Evaluation of the Canada Pension Plan (Disability Component) - Final Report - September 1996

Previous

Contents

Next

3.0 Program Success: Objectives Achievement and Impacts, Effects

This section of the evaluation deals with questions of impacts and objectives achievement, such as: "What was achieved?"; "Was this program component successful?"; "What has happened as a result of CPPD benefits?"; "Were there any unintended impacts and effects?"

The relevant findings below discuss key features of CPPD, including eligibility criteria, caseloads, characteristics and activities of beneficiaries, adjudication, appeals, treatment of abuse and reassessments, rehabilitation and work, benefit levels, and linkages to other programs. The detailed evaluation questions are addressed throughout the findings in sub-sections of Section 3. Of particular interest were the causes of the recent increase (1991-94) in the utilization of the CPPD.

3.1 Profile of Beneficiaries

This section addresses the evaluation question: What are the claim patterns, profiles and activities of these claimants?

As can be seen in Exhibit 2, the number of CPPD beneficiaries has risen steadily since the first benefits were paid in 1970, with a greater percentage of females being eligible for benefits and being granted them by the 1990s. The number of all CPPD beneficiaries increased by 231% between 1981 and 1995, with female beneficiaries rising by 358% and males by 176%.

Exhibit 2^* Numbers of Beneficiaries (end of June of each year)

Year	Male	Female	Total
1981	62,995	27,185	90,180
1985	93,967	38,735	132,702
1990	124,667	57,473	182,140
1995	173,971	124,841	298,612

* See: Statistics Related to Income Security Programs (unpublished), Income Security Programs Branch, HRDC, March 1996. These are the numbers of CPPD disabled beneficiaries and exclude the benefits to their children, which are about one-third of the above. QPPD beneficiaries rose from 23,271 in 1981, to 37,653 in 1985, to 45,256 in 1995 (same source).

CPPD beneficiaries generally have received earnings replacement from the federal Unemployment Insurance program prior to being granted CPPD benefits. This is demonstrated in Exhibit 3 by the average number of weeks of employment and unemployment in the year before CPPD benefits were claimed, and further illustrated by the average number of weeks of UI benefits and UI sickness benefits in the year before CPPD benefits were claimed.

Exhibit 3 CPPD Beneficiaries' Average Weeks of Employment, Unemployment, UI Benefits and UI Sickness in the Year Prior to Receiving CPPD⁴⁸

	1984	1989	1993
Weeks of Employment	32	31	34
Weeks of Unemployment	11	6	7
Weeks of UI Benefits	9	4	6
Weeks of UI Sickness Benefits	1.3	1.3	1.6

A profile of beneficiaries was further indicated by data collected through a survey of a sample of over 3,000 CPPD beneficiaries which was conducted by Statistics Canada in 1995. This profile is, therefore, based on the reports of those persons with disabilities who receive earnings replacement through CPPD benefits, with estimates subject to the usual limitations of sample surveys.⁴⁹

Sources of Information for CPPD: Respondents to the 1995 beneficiaries survey reported that they first heard about CPPD benefits from many sources: mainly from doctors or nurses (32%) and employers (22%), and, in some cases, from Federal or Provincial government departments (15%) or family and friends (11%).

About one-third of respondents reported that they requested information from Health and Welfare Canada or Human Resources Development Canada when they first applied for CPPD, and 97% said that they were able to get the information needed.

Age: The age of beneficiaries responding to the survey ranges from 20 to 64 years with an average age of 54 years (Exhibit 4, below). Older respondents (60 to 64 year-olds) comprise the largest single group of those receiving benefits (34%) and the smallest group (4%) of respondents are aged 20 to 34 years. The 55 to 59 year-olds form the next largest group, after 60 to 64 year-olds.

Exhibit 4
Distribution of CPPD Beneficiaries by Age
(based on 3622 respondents, 1995 Beneficiaries Survey)

Sex, Marital Status: The sex of CPPD beneficiaries responding to the survey somewhat reflects the Canadian working population with 58% males and 42 % females currently receiving benefits. Most were married or living with a partner (71%), with only 19% being separated, divorced or widowed, and the remaining 10%, single or never married.

Living Situation: Of all CPPD beneficiaries responding to the survey, 81% were living with family member(s) in their household, with 18% living by themselves. Very few beneficiaries (2%) were so sick or disabled that they had to live in a nursing home, chronic care hospital or other health care institution. Among those who were living with other family members in their household, only 19% had dependent child(ren), that is, child(ren) under the age of 18. The size of beneficiaries' households varied, with up to as many as 8 members.

Beneficiaries' Education: 28% of respondents reported having no education, or only elementary education, while 30% of respondents had some secondary education. Another 20% had completed their secondary education. The proportion of respondents having completed some post-secondary education, including a diploma or university degree, was 22%. Thus CPPD beneficiary education levels were generally low compared to the general Canadian population.

Diagnosis/Illness: There was a wide range of diagnoses reported by survey respondents who received CPPD benefits (Exhibit 5, below). Back/joint problems, heart/stroke/high blood pressure, and psychiatric illness were the three major health problems for which CPP disability benefits were most frequently granted, according to survey respondents.

For example, 54% of CPPD survey respondents reported that they received CPPD benefits due to their back/joint problems; 30% due to heart/stroke/high blood pressure; and 20% due to psychiatric illness. There was also a considerable number of people who received CPPD benefits because of other health conditions such as spinal cord injuries (12%); nervous system (11%); and allergies (11%). Smaller proportions (less than 10%) of other health conditions existed among respondents (note that reports of causes of disability could be multiple, and so add to more than 100%, and were not always necessarily the basis on which benefits were actually granted).

Exhibit 5 Beneficiaries' Diagnosis/Illness^* (based on 3622 respondents to the 1995 CPPD Beneficiaries Survey)

		%
(a)	Back/joint problems (e.g. Arthritis/rheumatism)	54
(b)	Heart/stroke/high blood pressure	30
(c)	Psychiatric illness/depression	20
(d)	Spinal cord injuries (e.g. Paralysis)	12
(e)	Nervous system (e.g. Multiple sclerosis)	11
(f)	Allergies (e.g. Asthma, environmental hypersensitivity)	11
(g)	Diabetes	8
(h)	Deaess/Blindness	7
(i)	Lung disease	6
(j)	Cancer	5
(k)	Infections/immunity disorders (e.g. AIDS, tuberculosis)	4
(l)	Substance abuse	**
(m)	Other	25

Program statistics demonstrate that the percentage of benefits granted for specific diagnoses varied significantly over the 1980-92 period. More specifically, a significantly higher proportion of disability pensions was granted on the basis of diseases of the musculoskeletal system (increased from 22% to 32%) and mental disorders (increased from 8% to 12%) in 1992, as compared to 1980. The relative share of other morbidity causes declined significantly over the same period, such as diseases of the circulatory system (from 30% to 16%).⁵⁰

Disability and Activity Limitation: The CPPD Beneficiary Survey was not designed to develop a specific index of CPPD severity. However, an attempt was made to collect some data from respondents on their limitations in daily activities as a proxy for disability "severity". The scope of disability was measured based on the respondents' reports on their daily activity limitations using some of the objective questions which Statistics Canada had previously developed in the HALS.⁵¹ These limitations do not match the CPPD definition of disability, which has key components related to time (the disability is "prolonged") and employability (the individual is deemed incapable of regularly pursuing substantially gainful employment).

The scope of disability among CPPD recipients was suggested by survey respondents' self-reporting on their activity limitations. As the exhibit below indicates, CPPD beneficiaries face many activity limitations. The most common problems for CPPD beneficiaries regarding activity limitations were their body movement problems. Over two-thirds of beneficiaries reported difficulty walking 400 yards (about three city blocks) without resting (74% of reporting cases), walking up and down a flight of stairs (69%), and/or bending down and picking up an object from the floor (67%). Among those who reported body movement limitations, a sizable proportion of them indicated total inability to undertake such activities.

Exhibit 6 Activity Limitations of CPPD Beneficiaries

Activities	Incidence of* Limitations	Totally Unable to Do So
Walking 400 yards/400 metres without resting	74%	31%
Walking up and down a flight of stairs	69	14
Bending down/picking up an object from the floor	67	18
Seeing ordinary newsprint with glasses	20	4
Hearing what is said in a group conversation	19	4
Speaking or being understood	17	2

Problems such as seeing, hearing and speaking seemed not to be a great concern for most CPPD survey respondents. Only about 20% or fewer of respondents reported difficulty seeing ordinary newsprint with glasses (20% of reporting cases), hearing what was said in a group conversation (19%), and/or speaking or being understood (17%). Very few recipients (4% or fewer) indicated that they were totally unable to see, hear and/or speak.

Among the survey respondents in this study, a majority of them was likely to attribute their activity limitations to their physical health condition versus their psychological or mental condition. Some 88% of respondents reported themselves to be limited by their physical conditions in the kind and the amount of activities they could do at home and/or in some other activities such as travel, sports or leisure. On the other hand, about 40% of CPPD beneficiaries also attributed such limitations to their psychological or emotional condition. (Thus a substantial number reported both physical and psychological/emotional limitations). These data suggest that the CPPD beneficiaries population faces a variety of disabilities.

Discussion: The survey of beneficiaries points towards a population that is severely disabled. In this respect, the results suggest that the population receiving benefits is appropriate for receipt of CPPD benefits. Yet, like other results to be noted below, these results suggest that some CPPD beneficiaries (7%) may, in fact, have relatively modest levels of disability. This points towards the importance of efforts to develop methods, both to monitor adjudication, and to emphasize ongoing reassessment of the CPPD population, as discussed below.

3.2 Income from CPPD and Related Programs

Some questions posed in this section include: What proportion of earnings replacement is supplied by CPPD benefits? How do these benefits compare with those of Canada's key trading partners?

3.2.1 The CPPD Benefit

Basic CPPD Benefits: Benefits received by individuals from CPPD can vary substantially, depending upon the individual's contribution history. A minimum CPPD benefit is set by the flat rate portion of the CPPD pension, which was $319.85 per month in 1995, and which is received by all CPPD beneficiaries. The balance of the pension, however, is contribution-based and corresponds to 75% of the retirement pension that disabled claimants would have been entitled to receive, calculated as if they had reached retirement age.

