per Cent
Source: Simulation using CPP Actuarial Model. E. Credit Splitting 1. Opinions on the Appropriateness Under the current CPP provisions, married people who subsequently separate or divorce, as well as those who terminate a common law relationship that has lasted at least one year, can split any CPP pension credits they have accumulated during the period they have lived together. Credit splitting results in a permanent amendment to each spouse’s record of earnings and may affect the level of current or future CPP retirement, survivor, disability and death benefits. Respondents to the general population survey were posed two questions 24shown below:
With respect to credit splitting on divorce, more respondents favour negotiation, but the level of no opinion is quite high. Female respondents were somewhat more disposed than men to support the current mandatory position (in most provinces); no significant age differences were evident in the level of support for one side or the other on this issue. With respect to credit splitting on separation, there was a strong preference, especially among both younger male and female respondents, for having credit splitting remain negotiable; older males were also supportive of the idea. 2. Simulating Removal of the Credit-Splitting Provision Since credit splitting was, at least in part, implemented to balance CPP benefits between spouses that separate or divorce, one might expect that the removal of credit-splitting to reduce expenditures on benefits. In fact, the introduction of credit splitting actually reduced the costs of CPP in the long term. Results of simulations using the CPP Actuarial Model show that the removal of the credit-splitting provision would result in a net increase in CPP expenditures. Exhibit V-6 shows the simulated percentage increase in expenditures year by year from 1997.
EXHIBIT V-6 Percentage Increase in Expenditures Resulting from Removal of Credit Splitting Source:Simulation using CPP Actuarial Model. The exhibit shows a relatively complex pattern of increases arising from the simulation. In fact, if the change applies to new beneficiaries only, the result would seem to be a slight reduction in expenditures over the medium term, with the increases not occurring until over twenty years hence. If applied to existing beneficiaries as well, however, this change would result to an initial increase in expenditures of 0.57% of total CPP expenditures, with this proportion declining to just under 0.4% within about 15 years. 3. An Explanation for the Finding Credit splitting should by and large be transferring credits from men to women and from those with a consistent earnings record to those with a more erratic earnings record. Both of these effects should increase costs. Women also live longer than men, so equal benefits transferred to women should be more costly. Secondly, some lower earnings obtained through the credit split should be eliminated by general dropout for those with a consistent earnings record. In fact, these effects are outweighed by the effective reduction in the cost resulting from the child rearing dropout. This is best illustrated by an example,25 illustrated graphically in Exhibit V-7.
EXHIBIT V-7 Illustration of the Effects of the Interaction of Child Rearing Dropout Provision and Credit Splitting
This example explains the simulation results: the wife does not need the credits transferred during child rearing years as these can be dropped out anyway. The husband has his earnings reduced by his wife’s years of low earnings during which she left the workforce to raise children, but he has no access to the child rearing dropout. The wife also has 14 years of earnings at 50% of the YMPE, which in this example is of no use to her, as they would reduce the average. If her average earnings outside the child rearing period were low, these credits could be useful. It should be noted that this is a worse case scenario, namely full attachment to the workforce outside the child rearing period, and no attachment during. Less extreme patterns would reduce the anomalies illustrated here. F. Summary Our assessment, in this chapter, of the achievement of objectives for features of the CPP other than the Surviving Spouse’s Benefit has relied mainly on evidence of the perceived appropriateness of the measures. By and large, experts and our key informants favoured retaining the death benefit, orphans benefits, the general and child rearing dropout provisions and credit splitting. We have not been able to assess whether every survivor entitled to the death benefit received the lump-sum payment (that averages around $2,500). We found that among females who could recall the adequacy, the majority found the benefit adequate. Males were less inclined to this view. Regarding the Orphan’s Benefit the public supported the status quo, but there was fairly strong support for extending OB to at least 22 years of age, even if the child was not in school. With respect to the dropout provision, the public was less inclined to comment on the achievement of objectives. They favoured extending it to cover other forms of family-related care-giving beyond the current rules, but there was no agreement on lengthening or shortening the number of years of general dropout permitted. The experts and the public differ on the mandatory aspect of credit splitting: the public is more disposed to making it or leaving it negotiable in the case of divorce and separation; the experts favour mandatory provisions. The simulation results showed that the removal of credit splitting would actually increase CPP costs, a consequence that would appear to be at odds with an objective of the program. (Credit splitting was intended to be cost neutral, or even to incur a small cost.)
A. The Effect of Survivor Benefits on the Employment Status of Survivors 1. Introduction It has already been established that Surviving Spouse’s Pensions go mainly to women and, in particular, to women of post-retirement age. As discussed earlier, we expect that the proportion of women who are employed at the time of the death of a spouse, or who have longer periods of contribution to the CPP, will be higher in future than is the case today. More women are employed today, so that the probability of their being employed at the time of their spouse’s death—particularly for surviving spouses of pre-retirement age—will be greater than in the past. This factor relates to the issue of the proportion of surviving spouses that can be expected to have CPP entitlement as a result of their own contribution record, in addition to the benefits they may obtain through survivor benefits as the result of a death of a spouse. We note that, despite the increased participation of women in the labour force, and therefore the increased numbers of contributors to CPP, women continue to earn less than men. They are to a significant degree engaged in non-standard and temporary jobs that imply lower retirement incomes. 26 The adequacy of CPP retirement benefits was addressed in the Phase 1 Evaluation. The focus here is on survivors. In the next section, we present data on the employment status of survivors before and after the death of a spouse. The data come from HRDC’s administrative data 27and the 1996 Survey of Beneficiaries of a CPP Surviving Spouse’s Pension conducted for this evaluation. 2. Administrative Data Earlier analysis has shown a clear decline in labour force activity among beneficiaries. The focus of this section, therefore, is on a comparison of labour force indicators before and after the start of CPP survivor benefits. This form of comparison suffers a weakness in that the analyst can never be certain that observed changes are solely the result of the benefit. Obviously the loss of a spouse could have a much greater effect on the lives of the survivors than the receipt of several hundred dollars a month. Impacts on health, decisions about where to live, and a number of social and psychological effects are very likely much more significant. On the other hand, no comparison group is available to represent what might happen to the survivors if no survivor benefits were available. Therefore, the analysis derives as much information as possible from the comparison of conditions before and after the death of the CPP contributor. The data and discussion that follow are derived from analysis of information in the form of annual data for five years before the year in which the survivor benefits started and for five years after that year. Limitations on data availability occur at both ends of the time period covered. In the early years, adequate historical data are unavailable. At the most recent end of the period, insufficient time has elapsed to allow for five full years of experience to have accumulated. The key indicators reported here are averages (including zero values) of the following:
EXHIBIT VI-2 Average Annual Earnings (1996 $), by Gender, for Starts in 1988
Source:HRDC Longitudinal Labour Force Data Base and CPP Master Benefits File. 3. The Effect of SB on the Employment Status of Survivors—Survey Results The Survey of Beneficiaries found that 30% of the female survivors were employed prior to the death of the spouse. Most of these women were of pre-retirement age. Exhibit VI-3 shows the proportion of each age group of female survivors who were employed in the year prior to the death of the spouse. The experience of pre-retirement survivors is very different from post-retirement survivors, as reflected in the much higher proportions of women who were then working. Also noteworthy is the fact that, for those who were working, the majority were in full-time jobs.
EXHIBIT VI-3 Employment Experience of Female Survivors in the Year Prior to the Death of Their Spouses
Source: Survey of Beneficiaries, CFO Panel questions 11a, 12, 13 *Indicates less than 5 cases. What do the survey results indicate about the reaction of women to labour market attachment after the death of a spouse? The broad answer to the question is provided by the following profile: for most there is no effect. When all female survivors who were employed before the death of their spouse (n = 225) were asked "What effect, if any, did the death of your spouse or common-law partner have on your employment?", the answers were as follows:
To show the impact of the death of the spouse more clearly, we have combined immediate post-death employment with "employed in 1995" to summarize the pre-post employment status more sharply. Exhibit VI-4 shows that the vast majority of the women who worked before the death of their spouse (disproportionately spouses who are currently of pre-retirement age) also worked at some time after the death of the spouse. EXHIBIT VI-4 Employment Status Before/After Becoming a Beneficiary of a Spouse’s Pension
Source: Survey of Beneficiaries, CFO Panel questions 11a, b, 14, 15. 23 cases are missing values in the table as a result of a Don’t Know/No Answer response to any one of the questions. The vast majority of women who were not working before the spouse’s death were not working afterwards (88%). This was true for all age groups. The age distribution of the female sub-sample that was not working before but was working after (n = 64) was distributed as follows:
These survey data are a useful counter perspective to the administrative data shown earlier. For those who are working there is a reduction in work following the death of a spouse, but most women were not working before their spouse’s death. B. The Effect of Other Sources of Income on Net Survivor Benefits SIMTAB (Simulation/Tabulation) was designed to simulate economic transfer systems and the impacts of policy alternatives. It takes into account the combined effects of other income support programs and taxes on both the size of benefits and on total family income. It thus permits an assessment of the contribution of SB to disposable income and it allows a calculation of the aggregate effects associated with survivor benefits. We regard it as the only reliable source of information to respond to the evaluation question: "What proportion of survivor benefits are recovered through the tax system or through lower costs of complementary programs, such as GIS, SPA, or provincial welfare?" The model was run with the CPP/QPP survivor benefits included as a source of income, then run again with this benefit deleted.28 The net effects on various other programs were then calculated, including GIS/SPA, the Child Tax Credit, GST Credit, federal and provincial taxes, and provincial tax credits, but excluding social assistance, worker’s compensation, and other sources of income. Exhibit VI-5 indicates what the federal and provincial governments pay for CPP survivor benefits and would pay for other programs in their absence. EXHIBIT VI- 5 Effect of Removal of Survivor Benefits on Government Revenues
*Includes GST credit, child tax credit, and provincial tax credits. N/A: not available. Overall, governments now issue about $3.2 billion per year in CPP/QPP survivor benefits, according to the model. In the absence of CPP/QPP survivor benefits, however, government expenditures on GIS/SPA benefits could be expected to increase, income taxes would be reduced, but CPP/QPP benefits would not have to be paid. The net result predicted by the SIMTAB model is that government expenditures would be reduced by over $2 billion. While the savings would be realized from CPP/QPP funds, the additional expenditures associated with the removal of CPP benefits would be financed from the Consolidated Revenue Fund. As noted, social assistance is not accounted for in the above analysis. The results of the MAPSIT examples suggest that there could be substantial liability for welfare benefits if CPP survivor benefits were no longer available. Inclusion of the effects of social assistance in the above analysis would greatly enhance its value and is strongly recommended as a direction for further research. C. An Unexpected Interaction Between Ancillary Benefits: The Effect of the Variable Contributory Period and Dropout Given the complexity of the CPP, and the fact that changes have been made to various program components, it is not surprising that interactions arise between ancillary benefits. At its inception, the CPP had a "fixed" contributory period, of age 18, or January 1, 1966, whichever was later, to age 65. Other than for those already aged 18 in 1966 (or immigrants coming to Canada after 1965), the contributory period was a fixed 47 years, giving a fixed dropout period of 7 years, and hence a "net" contributory period of 40 years. The introduction of the child rearing dropout as well as a flexible retirement age (60 to 70) has made the contributory period, and hence this general dropout period, variable rather than fixed. This has an impact on the supposedly cost-neutral nature of the early retirement provision. It could also distort the decision regarding early commencement of the CPP retirement benefit for those who exit the workforce prior to age 65. Exhibit VI-6 illustrates some of these effects. The table under Case 1 illustrates the impact of the variable contributory period on early retirement pensions. For example, in column 4 (Benefit Commences at 65; Scenario I), a contributor working for 32 years would have contributed for 80% of the total potential number of years to age 65, after the 15% general dropout (7.0 years). This would give rise to a monthly pension of $582 per month, starting at age 65. EXHIBIT VI-6 Examples of Impact of Dropout Provisions Data:
Note: "Percentage of maximum pension" is the percentage of early retirement pension at age 60 compared to an unreduced pension, i.e., $727, based on the given work pattern.