In turn, the retirement pension is computed on the basis of average adjusted contributions paid by individuals during their entire contributory period. The contributory period usually starts at age 18 and ceases when a retirement pension is paid, or when contributors reach age 65 (but it can be extended up to age 70), or at the onset of disability. Provisions exist for dropping-out years of no or lower earnings, years spent caring for a child under the age of seven, and years in receipt of a CPP disability pension.

Thus, the maximum earnings-related component of a disability pension - set as 75% of the maximum retirement pension - was $534.89 in 1995. When combined with the flat-rate portion of the benefit, the CPPD monthly pension could reach a maximum amount of $854.74. Additional benefits are provided for dependent children, in the amount of $165.00 per child per month in 1995.

3.2.2 CPPD as a Share of Income

Overall, CPPD benefits provided a significant portion of annual income of respondents to a recent CPPD beneficiaries survey, regardless of age (Exhibit 7). This CPPD portion was even greater for those beneficiaries who did not receive any WCB, auto insurance or private LTDI benefits, as shown in Exhibit 8.⁵²

3.2.3 Composition of Gross Income of 1995 CPPD Survey Respondents

Sources of Income : Stakeholders who were interviewed as part of the evaluation noted that beneficiaries had coverage from more than one program, so that in their view, many CPPD beneficiaries derived income from more than one public source. This is inherent in the current system.

Exhibit 7 Percentage Shares of 1994 Income by Source^* of CPPD Beneficiaries, who were Survey Respondents, by Age Group, 1994

	Age (%)				Total
	16-49	50-59	60-64
CPPD	55	51	49	52
Provincial Social Assistance	5	2	2	3
WCB, Auto Insurance, Private Disability Insurance	27	29	22	25
UI, Employment Income, Retirement Income**	13	18	27	20
Total	100	100	100	100

Exhibit 8 Percentage Shares of 1994 Income by Source of CPPD Beneficiaries who were Survey Respondents by Age Group Who Do Not Receive any WCB, Auto Insurance and Private Disability Insurance Benefits, 1994*

	Age (%)
	16-49	50-59	60-64	Total
CPPD	72	69	60	66
Provincial Social Assistance	11	5	3	6
UI, Employment Income, Retirement Income	17	26	37	28
Total	100	100	100	100

For example, individuals with severe disabilities caused by work-related injuries or diseases are usually entitled to benefits from both provincial WCBs and CPPD. Low-income workers who become disabled and who are not covered under a private LTDI plan often have to rely on provincial social assistance (PSA) to supplement their CPPD benefits. The same kind of 'double dependence' can also apply to public automobile accident insurance and, in some cases, UI and sickness benefits. Of course to the extent that these programs have different objectives it is entirely appropriate that individuals be covered by more than one program.

This multiple sources hypothesis is also suggested by the number of income sources reported in the 1995 survey of CPPD beneficiaries (see Exhibit 9, below). These data suggest that many CPPD recipients who were respondents to this survey - more than half - receive benefits from other sources, most importantly, from LTDI companies and WCBs.⁵³ Private sector LTDI providers indicated that, while many CPPD beneficiaries may receive payments from LTDI policies in addition to CPPD benefits, CPPD benefits are taken into account when LTDI benefit amounts are calculated, so that LTDI benefits are reduced by 100% of the amount of CPPD benefits.

Exhibit 9
CPPD Survey Respondents' Income Sources in 1994
Percentages Reporting Various Sources of Income
(based on 3,266 respondents to the 1995 Beneficiaries Survey)

(Source: 1995 CPP Disability Beneficiaries Survey)

The same survey, as noted earlier, indicated that overall, 52% of the CPPD survey respondents' income support was accounted for by CPPD. This statistic does not, by itself, reveal the fact that income shares were very different for different sub-groups of survey respondents: CPPD was a smaller share of income for those with WCB or LTDI incomes, but for many respondents, especially those with lower incomes, CPPD accounted for most of their income. (See Exhibits 7 and 8 on previous page.)

Where information-sharing agreements between CPPD and LTDI benefits providers are not comprehensive due to confidentiality concerns, LTDI providers have to rely on beneficiaries to self-report any such duplication or overlap in benefits. In other cases CPPD-LTDI agreements promote information-sharing and direct payments by CPPD to the LTDIs. Information-sharing agreements between CPPD and a number of WCBs are being negotiated, but these agreements are by no means standard across the country. Some agreements have been negotiated with regards to the reimbursement of overpayments, which occur when other disability benefits providers pay benefits to claimants who are awaiting an adjudication decision from CPPD.

This finding suggests the need for improved information-sharing between CPPD and other benefit providers as a means of reducing the providers' costs. However, such an initiative might also involve greater administrative costs for CPPD.

Discussion: While payments from multiple sources is an inevitable outcome of an income security system composed of a variety of programs with different objectives and rationales, many provincial stakeholders suggested that it is important to develop greater co-ordination and integration of PDI programs to avoid the problems which ensue from such duplications and overlaps.

Concern with multiple sources of earnings replacement also surfaced in the literature review conducted for this evaluation. Several authors examined in the literature review asserted that lack of coordination among programs produces a situation that is not only inefficient, but inequitable. They argue that "the lack of a unified system means that there is a considerable disparity in the treatment of people with similar disabilities". Different programs, both public and private, have not only different benefits but also differing eligibility criteria and, in the case of public programs, varying jurisdictions. This overlapping network of income and service provision may create considerable complications when programs interface or overlap for the same beneficiaries.

3.2.4 Earnings Replacement Rates and Program Offsets

Some of the evaluation questions examined in this section include: What proportion of earnings replacement is supplied by CPP Disability benefits? What are the impacts of the Year's Basic Exemption (YBE) and the Year's Maximum Pensionable Earnings (YMPE) on disability payment levels? Is there any duplication of benefits for public (federal, provincial) and private sources, i.e., CPP Disability, provincial Workers' Compensation and private long-term disability benefits? Within the Canadian income security system, what kind of offsetting practices have been adopted, especially in light of the interaction of CPPD/QPP Disability benefits with other government benefit programs, other earnings, or private insurance benefits?

Basic Offset practices: The evaluation examined the question: Within the Canadian income security system, what kind of offsetting practices have been adopted, especially in light of the interaction of CPPD/QPPD benefits with other government benefit programs, other earnings, or private insurance benefits?

These practices were found to be somewhat varied according to stakeholder interviews, depending upon the income security program involved, but basically took the following form:

• Some WCBs were found to reduce benefits they paid by 100% of the payments provided by CPPD. In those provinces, a person would receive a total amount no higher than the pre-offset WCB pension (usually 70% of gross or 90% of net pre-disability earnings), which would almost always be higher than CPPD. However some variations were noted.
• In particular, at the time of the research (Summer-Fall, 1995), in some Provinces/Territories (e.g. Alberta) beneficiaries could collect both a full CPPD pension and a full pension. The rationale for this was that the WCB saw itself as fully responsible for compensating a worker with a disability. At the same time CPP provides benefits to all who meet eligibility requirements, but does not take into account benefits from other programs. This resulted in the possibility that, in those jurisdictions, some persons might receive a post-disability income that (combining both pensions) would be higher than the pre-disability earnings. Thus, persons with similar disabilities could receive very different post-disability incomes depending on the offsetting practices of the Provinces/Territories providing their WCB benefit.
• PSA was found to reduce payments by 100% of CPPD in every province. Single individuals with very modest CPPD pensions (who had low incomes or made sporadic contributions), almost always received a "top up" from PSA, because PSA benefit guarantees were regularly greater than their CPPD pensions. This feature means that very low income individuals, who become eligible for both CPP and PSA, are no better off, in terms of the total benefits received because of their disability, for having contributed to CPP⁵⁴.
• LTDI generally reduced its payments by the full amount of the CPPD pension, so that, with the exception of overpayment situations, individuals with LTDI benefits, received a total equivalent set by LTDI (usually about 60-70% of pre- disability gross earnings, for some number of years).

An Earnings Replacement Analysis: The Modular Analysis Package for Systems of Income Transfers (MAPSIT) of HRDC was used to provide information on both the gross and disposable earnings replacement impacts of CPPD. Simulations were analyzed for single individuals with disabilities in four provinces (Alberta, Newfoundland, Ontario and Saskatchewan), chosen for their different provincial tax rates and their treatment of CPPD offsets. The analysis tested the replacement effects of CPPD for single individuals with pre-disability incomes of 50%, 100%, and 150% of YMPE.⁵⁵ Details are provided below.

At the inception of the CPPD, it was suggested in background studies and policy papers, that the gross earnings replacement objective should be about 25% earnings replacement at 100% YMPE pre-disability earnings.⁵⁶ The evaluation therefore employs this objective for pre-disability gross and disposable earnings replacement by CPPD.

Simulations of CPPD Impacts for a Single CPPD Beneficiary: Exhibit 10 summarizes the estimated maximum gross and disposable earnings replacement rates for a single individual in 1995 for CPPD benefits alone for pre-disability earnings levels (50%, 100% and 150% of YMPE). For purposes of analysis it was assumed that CPPD recipients have no income other than the CPPD benefit.

The percentage of gross income replaced by CPPD benefits alone for a single individual, is 41% at 50% of the average industrial wage (YMPE), declining to 29% gross income replacement at 100% YMPE and 20% gross income replacement at 150% YMPE. This is true for all beneficiaries, regardless of province.