EXHIBIT VI-6 Examples of Impact of Dropout Provisions (cont’d)
Years out of workforce prior to age 60
Note: "Percentage of pension without child rearing" is the percentage of pension in the table compared to the pension payable had the contributor worked 10 years instead of leaving the workforce for child rearing purposes, i.e. row (5) or (6) in Case II ÷ (5) or (6) in Case I. 1Child rearing dropout. This table illustrates the same early retirement anomalies that were noted for Case I, shown here on row (7). Row (8) in the table above illustrates the phenomenon in relation to child rearing. Figures in this row are obtained by dividing the monthly pension in rows (5) or (6) by the corresponding figures in the same rows in the table under Case I. If the contributor were to commence to receive the pension at age 60, there should be a 30% reduction applied, to give an expected pension of $407 per month. However, because of the variable contributory period, the 32 years contribution represent 90% of the total potential years to age 60 (not age 65) after the 15% general dropout (6.3 years). Applying the 30% reduction gives a pension of $458, as shown in the table. If the variable contributory period were changed to a fixed contributory period, the three entries in the bottom line of the table should all be 70%, not 79%, 76% and 70%, respectively. These percentages are obtained by dividing the monthly pension in row(5) by the corresponding figure in row (6) for each of scenarios I, II and III respectively in Exhibit VI-1. The table under Case II illustrates the impact of the child rearing dropout. Clearly, the application of the child rearing dropout gives a higher pension than if there had been no such dropout. However, if the objective were to provide roughly the same pension as if the contributor had been in the workforce during the child rearing dropout years, this objective of full compensation is not being met in some cases. For example, in Case II, Column 4 (Benefit Commences at 65, Scenario I), without the child rearing dropout provision, the contributor would have received $400 per month. However, had the contributor actually worked the 10 years instead of We draw the following conclusions from analysis of these cases:
The first issue (early retirement anomaly) would appear to be an unintended side effect of the flexible contributory period. A fixed contributory period (for example, to age 65) would eliminate this problem. The second issue (child rearing dropout anomaly) is caused by dropping an equal number of years from both the numerator (the years of earnings) and the denominator (the contributory period). This does not give the same result as crediting a contributor with years of service while absent from the workforce for child rearing purposes. The latter approach would more accurately calculate the pension as if the contributor had not left the workforce, but would add to the cost. D. Economic Effects of CPP Contributions on Employers/ Employees
1. Introduction The CPP is funded on a cost-shared basis through the compulsory contributions of employers and employees, with self-employed persons paying the combined employer-employee rate. From 1966 to 1986 the employer contribution was stable at 1.8% of the Year’s Maximum Pensionable Earnings (YMPE), and rose to 2.8% in 1996. The CPP employer payroll tax currently accounts for roughly 20% of all employer payroll taxes in Canada. This proportion varies from province to province and even within provinces depending on the specific features of other payroll taxes such as the employer health tax, which can vary depending on firm size. Overall, employer payroll taxes have increased during the decade of the 1990's as a fraction of total employer payrolls. Currently employer payroll taxes account for approximately 13% of payroll costs. The CPP employer tax has increased at approximately the same rate as all other payroll taxes combined, maintaining its roughly constant share.
2. The Incidence of the Employer Payroll Tax The economic impacts of the CPP employer payroll tax depend on the extent to which the employer, as the initial payer of this tax, bears its full burden. The labour market literature refers to this issue as the incidence of the payroll tax. A recent U.S. 29study indicates that, through lower wages, workers bear more than one-half of the burden of mandated insurance financed through the payroll tax. This means that employer labour costs rise by less than half of the tax paid so that the negative impacts on employment and competitiveness are reduced. This finding is confirmed in recent Canadian studies.30 3. Payroll Tax Impacts We find that the labour market literature does not provide unambiguous conclusions about the overall economic impacts of the employer CPP payroll tax. A recent major OECD study identifies a number of factors that contribute to the gap between employer costs per unit of labour and the consumption financed by this employer payment. 31 The study refers to this gap as the tax wedge. Some of the components of the gap reflect employer payroll taxes such as the CPP. However, the entire wedge is made up of the sum of the social insurance contributions of employers and employees, personal income taxes, and consumption taxes such as the GST and provincial sales taxes. Taken together, all of these elements of what the OECD study refers to as the employment tax wedge have impacts on labour supply and labour demand decisions in labour markets. Analysis of comparative data for a wide range of OECD countries from 1978 to 1992, showed that although the overall average of the tax wedge has not changed significantly from 1978 to 1991/1992, there have been changes for a number of individual countries. In fact, as the last column in Exhibit VI-7 shows, the rate of increase over this entire interval has been larger for Canada than for any of the other countries included in the comparison. Payroll taxes that are borne by the employer have the potential to reduce employment, increase unit labour costs and reduce the competitive cost position of firms in countries where such taxes are important. What matters is the extent to which the tax burden is shifted to workers. If the employer CPP payroll tax is not shifted fully to workers, the portion borne by employers will reduce employment. However, the extent of the employment reduction by firms depends on the elasticity of the demand for labour.