Exhibit 10 Comparison of Percentage of Pre-Disability Gross and Disposable Earnings Replaced by Maximum CPPD for a Single Disabled CPPD Recipient for Selected Provinces (1995)

Provinces	Pre-Disability Gross Earnings at YMPE57 Point	Gross Earnings Replaced by CPPD	Pre-Disability Disposable Earnings*	Disposable Earnings Replaced by CPPD
Alberta	$17,450.00 (50% YMPE)	41%	$14,193.00 (81%)	52%
	$34,900.00 (100% YMPE)	29%	$25,349.00 (73%)	42%
	$52,350.00 (150% YMPE)	20%	$35,669.00 (68%)	30%
Newfoundland	$17,450.00 (50% YMPE)	41%	$13,865.00 (79%)	53%
	$34,900.00 (100% YMPE)	29%	$24,315.00 (70%)	43%
	$52,350.00 (150% YMPE)	20%	$33,709.00 (64%)	31%
Ontario	$17,450.00 (50% YMPE)	41%	$14,294.00 (82%)	54%
	$34,900.00 (100% YMPE)	29%	$24,967.00 (72%)	44%
	$52,350.00 (150% YMPE)	20%	$34,949.00 (67%)	31%
Saskatchewan	$17,450.00 (50% YMPE)	41%	$13,803.00 (79%)	53%
	$34,900.00 (100% YMPE)	29%	$24,342.00 (70%)	43%
	$52,350.00 (150% YMPE)	20%	$33,711.00 (64%)	31%

However, the percentage of disposable income replaced by CPPD benefits alone, for a single individual, varies slightly by province, reflecting variations in provincial tax systems, with earnings replacement rates varying from 52% to 54% at 50% YMPE; 42% to 44% at 100% YMPE; and 30% to 31% at 150% YMPE; with Ontario allowing for the highest percentage of disposable income to be replaced by CPPD, and Alberta the lowest of the four provinces used in this analysis. The percentage of CPP earnings replacement decreases as pre-disability earnings increase.

Notes on the CPPD/PSA Linkage: In order to understand how PSA supplements CPPD, Provincial Social Assistance Ministries were asked to provide further information on the benefits that would be provided to a single disabled individual with a pre-disability income level of $12,500, and who is assumed to have had a contribution history entitling that person to a CPPD benefit of $450 per month. Results indicate that PSA exceeds and supplements low-end CPPD benefits in all of the Provinces and Territories. This analysis suggests that single CPPD beneficiaries, who must rely on PSA are (statistically speaking) not likely to benefit as a result of their CPPD contributions. However, their retirement benefits will likely be higher than in the absence of CPP contributions.

As well, if eligibility for CPPD makes a beneficiary ineligible for PSA this may have unforeseen negative effects by rendering beneficiaries ineligible for special needs allowances given to PSA beneficiaries.⁵⁸ These results are seen as undesirable for individuals with disabilities who require ancillary benefits for special needs such as housing, transportation, medication and assistive devices. In this sense, CPP contributors with disabilities may be required to apply for CPPD benefits that, in the end, leave them worse off than they would have been if eligible for PSA only. Stakeholders noted that some people with disabilities were reluctant to apply for CPPD benefits because they would lose these ancillary benefits provided by the province.

3.2.5 Impacts of YBE/YMPE on Earnings Replacement

The evaluation considered the question: What are the impacts of the Year's Basic Exemption (YBE) and the Year's Maximum Pensionable Earnings (YMPE) on disability payment levels?

In 1995 the CPPD benefit was the flat-rate component ($3,838.12 per annum) plus 75% of the CPP retirement benefit that would be payable if the beneficiary were age 65. In 1995 the maximum CPP retirement benefit at age 65 was $8,558.28 per annum, so the maximum annual CPPD benefit was $10, 256.88. In addition, there may be CPPD benefits payable to the children of a CPPD beneficiary, but these were not reflected in the following analysis.

Exhibit 11 shows the CPPD benefit payable in 1996, as a percentage of earnings in absolute amounts and as percentages of the 1995 YMPE, $34,900. CPPD benefits are calculated on the assumption that the CPPD beneficiary had always had earnings at the level indicated, and had otherwise met the eligibility requirements for disability. The lowest percentage selected was 11% because no credits are earned if earnings are less than the YBE, which is 10% of the YMPE (rounded).

Exhibit 11 CPPD Earnings Replacement Rates in Relation to YMPE

Pre-disability Earnings as % of 1995 YMPE	Earnings in 1995	CPPD Full year Pension in 1996 as % of 1995 Earnings
11%	$3,839	120.5%
25%	$8,725	63.5%
50%	$17,450	41.1%
75%	$26,175	33.7%
100%	$34,900	29.9%
150%	$52,350	20.0%
200%	$69,800	15.0%

As can be seen, the CPPD earnings replacement ratio decreases with increasing earnings. The YBE has no effect on earnings replacement ratios, but does have the effect of eliminating eligibility for CPPD benefits for those who always earned less than the YBE, $3,500 in 1995, and potentially reducing CPPD benefits for those who have sometimes earned under the YBE. (However, the YBE effect on those with earnings below YMPE reduces the effective cost of CPPD benefits to this category of beneficiaries).

The effect of the YMPE is to put a ceiling on CPPD benefits, which results in a decline in earnings replacement ratios for those earning above YMPE, as earnings over this level are not taken into account in determining benefits (or contributions).

3.2.6 Level of Benefits

CPPD benefits were compared to disability benefits in other major trading partner jurisdictions. Exhibit 12 illustrates the different levels of benefit for each country included in the international comparison which was conducted as part of this evaluation. There, the PDI benefit is compared to the average industrial wage (in local currencies). This comparison relates to the minimum and maximum amount of benefits for each national PDI program.⁵⁹

Exhibit 12 Minimum and Maximum Yearly Disability Pension Amounts, as a Percentage of Average Industrial Wage, Selected Countries, 1993 (Local Currencies)

Country	Average industrial wage	Minimum Yearly Pension	% of average industrial wage	Maximum Yearly Pension	% of average industrial wage
Canada	31,697	3,748	12%	9,754	31%
Australia*	30,924	n/a	n/a	8,115	26%
Germany**	51,257	none	0%	30,186	59%
The Netherlands	54,258	18,160	33%	52,252	96%
New Zealand*	30,772	n/a	n/a	8,634	28%
Sweden	173,900	67,424	39%	166,524	96%
United Kingdom***	14,113	n/a	n/a	3,416	24%
United States**	25,179	none	0%	14,736	59%

Sources: See Technical Note #1 at the end of this report for explanation of the footnotes which are the sources of the exhibit's data.

In the international context, these data indicate that Canada's benefit levels as a percentage of average industrial wages, are comparable to benefit levels in other countries. The minimum level of benefit compares advantageously with other insurance-based programs (higher than in Germany and the United States, and is comparable to the United Kingdom). The maximum level of benefits is significantly lower than other insurance-based programs, except the United Kingdom, and is very similar to the benefit levels of Australia and New Zealand, which have programs based on social assistance principles.

This results from the particular benefit structure of Canadian PDI, with a relatively high flat-rate benefit, but a lower earnings ceiling for contribution purposes (YMPE). Thus, the Canadian benefits structure contains strong redistributive elements, which makes it quite different from the American and German systems, which are solely earnings-related. Overall, however, CPPD is, by comparison to programs of Canada's major trading partners, no less generous than the United Kingdom, Australia and New Zealand, but is less generous than some other countries examined (U.S., Germany, the Netherlands and Sweden).

3.2.7 Beneficiaries' Satisfaction With Incomes

To further assess income adequacy, the evaluation examined the perceptions of beneficiaries responding to Statistics Canada's 1995 CPPD Beneficiaries Survey.⁶⁰ That survey provided respondent ratings of the adequacy of beneficiaries' family income and personal income in relation to meeting their needs and those of their family.

Generally, those participating in the survey reported that the adequacy of family income decreased after the individual became a CPPD beneficiary. For example, 44% of beneficiaries rated their family income as adequate to more-than-adequate after their disability, as compared to 92% of beneficiaries who rated their family income as adequate to more-than-adequate prior to their disability.

The rated adequacy of personal income decreased as well after the beneficiaries were disabled. For example, 36% of beneficiaries rated their personal income as adequate to more-than-adequate after the onset of their disability, as compared to 89% of beneficiaries who rated their personal income as adequate to more-than-adequate prior to their disablement.

Considering the beneficiaries' family size, living area (urban/rural), and family income, the beneficiaries were categorized as living above or below the Statistics Canada definition of Low Income Cut-Off Line.⁶¹ Based on respondents' self-reported data, it was estimated that slightly over half of the beneficiaries (58%) reported living above the Low Income Cut-Off Line while 42% were living below the Low Income Cut-Off Line.⁶²

3.2.8 Conclusions About Incomes

The accomplishments of CPPD are extensive in extending some level of earnings protection to hundreds of thousands of Canadians (reaching nearly 300,000 disabled persons and their households in 1996). CPPD would thus appear to provide an important element of earnings replacement for working Canadians who become disabled. It accounts for a significant proportion of pre-disability gross earnings replacement, exceeding the target of 25% set out at the time of the program's conceptualization.

However, CPPD benefits are often offset by other programs, potentially leaving some beneficiaries no better off financially during their disability than if they had received no benefit from CPPD.

3.3 Eligibility Criteria: Contributions/Defining Disability

The evaluation questions addressed in this section are: Are current definitions of disability and the eligibility criteria appropriate? What are the differences between CPP, and similar programs of other countries?

3.3.1 Contributory Requirements

The evaluation considered the questions: Are current eligibility criteria appropriate for the disability benefits in light of workforce trends in particular? What are the differences between CPP, QPP and other countries' programs in defining disability, coverage and claim rates, and their implications?⁶³

Historical CPPD Contributory Requirements: In order to qualify for benefits under the CPPD program, workers originally had to have contributed to the plan for at least 5 out of the last 10 years (in terms of contributory months), meaning that, as the program originated in 1966, no disability benefits were actually paid out until 1971.

Extensive changes to the contributory requirements have been made over the years that have affected eligibility to CPPD. For example, the child-rearing drop-out provision came into effect in 1984, allowing calculations of contributory years to remove those years spent in child-rearing, when contributions might be low or nil. This would have ensured that such pensions would be more generous to this category of applicant and encouraged such applications. The contributory rules were also changed in 1986 (Bill C-116) to allow claims from workers who had contributed for either 2 out of the last 3 years, or 5 out of the last 10 years. Retroactive claims were allowed after 1992 (Bill C-57) for those who were no longer working and who would have qualified had they filed a claim for CPPD at the time of the disablement.

By these measures, contributory requirements for CPPD had become less restrictive over time. By comparison QPPD had more stringent criteria until 1993, except for the latter's more lenient treatment of 60-64 year old applicants.

CPPD/QPPD Comparison: Until 1993, contributory requirements of the QPPD have been more stringent than CPPD's and this has resulted in a lower proportion of the population being eligible for the disability benefits program.