EXHIBIT VI-7 Overall Marginal Tax Wedges 1in Selected OECD Countries, 1978-92(Percentage Tax Rate as Described in Footnote 1)
1The overall tax wedge includes employees’ and employers’ social security contributions, personal income taxes and consumption taxes. Social security contributions and income taxes are calculated by applying tax rules to the level of earnings of an Average Production Worker (APW), as calculated in OECD, The Tax and Benefit Position of Production Workers. Consumption tax rates are calculated from aggregate tax and national income data. Non-wage labour costs other than social security contributions are not included in the calculations. Social security contributions include only those paid to the public sector, contributions to the private sector are ignored despite their importance in some countries. Social security contributions in some countries are closely linked to expected benefits; therefore, treating them in the aggregate as if they were simply taxes is a simplification. Payroll taxes which are not earmarked for social security are not taken into account in these calculations. No account is taken of "non-standard" reliefs, such as those for mortgage payments (see the Tax and Benefit Position of Production Workers for a detailed discussion of limitations). Furthermore, the tax wedges in this table only concern one point on the earnings distribution (the APW case), and someone on this income level may have consumption patterns which lead to different consumption tax payments from those derived from aggregate data. Source: OECD, The OECD Jobs Study: Taxation, Employment and Unemployment, 1995.
There is also an employee impact from the component of the CPP tax paid by workers. Although it is difficult to know how the full package of CPP benefits is valued by workers because of the extent of worker diversity, it is clear that on average, workers in early cohorts of the CPP should value these benefits more highly than later cohorts. In fact, the calculated generational rates of return appear so high for earlier cohorts that it is likely that the benefits were valued at the contribution amounts or more. As a result, there should have been no employment impacts of the system in the early years of the Canada Pension Plan. As CPP premiums increase, and if there is uncertainty The impacts of the CPP in terms of both the worker and the firm tax components also have a generational dimension. Consider a typical worker at the time that the CPP was introduced who was 55 years of age and ten years away from anticipated retirement. If that worker retired and received CPP benefits for the average lifespan of a 55 year old person, then that person's payouts from the CPP would substantially exceed contributions. As the Phase I CPP evaluation points out, this is an inevitable result of introducing a pay-as-you-go pension system. This means that generational rates of return from CPP will differ quite substantially. The effect is exacerbated by the demographic changes of the post-war baby boom. How important are CPP employer payroll taxes in the overall competitiveness picture? There are no published overall assessments in quantitative terms of the economic impacts of the Canadian CPP employer payroll tax. This is because such an estimate requires data on a series of intervening variables, including the incidence of the tax, the employer response to the net tax in terms of labour demand elasticities and the labour supply response. In comparing tax initiatives that affect the labour market, a complete perspective requires information on payroll taxes, but also on other taxes affecting the labour market. Currently, even among employer payroll taxes, CPP does not dominate in terms of cost impacts, since the CPP employer payroll tax has accounted and still accounts for one-fifth of all employer payroll taxes. Moreover, it seems likely that approximately half of the employer tax burden ultimately falls on employees. However, the employer cost impact for CPP has increased in the last decade. Looking only at the employers’ CPP contribution, the Canadian CPP employer payroll tax is smaller than for our major trading partners (see Exhibit VI-8). On the other hand, the contribution of employers will rise significantly over the next six years. According to draft legislation to amend the Canada Pension Plan tabled on February 14, 1997, contribution rates will rise over the next six years to 9.9% of contributory earnings and then remain steady. These contributions are split equally between employer and employees, so employer contributions will be slightly below 5% of contributory earnings. The self-employed pay the full amount.
EXHIBIT VI-8 Employer Social Insurance Payroll Taxes—Selected Countries
*Varies with wage level. Note: Maximum insurable earnings levels also vary from country to country. Some countries have matching employer and employee taxes as in Canada and the United States, with the self-employed paying both shares. Some countries supplement employer and employee financing with contributions from general revenue. Source: U.S. Department of Health and Human Services, Social Security Administration, Social Security Programs Throughout The World. E. Comparison of Costs of CPP Survivor Benefit with Private Insurance
1. Introduction The purpose of this section is to compare private sector sources of survivor income benefits, to review differences in funding approach that may have an impact on the long and short term actuarial costs of alternative private sector provision. In addition, we estimate the differential administrative costs of provision of the benefits from the CPP as compared to private sector sources. 2. Sources of Survivor Income Benefits in the Private Sector Members of private pension plans will be entitled to pre- and post-retirement survivor benefits on their death. Coverage of the workforce is far from universal: although almost 100% of employees in the public sector are covered by private pension plans, only about 35% of employees in the private sector are so covered.32 Since pension reform in 1987, private plans have had to offer a 60% joint and survivor pension on retirement (this percentage varies slightly from province to province). In addition, the plans have to provide pre-retirement death benefits, generally in the form of a lump sum payment equal to the "commuted value"33of the member’s pension, in other words, the amount that would have been payable had the member terminated membership while still alive. In addition to pension plans, many employers offer group life benefits. Based on input from the Canadian Life and Health Insurance Association (CLHIA) and other surveys, it is estimated that over 90% of the workforce have employer-provided life benefits. In some cases, this benefit takes the form of a "survivor income benefit", i.e. a monthly payment to the survivor, but the vast majority of these plans provide a lump sum equal to one to three times the employee’s salary at the date of death. In many cases the basic benefit is paid for largely or entirely by the employer, with additional optional group life available, paid for by the employee. Generally speaking, such coverage would be for pre-retirement death, although some plans provide for residual amount after retirement, e.g. in the public service a lump sum $5,000 paid-up benefit from age 65. Generally speaking, evidence of good health would not be required in order to be covered by group life insurance. Depending on the size of the group, evidence of good health might be required in some circumstances, e.g.:
Finally, Canadians have access to a well-developed life insurance industry. Policies of various types are easily available to most Canadians, at market prices. Policies are by and large expressed as a lump sum insured, although income benefits are available. A declaration of good health is generally required to be accepted for insurance and above a certain threshold medical evidence of health will be sought by the company. 3. Funding of Sources of Survivor Benefit
4. Integration of CPP with Other Forms of Benefit There is little evidence that group life policies take into account either the (relatively small) death benefit or pre-retirement survivor benefits under the CPP. Similarly, evidence is not clear as to whether individuals take into account the expected value of CPP benefits when planning on the amount of individual insurance they buy in the insurance market. On balance it seems that any such reduction would be minimal. However, most private pension plans do integrate benefits with the CPP benefits. Generally speaking such integration is in relation to retirement benefits only. Thus, for example, a typical pension plan may provide a benefit on retirement of 2% final average salary times years of service, integrated with the CPP benefit from age 65, by offsetting the retirement benefit by 0.7% times average YMPE34 times years of service. The survivor benefit would be 60% of the retirement benefit, in other words, it would reduce by 60% of the CPP offset at the member’s age 65 (even if the member were dead at that point in time). This does not accord with the CPP survivor benefit, which is payable immediately on the contributor’s death, and changes from a pre-retirement to post-retirement benefit on the survivor’s age 65, irrespective of the age at which the contributor would have turned 65. These anomalies are illustrated in Exhibit VI-9. This table illustrates that in this example the member would have been entitled to $16,000 pa from retirement to age 65, at which age the member’s pension would have dropped to $11,100 on account of CPP integration. In total, the private plan pension and the CPP is shown to be level, i.e. the total does not change materially at the member’s age 65. The survivor benefit from the private plan is computed as a percentage of the member’s benefit, namely $9,600 pa when the member would have been below age 65 and $6,660 pa after this age (i.e. 60% of the pension that would have been payable to the member had the member still been alive). The CPP benefit on the other hand changes at the spouse’s age 65 from a pre-retirement pension survivor’s pension (37½% of the earnings related pension plus a flat rate) to a post-retirement survivor pension (60% of the member’s earnings related pension). In this example the survivor pension reduces from $3,360 pa to $2,940 pa. In other cases it could increase. Therefore, the total survivor pension from the private pension plan could change at both the member’s putative age 65 and the spouse’s age 65. A different total is payable when both are (or could have been if still alive) below age 65, one below and one above and both over age 65. These changes are not related in any way to the survivor’s income needs, but are consequent upon somewhat different benefit design philosophies in the two plans. EXHIBIT VI-9 Examples of Integration of CPP Spousal Survivor Benefits in Private Pension Plans
5. Comparison of Actuarial Costs For the reasons explained above, it is very difficult to compare the actuarial costs of various providers of survivor benefits, versus the CPP. It is likely that given its broader coverage, the CPP would experience a higher average mortality than other plans, giving rise to a slightly higher cost. However, this cost encompasses among other things the fact that some of those covered by the CPP would be uninsurable or insurable only with extra premium. For example, it covers contributors with a weak attachment to the workforce who would not be covered elsewhere. It is difficult to separate out the cost for survivor benefits in private pension plans. In any case, the design of such benefits is far from uniform. Notwithstanding a certain degree of uniformity brought about by pension legislation, a number of models exist, for example a CPP-like benefit under the federal public service plan, to a strictly cash approach in money purchase (defined contribution) plans. In addition, for post-retirement survivor benefits, some plans provide these as "add-on" benefits, at no extra cost to members with spouse at the date of retirement, while other plans actuarially reduce the pension to account for the joint and survivor feature. In these latter plans, the survivor benefit is at the plan member’s expense, not the plan as a whole. Finally, the pre-funding of these benefits makes a direct comparison difficult. Again, the variety of group life and individual insurance designs, and the variability of different groups (white collar, pink collar, blue collar etc.), make direct comparisons very difficult. The overall conclusion is that the "wholesale" cost of the benefit (i.e. the proportion of contributors dying each year) probably does not vary greatly as between the CPP and other sources of survivor benefits. It is therefore unlikely that significant savings could be realized by any form of "privatization" of the survivor benefit, due to lower actuarial costs in the private sector. 6. Administrative Costs Again, a direct comparison of administrative costs of the various programs is fraught with difficulties. In the case of CPP, for example, it would not be possible to separate out the marginal cost of providing survivor benefits. The best we can do is indicate the cost per dollar of total expenditure. The same would be true for private pension plans, with the added complication that the total administrative cost of private pension plans is very difficult to estimate, as much of the direct and indirect cost is borne by the plan sponsor and is not readily available, even as a cost per dollar of benefit. Also private plans vary greatly in size. For the larger ones, costs per dollar of benefit are probably comparable to CPP costs (1-2% of contributions), whereas the smaller ones have more significant costs per dollar of benefit. For group and individual life insurance we have estimated, based on internal Mercer sources, the overall administrative cost as a percentage of the total premium paid for life benefits. The estimated cost for large employer plans would range from 2% to 4% of total premium (excluding applicable premium taxes). This is slightly higher than the estimated 1.3% of total expenditures required to administer all aspects of the CPP program. However, the cost for smaller employer groups is much higher, at 10% of total premium or more, depending on the employer size. The cost to administer individual life is much higher again, with administrative costs ranging from 30% to 40% of the total premium. The conclusion is that the CPP cost per dollar of total expenditure is low in aggregate. There is no reason to believe that survivor benefits create costs out of line with the average. Private pension plans, at least the larger ones, are also low cost providers, as are group life plans. However, the individual life insurance industry tends to have relatively high administrative costs, compared to the other sources of survivor benefit discussed here. 7. Conclusions It is very difficult to compare CPP actuarial costs and administrative costs for