Originally, QPPD eligibility rules required that claimants contributed to the program for at least five of the past ten years and one-third of their contributory period to be eligible for disability benefits. A somewhat relaxed provision applied after 1984 for people aged 60-64, where the "5 of 10 years" rule was waived to require that claimants had only contributed for 10 years and one-third of their contributory period. Also, it only required that beneficiaries in the age range, 60-64, be incapable of regularly carrying out gainful employment at their own occupations, rather than at any occupation.

In July, 1993, eligibility requirements were significantly changed for QPPD. Under the new provisions, contributors are covered under QPPD when: (i) they have contributed for at least 2 of the last 3 years; or (ii) they have contributed for at least 5 of the last 10 years; or (iii) they have contributed for at least half of their contributory period. Thus, while QPPD eligibility requirements were more stringent than CPPD until 1993, QPPD's eligibility rules became somewhat more liberal than CPPD's after July, 1993.

To summarize, differences in the eligibility requirements under the two programs have varied in the following ways over the past 25 years: (i) From 1970-1986: both programs required at least five of the last ten years of contribution, and a further requirement for one-third of the contributory period, but in 1984 QPPD deemed that beneficiaries in the 60 to 64 years of age range only be unable to engage in gainful employment at their own occupation,; (ii) From 1986-1993: unlike QPPD, CPPD allowed eligibility to the program under the more relaxed "2 of 3 years" rule and permitted retroactive claims, while QPPD kept the more stringent additional criterion requiring contributions for at least "one-third of the contributory period";⁶⁴ and (iii) Since July, 1993: QPPD eligibility rules are somewhat more liberal, since they have introduced their own "2 of 3 years" rule, and also allow people who have contributed for half of their contributory period (with a minimum of two years of contributions) to be covered under the program, even if they do not meet the recency-of-work test.⁶⁵ These differences have had significant effects over the past 25 years.

As can be seen in Exhibit 13, the "one-third of the contributory period" requirement of QPPD seems to have had little impact on the eligibility of males in the 1976-87 period, but a very significant impact on the eligibility of females during the same period. While the data indicate that the percentage of the female population aged 20-64 covered under the CPPD program has averaged 75%, the proportion of females covered under the QPPD program was significantly lower, with an average of only 52% eligible for QPPD. Thus historically, eligibility for all workers, but especially females, has been significantly less under QPPD than CPPD.

The difference is marked, but considerations of causes should be noted in interpreting these results. These data do not take into account the child-rearing drop-out provision that has been in effect in Quebec since 1977 and in the rest of Canada since 1983. The child-rearing drop-out provision has a significant impact on the contributory period of females, and thus would greatly influence the eligibility under QPPD's "one-third of the contributory period" requirement.

Exhibit 13
CPPD & QPPD Historical Eligibility by Sex
Percentage of Population, Aged 20-64, 1976-1994

Some of this difference may be attributed to demographic and sociological patterns which may have historically resulted in a smaller percentage of females in the labour force in Quebec compared to the other provinces. Such a trend may be indicated by the fact that the proportion of females eligible for disability benefits has increased more rapidly in Quebec between 1976-87 than in the rest of Canada (by 15% in Quebec compared to 6% in other provinces).

In the 1986-93 period, the gap in population coverage between the QPPD and the CPPD was further accentuated by the introduction of the "2 of 3 years" rule of eligibility under the CPPD. During this period, the difference in program coverage is significant for both males (an average gap of 13%) and, to a greater extent, females (an average gap of 25%).⁶⁶

Finally, in the 1993-94 period, with the introduction of the new contributory requirements in Quebec, most of the program coverage differences disappeared. Program coverage among males is slightly higher under QPPD (94%) compared to CPPD (91%), as would have been expected considering the somewhat more "liberal" eligibility criteria for QPPD, but slightly higher for CPPD females (84%) as compared to QPPD females (80%).⁶⁷

This "gap" in program coverage between CPPD and QPPD has no doubt contributed to the higher disability caseload at CPPD, since a higher proportion of the CPPD population met the insurability requirements.

In conclusion, it appears that program coverage by QPPD has been historically lower than that of CPPD. This difference may explain a noteworthy portion of the differences in the number of grants and total caseloads under these two programs.

International Comparison: While detailed data were not available for all countries with respect to specific insurance requirements, the available evidence suggests that the Canadian programs provide broader coverage of the working population compared to those countries where coverage is conditional on labour force participation. For example, at least two countries which have been studied for this evaluation (Germany and the United States) have more stringent requirements for coverage under their PDI programs.

In Germany, a claimant must have contributed for at least 36 months over the last five years and at least 60 months overall to be covered under the disability program. In the United States, while eligibility requirements vary according to the age of claimants, eligibility requirements are more stringent than Canada's for people aged 31 and over.

To be covered under the U.S. PDI program, claimants aged 31 and over must have contributed for at least one-quarter of coverage for each year that has elapsed since age 21 and at least 20 of the 40 quarters immediately preceding the onset of disability.⁶⁸

But coverage of the working population is higher in a number of other countries, such as Sweden, which offers a universal basic pension to disabled recipients and work-related benefits, and the Netherlands, which provides both earnings-related benefits for disabled workers and basic benefits for all residents. Such countries with disability benefit programs based on welfare principles offer universal coverage, but these programs are not directly comparable to PDI programs, such as the CPPD.

3.3.2 Definition of Disability

General Definitional Issues: An extensive literature indicates that what at first glance seems a simple matter - deciding who is and who is not disabled, for example, on the basis of a medical definition - is not simple at all. This issue, of defining disability, has been a problem for PDI systems throughout the world. Indeed, a considerable amount of the relevant literature portrays disability as a complex phenomenon that results both from the presence of medical impairments and from the demands imposed on workers by the structure of work itself (physical barriers, pace of work, etc.).⁶⁹

This complexity is further accentuated by a worker's personal vocational characteristics, since different types of workplaces make different demands of workers (e.g. the office vs. the assembly line) and therefore workers with the same severity of disability may face dramatically different opportunities for employment, e.g., job mobility.

Finally, socio-economic considerations may greatly affect the employability of persons with disabilities, as the particular jobs which lie within their capacities are reduced or eliminated by industrial restructuring.⁷⁰

This perspective is supported by findings which suggest that only a percentage of persons defined by HALS as having disabilities (using a different measure of disability from CPPD, as was noted earlier) are actually in receipt of government benefits, whether PDI or public assistance.⁷¹ These findings emphasize the important distinction between having functional limitations and being unemployable.

These considerations point to the need for definitions to reflect additional criteria. For example, the potential for the disabled person to undergo occupational therapy and rehabilitation can be considered, in order to determine whether a person can work or not. This is generally seen as a task which medical assessments alone cannot judge.

CPPD/QPPD Definition of Disability: CPPD Benefits are available to persons who have "a severe and prolonged mental or physical disability [..]. A disability is severe only if by reason thereof the person in respect of whom the determination is made is incapable regularly of pursuing any substantially gainful occupation⁷²; and a disability is prolonged only if it is determined in a prescribed manner that the disability is likely to be long continued and of indefinite duration, or is likely to result in death."⁷³

This official definition is supplemented by a number of adjudication guidelines which define, in operational terms, how disabilities are to be recognized for the purpose of granting CPPD pensions. These guidelines were revised in September 1995,⁷⁴ in order to ensure greater uniformity in the adjudication process. They provide information on the interpretation of the official definition in light of the jurisprudence of the CPP Pension Appeals Board and past practices of CPPD. For instance, a "prolonged" disability has been defined in operational terms as a disability "likely to last for at least one year". On the other hand, the QPPD definition of "prolonged" is that the disability will last indefinitely.

The QPPD treatment is somewhat different, even though the statutory definition of disability is very similar under both CPPD and QPPD. Both programs compensate contributors who are subjected to a "severe" and "prolonged" disability. The QPPD has been more rigorous in that it has established detailed standards on how medical conditions affect work capacities and at what stage they should be considered disabling. This is offset, however, by other aspects of adjudication which make the program more lenient for the pre-retirement group (ages 60-64) only.

QPPD's Emphasis as a Pre-Retirement Program: While the overall approach of QPPD to the definition of disability is strict, its approach to older workers produces a significant counter-effect. This approach was enshrined in Quebec legislation in 1984, making the criteria for workers aged 60-64 much less onerous than for other applicants. Workers in this older age group are tested much less stringently by a test of "ability to do one's own job" (emphasis added), rather than ability to engage in any substantially gainful employment. Similarly, CPPD evidenced a priority for pre-retirement applicants, (55 to 64 years of age) as outlined in administrative directives in 1989-95, but this directive was eliminated as part of the September 1995 revisions to the guidelines. However, CPPD's pre-retirement emphasis was less than that found for QPPD, as suggested by caseload statistics, and was removed from new CPPD policy guidelines in 1995.

The pre-retirement emphasis of QPPD is suggested in Exhibit 14 (next page), which shows that nearly half of new QPPD pre-retirement beneficiaries (48%) in 1993 were in the 60-64 age group. In contrast, fewer than one-quarter of CPPD's new 1993 beneficiaries were in this age group.

Exhibit 14* Percentage of New QPPD and CPPD Beneficiaries by Age Groups in 1993

Age	QPPD Count	Percentage	CPPD Count	Percentage
20-54	2,761	33%	30,274	52%
55-59	1,544	19%	14,990	26%
60-64	4,031	48%	12,707	22%
Total	8,336	100%	57,971	100%

The pre-retirement orientation of QPPD has two important effects. First, this policy may have encouraged proportionately more applications by 60-64 year olds.⁷⁵ Second, because QPPD applies more stringent criteria to younger persons with disabilities, the evidence suggests that those rejected by QPPD seek other PDI supports or PSA. As will be noted later, many of these individuals ultimately become beneficiaries of provincial social assistance.

Findings of a QPPD Study on the Eligibility for QPPD of a Random Set of CPPD Applicants in 1993-94: Physicians with the Quebec Pension Plan Disability (QPPD) program recently carried out a random file review of 477 CPPD adjudication decisions made in fiscal 1993-94, comprising 340 grants and 137 denials. This was to ascertain what proportion of these past CPPD applications would have met the QPPD adjudication criteria⁷⁶.