survivor benefits with comparable benefits provided by private pension plans, group life plans and individual life insurance. It is likely that the actuarial cost of the CPP is slightly higher than that for the other plans, due to the higher mortality likely experienced by the broader group covered by the CPP. However, this cost could not be reduced by privatization—this would merely deprive those who might find it difficult to replace such coverage of a benefit. Insofar as administrative cost is concerned, the CPP has comparable costs to those found for large pension plans and group life plans, all of which have low costs, although it is difficult to be sure that all costs are being measured properly. CPP costs are significantly below those for individual insurance. We also looked at what evidence there was for integration of other sources with the CPP. Not much work has been done in this regard, but it would appear that there is little or no integration with group or individual life insurance sources. There is integration with private pension plans, but it tends to be somewhat indirect, as indicated above. F. Summary The impact of survivor benefits on the labour market behaviour of female survivors following the death of a spouse was limited. Most of the current beneficiaries were not in the labour force at the time of the death of a spouse and for most the situation did not change afterwards. Trend analysis shows that the number of weeks worked by women revealed a downward trend both before and after the start of benefits. This is reflected in annual average earnings. In future, it is not likely that significantly more widows will be employed at the time of the death of their spouse/partner. In part this is because the longevity of males is greater now than in the past and also because of a trend to earlier retirement. This means that it is likely that neither partner will be working at the time of the death of a spouse, and there is little reason to think that many will seek/find employment afterwards. Changes in women’s labour market participation have had an impact on how successful the child rearing dropout has been in compensating women for dropping out of the labour force for the care and nurturing of young children. Women who benefit the most are those who leave the workforce during child rearing years and have a strong attachment during years when they are not caring for young children. This was probably the model most prevalent in the 1970s, when this provision was first introduced. Women who leave the workforce for only a short period on the birth of a child and continue to have a strong attachment to the workforce during child rearing years benefit relatively little from this provision. In fact, they pay CPP contributions for periods of service that they would have been credited with anyway, because of the child rearing dropout. The model of the mother who returns quite soon to the workforce after the birth of a child is much more prevalent now than it was previously. The CPP employer payroll tax has been increasing in recent years, but it is relatively smaller than for our major trading partners. Currently, the CPP does not dominate the cost impacts on employers, since it accounts for only one-fifth of all employer payroll taxes, but it is scheduled to rise significantly in the next fifteen years. Other impacts noted were the following:
A. The Main Findings of the EvaluationThe main findings of the evaluation support the continuation of the Surviving Spouse’s Pension and the ancillary CPP benefits we have examined. The review has established that the rationale persists for the Surviving Spouse’s Pension, the Death Benefit, Orphan’s Benefit, the general dropout provisions, the child rearing dropout provisions, and credit splitting. With respect to survivor’s benefits, the evidence most strongly supported the rationale for post-retirement survivor benefits. Almost 90% of current beneficiaries are women and, of these, nearly three quarters are over the age of 65. Adequacy of the benefit was validated in a qualitative sense in relation to the original implicit replacement target. Analysis of current data on the significance of survivor benefits within total family income showed that for about a fifth of female survivors, the benefit represents a significant share (at least 20%) of total family income. Reliance on survivor benefits was found to have peaked in the late 1980s and to have declined somewhat since, although the proportion of family income replaced by SB and SB plus other CPP benefits has increased for those in the lowest income categories. The general public thinks that pre-retirement benefits should not be restricted only to those with children or who are disabled. Although not fully enthusiastic about the current eligibility for pre-retirement survivor benefits, most informed sources accepted the income support implications of the flat-rate feature of pre-retirement benefits. Current beneficiaries tended to support the status quo with respect to whether benefits should vary with the age of the surviving spouse and whether benefits should be related to the survivor’s income. They also support the current arrangements with respect to Orphan’s Benefits. Except for survivors, evidence supporting the retention of death benefits and orphan’s benefits was not so strongly presented; on the other hand no convincing arguments for dropping these features came from informed sources or the general public. The case for the general dropout provisions of the CPP was supported, even found to be enhanced in view of current labour market instability. Also, and despite the increased participation of working mothers, no convincing argument was made for ending the child rearing dropout provision of the CPP. Removal of the dropout provision would, of course, substantially reduce expenditure under the CPP. Evidence on the impacts of the program features we examined offered no major concerns for unintended consequences, although there were some:
Although we found that employed women reduce their average weeks of work after they begin to receive survivor benefits, there was no clear case to support an argument that employed beneficiaries differ from those with no employment income with respect to the adequacy of benefits. In other words, we found no justification for varying the benefit structure for employed survivors to take into account differences in income. Notwithstanding some unintended consequences, the three basic components of the CPP we have examined—SSP, DB and OB—should be retained. In the next section we examine possible enhancements or alternatives to the current system. B. Consideration of Program Alternatives1. Introduction Many suggestions for changes, both major and minor, were elicited from key informants including major stakeholders, the expert panel, as well as CPP program officers. Input on alternatives was also provided by the surveys of survivors and the general public. Ideas were also provided by reviewing the study of how Canada compares with selected other countries. This section begins by discussing the context which might create a climate for giving serious consideration to alternatives. It then goes on to discuss changes that have been suggested for each of the components. 2. Context for Change A number of issues relating to the changing environment in regard to survivor benefits and other ancillary benefits were discussed in Chapter 3. These included such items as:
The CPP has already changed in response to some of these issues.
Additional contemporary contextual issues include:
To address these contextual issues we look at a number of reforms that did not proceed, as well as some current suggestions for change. 3. Post-Retirement Spousal Survivor Benefits Few alternatives were offered to the current spousal benefit. The only suggestion put forward involves mandatory credit splitting for all couples, with survivor benefits payable on the death of one partner based on the net credits of that partner. One illustration of how this could work is shown below:
Notes: The table ignores the combined benefit rules, to illustrate the point. The table shows the benefit to the survivor on the death of the other partner. In Scenario I, the higher earning spouse (HES—usually the husband), receives a 100% pension, while the LES (usually the wife) has no CPP entitlement. The total pension paid to the couple while both are alive is 100% of the YMPE. On the LES’s death there is no change in the retirement pension, while on the HES’s death, the Surviving Spouse’s Pension would drop to 60% of the YMPE. Scenario II illustrates a working couple where the total income is 150% of YMPE while both are alive, dropping to 130% on the LES’s death, but 110% on the HES’s death (ignoring the combined benefit rules). The proposal is to split credits equally between spouses, so that in Scenario I both HES and LES would be entitled to 50% each of the YMPE, for the same total 100% as before, when both are alive. On either death, the benefit payable to the survivor is the 50% of the retirement pension continuing in the survivor’s own name plus 60% of the deceased partner’s retirement pension, for a total of 80% (50% + .6 x 50%). Similarly in Scenario II the pension would be split 75%/75% (i.e., 100% plus 50% divided by 2). On either death, the survivor would receive 75%, plus 60% of 75% or 120% of the YMPE. In Scenario III, where both parties have equal entitlements, both the current and proposed approaches give the same result, again ignoring the combined benefit rules. The following points can be made about this suggestion:
In regard to the combined benefit rules, these have been ignored to illustrate how this proposal might work. If the combined benefit rules were to be applied, all the percentages in the table above would have been capped at 100% and hence there would have been no difference in Scenario II or III between current and proposed post-retirement survivor benefits. (These simplified examples assume cohabitation throughout each contributor’s career. The results would be different if we were to model periods of cohabitation and non-cohabitation.) The fact that there are more and more working couples means that a growing proportion of survivors will be affected by the combined benefit rules. Effectively, the result of the survivor benefit in these cases is to top up the existing retirement pension to 100% of the YMPE, irrespective of what the survivor benefit mechanism is (i.e., current mechanism or some variant of the credit splitting mechanism discussed above). These issues appear to be of concern in the U.S. (called "dual entitlement limitation") as well,35 although no specific recommendations to remedy them have been put forward. 4. Pre-Retirement Spousal Survivors Benefit As has been pointed out before, relatively few dollars go into pre-retirement benefits as compared to post-retirement benefits. Improved mortality will further reduce the financial significance of this component. Nonetheless, this has been the component that has generated the most debate, and the greatest number of alternative proposals. Alternatives to be discussed are:
5. Orphan's Benefits As mentioned previously, orphan’s benefits elicited less interest, including fewer suggestions for changes, than spousal survivor benefits. Some discussion of alternatives to the current benefit has already appeared above, in relation to changing the current structure to a "family benefit". Nonetheless, some alternatives are discussed below. These are:
6. Death Benefits As with the orphan’s benefit, there was limited discussion of this benefit. The alternatives appear to be:
The considerations with respect to these options are as follows:
7. Credit Splitting and Assignment The current arrangements for credit splitting on marriage breakdown were strongly supported by both key informants and the expert panel. Three areas could be looked at:
8. General Dropout
Initially, after the transitional phase-in period, the general dropout was designed to be fixed, namely 15% of the years between age 18 and age 65. This arrangement gave a 7 year dropout, for a net contributory period of 40 years. This analysis ignores the dropout for months while on disability pension. Changes to the CPP, namely the child rearing dropout, credit splitting and flexible retirement ages have complicated the originally simple design. Some of these complexities and interactions have been discussed in Chapter 5.
Alternatives discussed below are:
9. Child Rearing Dropout The concept of child rearing dropout was strongly supported by most respondents, although a few felt that it was a "social measure" that had no place in a contributory pension plan.
Alternatives discussed are:
C. Summary: A Weighing of the Alternatives
1. Changes Worth Pursuing
2. Changes Not Worth Pursuing
The alternatives identified as worth pursuing create benefits from "modernizing" provisions of the CPP to respond to the changed context. They appear not to create countervailing negative effects, especially increased cost and administrative complexity.