The conclusions of this QPPD study on the applications declined by CPPD generally confirmed the original CPPD decision to deny these applications. On the other hand, QPPD would have granted only 27% (92 cases) of the 340 CPPD grants on the basis of the evidence at hand. A further 44% (149) of these CPPD grants were identified as requiring more information, before a decision to grant or deny could have been made. About 14% (47 cases) of the CPPD grants would have been denied by the QPPD⁷⁷. Of the 47 cases denied by the QPPD (74% did not meet the "severity of disability" criteria of QPPD. Particularly noteworthy were the differences in QPPD decisions regarding musculoskeletal and mental disorders accounting for about 40% (135) of the CPPD grants examined by the CPPD physicians. The QPPD would have denied 30% of the musculoskeletal cases, and requested additional information for 33% of these cases before making a decision. In the case of mental disorders, the QPPD would have denied 9%, and requested additional information in 60% of these cases. The usefulness of this QPPD study is somewhat limited since the CPPD grants examined were not made on the basis of the more strict current CPPD adjudication guidelines introduced in September 1995⁷⁸.

In a further analysis, Income Security Programs (ISP) Branch, HRDC reviewed a randomly selected sub-set (119) of the 1993-94 grants which would have been denied by the QPPD physicians against the CPPD September 1995 Guidelines, and found that it would now deny about 11% of these grants and require more information for 21% of them⁷⁹. ISP Branch also reviewed 25 CPPD decisions (24 grants and one denial), where the QPPD decision would have been different from that of the CPPD, against the September 1995 guidelines; one application would have been denied, seven (28%) granted and more information requested in the case of 12 applications (48%).

Conclusion: CPPD and QPPD differ somewhat in terms of operational definitions of disability, but especially in terms of how this affects access to the program. Key differences are the extent to which QPPD operate more as a pre-retirement program, and its more stringent medical criteria, with the result being somewhat lower caseloads.

QPPD generally requires more documentation before making initial grants; it does not recognize certain conditions as severe enough to prevent one from working (certain kinds of musculoskeletal conditions and some mental disorders). Determining the eligibility for both CPPD and QPPD disability is very complex and requires expert medical judgement.

3.4 Incidence: Economic Grants and Other Explanations

3.4.1 Background on Incidence

Some specific evaluation questions posed in this section are: Why has there been an increase in CPPD Disability claims during the last 15 years, including the recent recession? Is the increase in CPPD Disability levels justified or is it caused by other factors, which, if addressed, would lower incidence levels?

Exhibit 15
Number of CPPD Applications
Number of New Pensions Put in Pay, Grants as a % of Applications
Fiscal Years 1969-70 to 1995-96⁸⁰

CPPD Claim Patterns and Profiles: The pattern of CPPD claims has shown an almost steady increase in applications over the past twenty years (Exhibit 15). This increase was from an annual number of 27,966 applications in 1975-76 to a high of 109,000 in 1993-94, then a decline to 90,449 applications in 1995-96. It has also been marked by several years when numbers of grants to new beneficiaries have jumped significantly. The most noticeable peaks for new applications occurred in 1985-86 and again in 1992-93, while the years with the highest percentage of applications which were granted CPPD benefits were 1980-81, 1985-86, and 1988-89.⁸¹

It is important to note, however, that the percentage of CPPD applications granted benefits between 1991 and 1995, 58%, was significantly lower than the 65% over the 1986-90 period.⁸² In comparison, the growth in the number of applications at QPPD has been generally slower than at CPPD. While QPPD has, like CPPD, experienced surges of applications in periods of economic recession and high unemployment, applications have, unlike CPPD, tended to level off and decrease more rapidly in periods of stronger economic activity.

This seems to indicate that differences in program caseloads and incidence rates are, to a large extent, driven by applications, especially since grant rates have not varied significantly between the two programs.

Reasons for QPPD's Lower Incidence Rate: The causes of lower incidence of QPPD payments in Quebec are complex and no single factor is likely to provide a satisfactory explanation as to why the caseloads and costs of the QPPD program have been kept at a much lower level than its counterpart in other provinces. Several of these explanations will be discussed.

Generally, it appears that QPPD is situated differently within the broader system of income security for individuals with disabilities in Quebec. This is exemplified by the clear delineation between work-related disabilities (which are, as a rule, solely compensated through the Quebec-WCB), and other disabilities which can give rise to QPPD pensions. The separation of most WCB cases out of QPPD means that QPPD caseloads are reduced by this rationalization of the two programs.⁸³ Another example of the particular emphasis of the QPPD program is the fact that more rigorous criteria seem to be applied to the population of younger persons with disabilities (aged under 60), while older workers (aged 60-64) are much more readily admitted on the disability rolls. Thus, demographics and work force participation would explain a substantial part of the QPPD caseload's difference from CPPD's.

The nature of QPPD as more of a pre-retirement program, and its concentration on older workers, means that its caseload turns over at a rapid rate, particularly by beneficiaries reaching age 65 and transferring to the QPP retirement program.

Certain data also indicate that some of the difference between CPPD and QPPD caseloads may be related to a difference in disability incidence rates in Quebec versus other provinces. According to the 1991 HALS, disability incidence rates, measured with objective questions on physical ability and limitations, (although not by the "severe" and "prolonged" definitions used by CPP and QPP) indicated that the incidence of disability (any limitation) is higher outside Quebec by 3.3%. Whereas disability incidence was 13% for Quebec, incidence was 16.3% in the rest of Canada. Similar results are reported in the 1991 Census. This finding is little understood, however, as there is no obvious reasons why disability should be less in Quebec, and should probably be the topic of further research.⁸⁴

Use of Other Disability Income Programs in Quebec: One element that is often overlooked in comparisons of CPPD and QPPD is that lower incidence of disability payments at QPPD does not necessarily mean that Quebec faces lower public costs for disability in general. Overall, while the Quebec income security system appears to achieve considerable efficiencies in caseload and cost controls resulting from the unique orientations of QPPD (retirement orientation, tighter adjudication including detailed standards on medical conditions and more use of independent medical assessments), it appears that it simultaneously incurs increased cost in other public programs, since more persons with disabilities appear to be channeled to PSA for income support.

This is illustrated in Exhibit 16 (next page), where it can be seen that in fact a higher proportion of Quebec's persons with disabilities are in receipt of public assistance of some type than in the other provinces (HALS, 1991).⁸⁵ Indeed, the overall total reliance of persons with disabilities in Quebec on public funds as a source of income (23.9%) is not less than that of people with disabilities in other provinces (18.1%), but apparently somewhat greater.

Exhibit 16 Income Support Programs Used by Persons With Disabilities, Canada, 1991* (Percentage of all Persons Reporting Disabilities Obtaining Income
From Any of these Sources, HALS, 1991)

Program	Quebec	Other Provinces
CPPD/QPPD	9.6%	11.9%
WCB	5.5	7.1
Social Assistance	16.4	9.0
LTD	4.1	5.2
Any public program**	23.9	18.1

While these data are only indicative, they highlight the importance of looking at the phenomenon of disability and income replacement from a broader system perspective (including WCB, PSA⁸⁶), rather than judging the efficacy of a system from the point of view of only one type of program such as CPPD or QPPD.

The Single Payer Approach in Quebec: In Quebec, persons with disabilities tend to receive their disability benefits from a single program (QPPD, WCB, the Quebec Public Auto Insurance Plan, or Provincial Social Assistance), and specifically from only one payer in that same integrated system. A person with disabilities in Quebec is thus less likely to receive public benefits from more than one source, than is a person with disabilities in any other province. As a result, the "single payer" system in Quebec seems to avoid much of the complexity that exists for CPPD because of the overlaps and administrative duplications of a "multi-payer" system.

In conclusion, these findings suggest that while some lessons from the QPPD program are potentially relevant to CPPD (more rational program linkages, tighter adjudication including greater use of independent medical examinations, etc.), other features (pre-retirement emphasis, lower coverage of the female population, shift of costs to social assistance) imply a program model which may not be totally desirable for emulation by CPPD.

Yet QPPD has moved closer to CPPD in eligibility criteria in recent years, and harmony between these two programs may be desirable from a national perspective, as will be discussed further in the conclusion to this report.

International Trends: International comparison demonstrates that increasing PDI claims are a common phenomenon (see Exhibit 17, next page). Indeed, the comparative review conducted for this evaluation indicates that problems faced by the CPPD program are very much like those facing the PDI programs of all modern industrial nations today. This has raised a number of concerns and has prompted extensive reforms of disability benefits programs in several countries.

Most countries examined, except Germany, have experienced rapidly rising PDI caseloads and report substantial struggles with the challenges of definition, adjudication and program administration. These trends also appear to have been affected by long-term unemployment patterns found in these countries in the 1980s and 1990s. The specifics of individual countries' PDI program criteria, partial disability benefits, and the age distribution of the population are also factors.

Disability Caseloads as a Percentage of the Population: The evolution of PDI caseloads is only one indication of international trends in disability benefit programs. A somewhat better guide for assessing PDI caseloads is the incidence of disability payments to the overall population covered by the program. However, as all programs have different eligibility requirements, it is very difficult to obtain reliable denominators which would represent the number of people covered by the program.

In the number of people aged 20-64 was used as a proxy for the denominator in such an analysis. These data must be interpreted cautiously, as programs vary considerably in terms of population coverage (eligibility), coverage of partial disability, provision of partial disability benefits, and linkages with other disability benefit programs (see detailed exhibit notes). Nonetheless, the data are indicative of the relative size of different countries' disability programs.

Exhibit 17 International Caseload Growth, CPPD/QPPD, and Other National Disability Benefit Programs, Selected Countries, 1983-1993

(Program) Country	Caseload (to nearest 1,000)		Percentage Increase	1993 Caseload
	1983	1996		As % of Population Aged 20-641
(CPPD)*	102,000	234,000	129%	1.8**
(QPPD)	30,000	43,000	42%	1.0
Australia	220,000	407,000	85%	3.9‡
Germany***	1,724,000	1,405,000	(-19%)	3.4†
United Kingdom	700,000	1,580,000	126%	4.6
Netherlands	728,000	921,000	27%	9.7†
New Zealand	19,000	35,000	84%	1.8‡
Sweden	308,000	402,000	31%	8.0†‡
United States	2,569,000	3,726,000	45%	2.5

* Fiscal year ending, from Annual Statistics, December, 1994, HRDC.