As part of the review of CPP survivor and other ancillary benefits, we have conducted an international comparison of six countries – Argentina, Australia, France, Sweden, United Kingdom and the United States. The Technical Report "CPP Evaluation Study—International Comparisons" reviews programs in these countries for any assistance they can lend in regard to assessing CPP provisions under study. The countries have been chosen to provide a cross-section of social, economic, political and demographic conditions. These countries form a broad spectrum of generosity in terms of social benefits. The report points out some of the contextual differences that make comparisons difficult to analyze. The report summarizes contribution levels, retirement benefits, survivor pensions, lump sum death benefits, division of benefits on marriage and drop-out provisions (or equivalent provisions for these countries). Please see the working report for more details. Most countries provide surviving spouses’ and children’s benefits similar to Canada, although Australia and the UK confine these spousal benefits to widows. Age and family status dependent pre-retirement spousal benefits seem to be unusual—only the UK had a similar pattern. Lump sum death benefits are also present, but there is more variation. Some countries do not provide such benefits, while others provide much more generous benefits than Canada. Insofar as other benefits are concerned, some countries (Sweden and the UK) provide some type of drop-out benefits for child care and other family-related duties. It is more difficult to compare general drop-out rules. Most countries seem to favour a fixed contributory period instead of the Canadian approach. Credit splitting is unique to Canada among the countries studied. Argentina and France pro-rate survivor benefits for multiple marriages instead. In terms of demographic profile, Canada is a relatively "young" country, similar to Argentina and Australia. Other countries have older population profiles, similar to the expected Canadian profile by the end of the first quarter of the next century. Most countries show similar proportions of women in the workforce (70% of prime working age women). The report notes a number of proposed changes. For example, some countries are moving toward a defined contribution system. In other countries, benefit reductions are scheduled to occur in the next century. Finally, Australia is considering credit splitting on marriage breakdown and the UK system, while not splitting state pensions, does consider them as family assets in cases of marriage breakdown. Similarities and differences with Canada in contributions and benefits, described in more detail in the tables that follow, are summarized below:
Surviving spouse's benefits
Surviving children’s benefits
Similar to Canada’s although more variation than spouse’s benefits.
Death Benefit
Wide variety of provisions, including no provision.
Division on marriage breakdown
Unique to Canada.
Drop-out provision
Limited similarity in some countries.
Tax deductibility of contributions
International comparisons establish full tax deductibility of employer contributions. Treatments of employee contributions (tax credit) is not found anywhere else.
We were commissioned by Human Resources Development Canada (HRDC) to conduct a survey of beneficiaries of CPP Surviving Spouse’s Pensions. The purpose was threefold:
The respondents to this survey are derived from an attempted census of all panelists in Canadian Facts’ Canadian Family Opinion (CFO) Panel who indicated in a survey sent to all 22,000 panelists in January 1996 that they were in receipt of a CPP Surviving Spouse’s Pension. The CFO Panel is a continuously-maintained panel broadly representative of Canadians across the country who participate, from time to time, in self-complete questionnaire surveys on a variety of subjects.
A total of 1,153 questionnaires were mailed. As a result of the process of qualification of respondents as being within the target population (see Questions 1a and b), 116 panelists indicated that they did not qualify. Of the remainder (whom we will call qualifiers), 38 provided information that was so incomplete that it was not useable; 158 did not respond; and 841 responded with complete or virtually complete questionnaires. The response rate from qualifiers was, therefore, 81%. Canadian Facts offered an incentive (a "final lucky draw" for prizes) for those who qualified to complete the questionnaire and who completed it.
Careful edit checks were made on the household income data to ensure that data reported on survivor benefits (or, in some cases, SB plus Orphan’s Benefits) did not exceed stated program maxima. The data were then weighted to adjust by age and sex within region to the known population of Beneficiaries of the Surviving Spouse’s Pension as recorded in CPP Administrative data reports, by age and sex within province, March 1996.
The main effect of the weighting was (a) to increase the number of males (to 93 from an unweighted n of 41) and to reduce the number of females from 800 to 748 (weighted) and (b) to decrease the number of younger female recipients and increase the number of those over 74 years of age.
Exhibit 1 compares the demographic characteristics of respondents to Statistics Canada’s 1987 survey with the results of the CFO panel survey.
The 1996 survey respondents have more education and some of this difference will be due to cohort effects. However, we have concluded the CFO data over-represent survivors with higher education levels. EXHIBIT 1 Profiles of CPP Survivors—1987 and 1996 Surveys—Totals
As part of the Evaluation of Survivor’s Benefits and Ancillary Program Features, we were asked to conduct surveys of the general public’s views on a number of program-related features. As was the case for the survey of beneficiaries of the Surviving Spouse’s Pension, we surveyed members of Canadian Facts’ Canadian Family Opinion Panel. 39 Two samples were drawn randomly from CFO panelists in the target age populations, excluding persons who responded to the Survey of Beneficiaries of a Surviving Spouse’s Pension. We drew a sample of 1,500 names for the 45 and older sample and 1,000 households for the 25 to 44 age group. Respondents within households were selected randomly.40 The response rate was better for the older age group sample:
Since the same questionnaire items were administered to both samples, we designed the study so that the two samples could be analyzed separately or, after proper weighting to the age/gender proportions of the Canadian population outside Quebec, they could be analyzed at the level of the total population 25 years of age and older. The combined result would provide a representative profile of the views of Canadians outside Québec aged 25 years and older.41
We have made an effort to compare the achieved sample with 1991 Census data or other Statistics Canada data for more recent years, but with mixed results. Too often the categories reported do not precisely match those we have used and since we must use data only for those 25 years or older, the comparisons are limited. Age distributions match reasonably well, although after weighting by gender and age, our total sample somewhat under-represents the 65+ population (compared to the 1991 Census) and somewhat over-represents the under 35s (16.8% compared to 13.5% for the 65+ and 36.6% compared to 27.7% for the <35). The gender match is good, as is the case with residence ownership status (owning with and without mortgage). Where household income groups can be compared—under $20,000 and $20,000-$29,999—the match is also good.
Exhibit 1 shows several selected features of respondents of the two samples The results selected here show that:
EXHIBIT 1 Selected Demographic Characteristics of the Two Samples—Survey of the General Public
*Less than 0.5%.
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
![[Table of Contents]](/web/20060126234402im_/http://www11.hrdc-drhc.gc.ca/edd-img/content.gif)
= 100% of maximum pension
= 92.7% of maximum pension1
= 82.5% of maximum pension
= 96.3% of maximum pension1
![[To Top]](/web/20060126234402im_/http://www11.hrdc-drhc.gc.ca/edd-img/top.gif){kind=small}