** It should be noted that the 1993 figure for CPPD does not capture the full impact of the recent surge in CPPD incidence to 1994.

*** Caseload growth figures for West Germany are from 1985 to 1994 and the ratio is pro-rated. Ratios were estimated based on partial time-series.

† Programs include provisions for partial disability.

‡ Universal coverage.

N.B.: It is important to note that these figures show only the CPPD and equivalent programs, i.e. nationally mandated PDI programs, without regard to the greater context of work injury compensation, public social security, private LTD insurance, etc. Employment injury in these countries is covered by the programs; the exceptions are Canada where this is done by the WCBs and LTDIs , Germany by trade and professional associations and in the US by LTDIs.

¹ Note that these figures do not control for age distributions within the 20-64 group. This varies from country to country, with some countries having "older" population distributions within the 20-64 group.

The two countries with the highest percentage of beneficiaries - the Netherlands and Sweden - are also the two countries offering the highest levels of benefits, very broad coverage and partial disability benefits. Germany, following an increase in the required contribution period in 1984, which initially reduced the benefits paid to women by one-half, still maintains a relatively high percentage of beneficiaries considering the stringency of its contributory requirements.⁸⁷

The United Kingdom, the United States and Canada share many characteristics, being the only countries whose programs are based primarily on insurance principles and not covering partial disability. Their respective caseloads, however, are somewhat different. The United Kingdom has a relatively high caseload (4.6%), considering that coverage is not universal and that benefit levels are similar to those of Canada.⁸⁸

With respect to the United Kingdom, it should be noted that high disability caseloads have been seen as a serious problem; consequently eligibility was sharply curtailed in 1995 by the development of strict guidelines for determining disability.

The Canadian and American systems have very similar definitions of disability and thus present the best comparison. Both programs require a disability that prevents claimants from pursuing any "substantially gainful occupation," and one that is likely to last for at least one year. The two programs also consider a number of vocational criteria (such as the age, education and work experience of claimants) in the adjudication of benefits, but exclude socio-economic factors. Finally, they have broadly similar program coverage, although contributory requirements are generally somewhat less stringent in Canada. Yet, CPPD was seen to compare favourably, with a difference in the 1993 disability caseloads between the two programs (1.8% of the population aged 20-64 in Canada, compared to 2.5% of the population aged 20-64 in the United States).⁸⁹

Despite the fact that Canada has experienced the most rapid increase in disability caseloads over the period 1983-93, the percentage of disability recipients in the population aged 20-64 still remains one of the lowest at the international level. This is not an indication that there is no "increasing incidence" problem in Canada.⁹⁰ However, these data suggest that, from the point of view of international comparisons, Canada is not dissimilar from its major trading partners. In fact, since CPPD is younger than most similar programs in other countries, it is possible to argue that Canada is only starting to experience the same kind of problems that other countries have encountered in the early 1980s, and again in the 1990s. This is no consolation for CPPD, since the international comparison has demonstrated that while PDI caseloads tend to expand rapidly in periods of weak economic activity and high unemployment, disability caseloads do not automatically decrease when the economy has recovered.

Economic/Intergovernmental Factors and Caseloads: Comparisons with other countries suggest that the use of vocational criteria (socio-economic factors, personal characteristics) has often been identified as one of the causes of increasing disability rolls. Disability benefit programs that apply a broad definition of disability have been particularly vulnerable to variations in economic activities and have often had to deal with larger disability caseloads in recent economic downturns.

Higher benefit levels (amount of the CPPD pension relative to average industrial wage) and more generous eligibility criteria, which were introduced in various changes to the CPPD program effective in 1983-84, 1986, and 1992-93, also have apparently had significant effects in increasing caseloads.⁹¹

These effects are often difficult to distinguish, when multiple impacts occur over the same time periods. For example, 1987 saw two important program changes for CPPD: (1) expanded eligibility; and (2) a significant increase in the dollar value of CPPD pensions. Similarly, the 1992 changes were complex in allowing retroactive eligibility, opening the door for a significant movement of cases from PSA to CPPD.

As a result of the 1992 program changes, and a special Federal-Provincial agreement with Ontario, a special project was established in Peterborough to process applications for CPPD which were brought forward by the Ontario Ministry of Community and Social Services. In 1993-94 this project, associated with a special CPPD project office located in Peterborough, processed over 16,000 CPPD Ontario applications from PSA recipients. This Federal-Provincial project resulted in over 6,000 new CPPD pensions, a significant part of the accelerated caseload increase for CPPD in 1993-94.

The role of the economy has also been identified as a significant influence on the rise in applications for CPPD. This is discussed in the next section.

Demographic Effects: Demographic changes also played a role in the uptake of CPPD. The aging of the "baby boom" generation⁹², those born between 1946-64 (which augmented the number of people in the middle-age category in the late 1980's-early 1990's), the increasing labour force participation of women and the credit-splitting provision (which increased the number of women covered by CPPD⁹³), were associated with increased CPPD applications and grants.

These demographic effects were illustrated in a statistical analysis of CPPD caseloads (see Lazar paper, footnote 64). In this analysis it was estimated that, controlling for economic and other factors, CPPD caseloads over the period 1986-94 would have been 14% lower for males and 12% lower for females, if population and its age/sex characteristics had remained constant at their 1986 characteristics.

3.4.2 Economic Grants

The evaluation considered the question: Do CPPD disability benefits constitute economic grants (do they act as a form of unemployment benefit or bridge-to-retirement benefit)?

A specific question for the evaluation centered around the issue of "economic grants" or the extent to which CPPD acts as a form of unemployment benefit or bridge to retirement (because of pensions being awarded for these reasons, rather than on medical disability grounds). This issue has come to the fore since a 1989 policy guideline suggested that CPPD claimants over the age of 55 should have their claims adjudicated on the basis of their ability to perform (own occupation) employment as well as on the basis of the strict program criteria. This issue has gained wide attention since research in the early 1990s (particularly in the U.S) pointed out that a very high percentage of males aged 55-64 are not in the labour force and are receiving disability pensions.

"Economic Grants" As Seen in the Literature: The literature review for this evaluation addressed the issue of possible economic causes behind increasing beneficiary caseloads and found an extensive body of research.⁹⁴ Maki, Leonard, and other econometricians argue that the provision of PDI benefits has caused a decline in labour-force participation, particularly among males over the age of 55. The labour-force effects of PDI depend on the extent to which workers who are capable of working, choose to become PDI beneficiaries instead. Leonard, for example, quotes Parsons's 1980 observation of a correlation between the growth of PDI beneficiary rolls (in the U.S.), and the growth of labour force non-participation among males in the same age range, and asserts that it is difficult to "explain away" this correlation without reference to labour-supply effects of PDI.⁹⁵

However, there is an alternative view among researchers⁹⁶ (mostly in the U.S.) which asserts that the labour-supply effects of disability insurance are minimal, that health considerations are salient, and that labour force withdrawal is affected by a variety of other factors, which are socio-economic rather than medical, but which cannot be affected by manipulating benefit rates or eligibility criteria. Haveman and Wolfe, for example, employ econometric analyses similar to those of Leonard and Maki, but describe much smaller labour-supply effects.⁹⁷

The term "economic grant" is often used loosely and it is important to distinguish between a variety of economic factors that may be considered. First, disability rolls are affected by economic behaviours of people with some disabilities who must determine whether or not to apply for disability pensions. Economic factors such as benefit levels, high unemployment, accessibility of other programs, and contributory requirements, play an important role in these economic decisions, creating incentives or disincentives to apply for benefits. Secondly, adjudication criteria must serve as a test to prevent people with mild disabilities from entering the PDI program. A relaxation of these criteria over time will result in a greater number of people being admitted to the rolls, regardless of employment potential.

These two factors, individually or combined, can have a marked effect on the number of disability benefits which are granted. They are often discussed in the context of "economic grants", but one factor is primarily external to the CPPD, while the other relates directly to the administration of the program. For this reason, the two factors have been separated out in this report. This section addresses the issue of economic behaviours affecting disability grants, while the next section will focus on the adjudication of benefits and the uniformity of criteria used over time.

Economic Analyses: For many people, over the last few years, a number of factors may have "tilted" the balance of economic considerations in favour of applying for CPPD benefits. These factors have included a significant increase in benefit levels in 1987, broadening of the eligibility criteria, changes to the rule for retroactive eligibility, tightening of eligibility for other disability programs and alternative sources of income, and cost-containment pressures from other social security programs that encouraged contributors to apply for CPPD benefits. While some of these factors preceded the sudden increase in disability grants at CPPD by several years, their effects, combined with the recent economic recession and higher unemployment, may have encouraged more people with disabilities to apply for CPPD benefits. However, generally speaking, the decisions to apply for benefits did lag these factors and events.

To assess the role of economics versus other factors in the granting of benefits, econometric analyses were conducted, examining the role of benefits rates, unemployment rates, and other factors, in the flow of applications to CPPD.⁹⁸ The results provide mixed evidence, but generally seem to support the claim that broad economic factors have constituted a major factor in caseload increases. Evidence in favour of the conclusion that there have been economic grants, is seemingly provided by the econometric analysis of impacts of unemployment and economic fluctuations, and benefit rates. Particularly, the results suggest that in times of recession and economic restructuring, applications to CPPD have increased dramatically.

Indeed, regression simulations conducted for the evaluation suggest that, had economic conditions remained at 1989 levels to 1994, CPPD caseloads would have been significantly lower, at 199,000 to 215,000 cases as compared to the actual 295,000 cases.⁹⁹ As well, the results indicate that when pension benefits (cash amount of the pension) have increased, applications have also increased.

A separate study conducted for the evaluation examined the labour supply response to the 1987 rise in the flat rate portion of CPP disability benefits. The regression analysis focused on males in the 45 to 59 age group, and found a significant labour supply response. The study concluded that the number of males 45 to 59 droppping out of the labour force in response to the 1987 increase in benefits was in the range of 24,030 to 30,700 (which corresponds to between 1.8 and 2.3 percent of males in this age group in the provinces receiving CPP). The study also suggested that large increases in benefits may have an 'announcement effect'. This means that introducing a large increase in benefits may result in larger labour supply responses than introducing the same increase as a series of smaller benefit adjustments over time.¹⁰⁰

It should be emphasized, however, that these results do not provide evidence to support the view that individuals who have, in some cases, applied for CPPD when unemployed or when benefit rates were higher, should not have qualified¹⁰¹. In fact, a number of other factors can explain these findings with equal force, such as:

• people with disabilities may be disproportionately affected by economic downturns, especially where economic restructuring aims at a "lean, mean" corporate profile. Under this argument, many less productive individuals may have lost "sheltered" employment in the 1990 recession; and
• generally (and particularly in light of CPPD rate increases which occurred in 1987, and easier eligibility after 1992) CPPD may have led PSA administrations to encourage individuals to apply for CPPD. This argument is supported by evidence of referrals of PSA recipients to CPPD, particularly in Ontario in 1993-94. Some LTD administrators likely also did the same. Of course, many other disabled persons normally apply for PSA because they are not eligible for CPPD or WCB benefits.

Equally important, a number of the studies examined in the literature review have emphasized that applicants who apply for and are rejected by PDI programs are unlikely to re-enter the workforce. Rather, they are far more likely to receive income support from welfare. Thus most PDI applicants whose applications were rejected appear to wind up in poverty. This finding - based on the premise that employable individuals would return to the workforce to obtain an adequate income - suggests that most people denied PDI have reduced opportunities to attain for substantial employment.¹⁰²

This general evidence against the occurrence of economic grants is complemented by indications that new CPPD beneficiaries report a slightly higher number of disabilities over the period 1991-94. (This issue is examined further in Section 3.5.2. below). This indicates that, whatever the actual incidence of economic grants, these have likely not been more common in recent years, and are not correlated with the 1991-94 increase in CPPD caseloads. Before returning to this issue, however, we examine the issue as to whether CPPD has been used extensively as a bridge-to-retirement.

To What Extent Is CPPD a Pre-Retirement Bridge?: A review of CPPD benefits granted, by age and gender, over the past 15 years (calendar 1981-95), shows a steady increase in the distribution of new benefits to both females and to younger workers. Males' share of the total number of new CPPD benefits has declined steadily from a peak in 1983 of 71.3% to 55.5% by 1995 (Exhibit 18, next page).¹⁰³

New benefits for females over the same period showed an increase, with the highest rate of percentage point increase of 17% (from 13.6% in 1981 to 30.6% in 1995) for the youngest age group (20 to 44 year-olds), followed by 11.8% percentage point increase (from 23.5% in 1981 to 35.3% in 1995) for 45 to 54 year-olds. These changes relate at least in part to the increasing eligibility of women for CPPD, as a result of their increasing labour force participation, and their accumulation of more years of CPP contributions. Among males the rate of increase in new benefits for 20 to 44 year-olds has also risen steadily, from 13% in 1981 to 25.6% in 1995. The new benefit rate increase for 45 to 54 year-old males has also shown a trend upwards from 22.8% in 1981 to 30.5% in 1995. In contrast, the high percentage of new benefits for 55 to 64 year-old males peaked in 1986 and fell sharply starting in 1987, when early retirement provisions were implemented, with the rate continuing to fall to the present. Those aged 60-64 have declined most substantially as a percentage of new beneficiaries, in spite of the aging population.

These findings suggest that the effect of "older workers" using CPPD as an early retirement program may have been overstated as a major factor in the increasing CPPD caseload, although the actual numbers of 55 to 64 year-old males granted CPPD benefits have indeed increased, as have benefits granted to all other age categories, and the incidence of benefits for this age group has continued to rise.

Exhibit 18
Percentage Distribution of New CPP Disability Benefits
by Age Category and Gender, 1981-1995

The CPP Disability Incidence Study notes that the largest absolute increases in the CPP disability caseload are occurring among those men and women approaching retirement age.¹⁰⁴ From an incidence perspective it revealed that one in six men in the age group 60-64 in 1994 were current CPP disability beneficiaries. The corresponding incidence rates were substantially lower for women in the same 60-64 age bracket (one out of twelve). However, the incidence proportions for both men and women, aged 60-64, were much higher proportions than for the other age groups for men and women (18-34, 35-44, 45-54, 55-59, age categories), and have been growing over time.

The study concludes that some CPP disability benefits might have been awarded on some basis other than disability per se¹⁰⁵. At the same time, although the absolute caseload growth over the 1990-94 period occurred disproportionately among those approaching retirement, the relative growth in the CPPD caseload was concentrated in the 35 to 49 years of age group, peaking in the 45 to 49 years of age range.

The 1995 CPP Disability Incidence Study also conducted a comparison of the age-specific fractions of the 1991 census population that were CPPD beneficiaries and the corresponding proportions of the 1991 Health and Activities Limitations Survey (HALS) survey frame classed as HALS 'severe limitations'. It found that the CPPD incidence rate for those in the 55 to 64 years of age bracket was significantly higher than the corresponding HALS severe limitation incidence for this same age group, whereas for younger age groups the HALS severe limitation incidence was substantially higher than the corresponding CPPD incidence¹⁰⁶.

This evidence indicates that, with respect to these HALS self-identified activity limitations, CPP disability benefits are being awarded in a substantially greater proportion to those approaching retirement relative to this group's population, than for younger groups. It thus suggests the possible existence of economic grants that are providing bridging benefits to retirement, as noted in another recent study¹⁰⁷.

The CPP Disability Incidence Study concludes that the caseload increase arose from a variety of causes, including increased public knowledge of CPPD, legislative changes providing enhanced benefits and permitting for application for past benefits, and encouragement to apply for CPPD by Provincial Social Assistance (PSA) and private LTD insurance providers.

International Experience: "Economic grants" have been a concern in other countries besides Canada, especially those countries which had a broad definition of disability, including a wide range of vocational criteria.

The Netherlands has been historically one of the jurisdictions with the broadest definition of disability (i.e. benefits are related to percentage loss of earnings capacity). In the Netherlands PDI program, this was reflected not only by considering a wide range of personal characteristics in the adjudication of claims, but also by factoring into the decision-making process such aspects as labour market characteristics and the state of the economy. In this sense, Dutch disability benefits were often used as a form of unemployment insurance for people with mild disability who still retained good employment capacities. The same problem has also been noted in other countries such as Australia and the United Kingdom.

The international experience indicates that such an approach to disability benefits is not fiscally viable in the long-run. Such practices, for example, resulted in a remarkably rapid expansion of caseloads for the Dutch program. As a result, extensive reforms of the Dutch PDI system were introduced in 1987 and 1993 which considerably curtailed eligibility to the program, and especially the use of vocational and socio-economic factors in the adjudication of claims for disability pensions.

Similarly, the United Kingdom and Australia have also positioned their PDI programs to use a more medically oriented definition of disability over the last few years, as part of a general program retrenchment. These trends suggest that, to whatever extent CPPD allows or encourages economic or pre-retirement grants¹⁰⁸, it would be appropriate to review this aspect of the program based on future experience with the newly adopted 1995 Adjudication Guidelines.

The data at hand was not available to quantify the significance of the role of economic grants on the rise in demand for CPPD benefits.

3.5 Adjudication of Claims

Some specific evaluation questions posed in this section are: Are criteria applied in a consistent and equitable manner by medical practitioners and appeal boards? Are the claims being assessed properly?

There have been suggestions that more generous adjudication of claims at CPPD has been one of the primary factors in the increase in disability benefits. For example, HRDC's internal Disability Incidence Study suggested that some 3% of disability grants made in 1994 were not consistent with adjudication guidelines.¹⁰⁹ Moreover, the historically high incidence of reversals of first-level decisions regarding CPPD claims for benefits seems to indicate that the interpretation of the definition of disability has, in the past, varied between the different levels of adjudication, with the initial level providing a stricter interpretation of eligibility and disability. On the other hand, a 1995 CPP Random Review, provided a positive assessment of consistency of administrative procedures, but no data on the role of socio-economic factors in adjudication, or the extent of disability, or any indicators of employability of applicants.¹¹⁰ Similarly, file audits over the years have found little evidence of actual fraud in CPPD. As well, although the CPP Disability Incidence Study indicated that while some part of the CPPD caseload in 1994 may have been made up of persons who did not meet the then current CPPD definition of disability (i.e., were not regularly incapable of pursuing substantial gainful employment), it concluded that at most this would account for a very small portion of the overall caseload growth experienced.

3.5.1 A QPPD/CPPD Comparison

Data were analyzed from the 1991 HALS, which provides a picture of Canadians in Quebec and in the rest of Canada as regards their activity limitations, some of which may be associated with disability. The premise of this exploratory research was that by understanding the characteristics of those granted benefits under CPPD and QPPD, the similarities or differences of program criteria could be determined. One might view the analysis as providing a "window" on the administration of CPPD. This was done using the multi-variate technique of discriminant analysis.¹¹¹ Key results were:

• Objective criteria: the analysis suggests that objective factors (measured by the 1991 HALS, via self-reported activity limitations) could be used with about 80% accuracy to predict which activity-limited people would qualify for CPPD and QPPD, with the same general factors being prominent for both programs (such factors as age, number and types of activity limitations, need for assistance with activities of daily living, etc.).¹¹²
• Margin of CPPD/QPPD consistency: the analysis further suggests that CPPD and QPPD apply similar criteria overall. Thus, most persons (about 80%) who would have been eligible for CPPD, possess characteristics predicting eligibility for QPPD and vice versa. (The 20% difference was seen by the evaluators as relating to the QPPD priority for older workers, as opposed to CPPD's relatively higher incidence of grants to younger workers.)
• Program generosity: the analysis suggested that overall, CPPD was only slightly more generous than QPPD. The analysis suggested that more CPPD beneficiaries are similar to persons with disabilities who work full-time than are QPPD beneficiaries. Overall, about 10% of CPPD beneficiaries were thus identified as likely being able to work full-time as compared to about 7% of QPPD beneficiaries. CPPD beneficiaries were, however, in general, more likely to be severely disabled than QPPD beneficiaries.

3.5.2 Economic Grants, A Temporal Comparison

One additional test was conducted to consider the extent to which administration, or lenient adjudication, appears to be a factor resulting in "economic grants" and thus a factor in the 1991-94 incidence run-up.

Although recognizing that CPPD eligibility involves a complex of considerations regarding disability, duration and capacity to work, this analysis focused on the single indicator available at the time of the evaluation, namely a measure of the extent of disability. This analysis, while indicative only, addresses somewhat the important public perceptions of how CPPD operates.

To conduct this analysis, the 1995 Survey of CPPD Beneficiaries was examined as regards the number of activity limitation responses given by CPPD beneficiaries, and compared (using the same indicators) to CPPD beneficiaries who responded to the 1991 HALS. The analysis compared the exact same survey indicators, combined into one indicator - mean number of disability responses¹¹³ - given by CPPD respondents to the 1991 HALS and 1995 CPPD Beneficiaries surveys.¹¹⁴

The 1995 CPPD beneficiaries survey indicated a higher level of activity limitations for survey respondents, with a higher mean number of disability responses, as compared to 1991 respondents.¹¹⁵

Discussion: There is no conclusive evidence that economic grants are a substantial factor for CPPD's caseload increase in the period 1991-94. The HALS analysis, however, suggests that both of these programs (CPPD and QPPD) provide pensions to a noticeable minority of individuals who could potentially work (see Section 3.8.2) - judging from the behaviour of their employed peers with generally similar characteristics and comparable levels of activity limitations.¹¹⁶

This finding and previous ones point to the need for a fundamental determination on the extent to which and how, the CPPD process may or may not in fact encourage economic grants. The suggested remedy for this need is a major, ongoing case/file review, as discussed in Section 4.9 below.

3.5.3 Some International Comparisons

It is likely that CPPD adjudication system could benefit from the experience of QPPD and of other countries in terms of developing adjudication tools and standards to assist decision-makers. Serious efforts have been made in the United Kingdom, the Netherlands, the United States, and Quebec to develop very detailed adjudication guidelines. The CPPD, on the other hand, has been criticized on grounds that it needs clearer, more objective adjudication standards and guidelines.¹¹⁷

Recently, new adjudication guidelines, dated September 1995, were introduced at CPPD in order to address these concerns. The new guidelines clarify the interpretation of the legislation, the jurisprudence of appeals tribunals and adjudication policies. These new guidelines are expected to ensure a more uniform application of adjudication policies and to provide better guidelines with respect to important elements of the adjudication process such as the primacy of medical criteria.

The new CPPD guidelines appear to be less comprehensive, however, than adjudication tools developed in other countries, such as: the Dutch computerized system for matching functional capacities with job demands; the new British "all work test" which assesses the functional limitations of claimants in 15 work-related activities; the detailed adjudication guide currently being developed for QPPD adjudication officers; or the Baseline of Occupational Demands currently being developed by the American Social Security Administration.

The historically lower incidence of reversals of initial decisions at QPPD (especially at the second level of appeals) seems to suggest a better application of adjudication norms and guidelines and a more consistent decision-making process. At QPPD, extensive use is made of internationally recognized standards to assess disabling conditions, and the exclusion of vocational criteria from the adjudication process makes it easier to achieve consistent decision-making.

Some stakeholders suggested that too much weight was placed on beneficiaries' own doctor's reports. That is, many physicians, it was felt, tended to respond to CPPD evaluations to satisfy their patients rather than present the medical evidence in a fully unbiased way. This concern of stakeholders was not an objection to medical evidence per se, but rather an objection to the adjudication of claims relying solely on the report of the patient's physician, rather than an independent medical assessment, even though the instructions to the physicians focussed on objective medical standards as opposed to physicians' opinions regarding disability.

At another level, a particularly important difference in adjudication procedures at QPPD appears to be the more frequent use of independent medical examinations at the administrative review level to gather evidence on the medical condition of claimants. This aspect has been suggested as one of the major causes of a lower reversals rate of initial level decisions at QPPD. This practice appears to be equitable and to contribute to a more consistent adjudication process.

Administrative practices related to the adjudication of CPPD applications have lagged and cumulative effects. CPPD applications and decisions on CPPD applications only occur some time after legislative, economic and other causes of the changes in demand for these benefits. The evaluation was not able to fully explore these lag effects.

3.5.4 Discussion of Adjudication

Overall, these indicative findings comparing CPPD and QPPD clients, and 1991 and 1995 CPPD beneficiary data, suggest that the adjudication process is generally objective and predictable, and that the process for CPPD is generally comparable to that for QPPD.¹¹⁸

Areas for improvements to CPPD, with regard to claim adjudication and the appeal structure, have been addressed to some extent through the Income Security Programs Redesign Project (Redesign). The Redesign project will provide a much needed introduction of new computerized resources (for a program which has operated for many years with a mix of 1960s "paper file" systems, and 1970s vintage computer systems), and reorganization to improve access for clients. The Redesign project is in mid-course, and scheduled for completion in 1997.¹¹⁹

Other steps have been taken to improve adjudication, such as the September, 1995 implementation of new adjudication guidelines clarifying the primacy of medical determination criteria in adjudication. There is a need to develop a framework in which the new adjudication guidelines will be fully articulated, with benchmark parameters for determining disability to support adjudicators' decisions. Better processes have been implemented to ensure complete case development at the initial level or, alternatively, at the administrative review (first level of appeal) level. Case development and the collection of evidence is primary in reducing the number of reversals of initial decisions at the appeal level.

Need for Better Program Information: This research is only indicative, however, because of the limited HALS sample and the preliminary nature of the 1995 Beneficiaries Survey data and other differences between the two surveys, and many questions remain to be clarified. This inability to draw strong conclusions stems in great part from the lack of certain types of information about the ongoing adjudication process, which is needed for purposes of quality assurance, and for the strategic management of CPPD.

CPPD has, historically, maintained a variety of reviews of the program's administrative decision-making, i.e., on process and results, but these have not provided ongoing information on the quality of the adjudication processes, and the role of key factors in adjudication. For example, the recent random review, while no doubt extremely useful to managers in its treatment of operational aspects of adjudication (treatment of eligibility decisions, filing of permissions, completion of forms etc.), did not (and no doubt was not intended to) provide information which would be useful to the strategic and policy management of the program¹²⁰ (data of the sort which would be required to measure the extent of economic grants, provide an independent view of decision-making, explicitly measure extent of applicant disability and employability, and provide quality assurance on adjudication).

Such a flow of strategic information is much needed by the program, to allow policy makers to deal with questions of program appropriateness and cost. As well, it is exactly this type of data which is needed for ongoing evaluation of the program (for example, to start developing the background operational data for the next round evaluation, which should be conducted about four years from now).

Thus, to examine this issue in more detail, considering the full range of CPPD adjudication criteria, a specific, ongoing case-file monitoring program may be desirable and useful in light of the recent introduction of more rigorous adjudication standards (see Section 4.9). Such an exercise would also shed light on the role of economic grants in increasing caseloads, specifically for the pre-retirement age group.

The available evidence suggests that the CPPD program administration has been relatively constant in recent years in its adjudication of applications and claims for CPPD benefits (in spite of labouring with limited resources and an antiquated system based largely on paper files). The background research for the evaluation has pointed to economic fluctuations, legislative changes and population demographics as being the true drivers of CPPD caseloads.

3.6 Appeals

Some specific evaluation questions posed in this section are: Are criteria applied in a consistent and equitable manner by medical practitioners and appeal boards?

Claimants whose applications for CPPD benefits have been denied can appeal the decision. Until recently, there were four avenues of appeal: appeal to the minister (the "81" level of appeal); to the Review Tribunal ("82"); the Pension Appeals Board ("83"); and to the Federal Court of Canada. For the last two stages, the claimant has to request and receive leave to appeal.

Recently, the "81" level of appeal has been streamlined into an administrative review whereby claimants provide additional information to support their claims for benefits.

This change is based on the fact that, in the past, appeals frequently involved the submission of new or updated evidence by the applicant. In the first level of appeal, this new evidence often prompted a decision to award benefits before a further review process got underway. In general, "it is safe to say that many, and perhaps most, of these appeals-stage grants [...] are the result of additional information being submitted by the applicant [...]."¹²¹

The appeal stages are progressively more formal, and it has been suggested that they tended to involve a shift in the burden of proof from the claimant to the bureaucracy which is maintaining the denial. At the "81" level of appeal, or reconsideration, the decision is made again by a single adjudicator, as is the case with the initial claims adjudication.

At higher levels, progressively more individuals become involved in evaluating an appeal. At the Review Tribunal and Pension Appeals Board (PAB) levels, a medical adjudicator prepares a case for submission to the Tribunal or Board. Tribunals consist of three paid professionals trained for this job, including one qualified medical practitioner. Pension Appeal Boards (PABs) consist of three judges or retired judges; at this level, the applicant is permitted legal representation. Very few cases are appealed to the court system. In addition, a decision to award benefits made by the Review Tribunal can be appealed to the Pension Appeals Board by the Minister, although this is very rare.¹²²

Compared to most other countries examined, CPPD seems to be a program with a historically very high rate of appeals (although Canadian and American experiences are somewhat similar). There is a perception, particularly among stakeholders, that CPPD has a "loose" appeal structure, where persistence through several appeal levels usually pays off in the granting of benefits. A high proportion of appeals is not only costly for the system, but also suggests that shortcomings exist in the overall adjudication system. Also, the possibility of subjecting claimants to long and complicated appeal procedures to obtain benefits to which they are entitled is a serious concern.

This situation with CPPD appeals may stem partly from the fact that the responsibility for case development lies primarily with the claimant, who may fail initially to gather all the evidence necessary to produce a fair adjudication and thus, case development is minimal. When compared with the practice of other programs described in the international comparison, the CPPD procedure has a number of potential pitfalls, with decisions being based primarily on a medical assessment which the applicant has to arrange for him/herself, and on any other information that the applicant sees fit to include in the application.

In contrast, the more common practice internationally is to assign case decisions to adjudication officers or teams of experts, who are responsible for ensuring that cases are fully developed before being submitted as claims and before adjudication decisions are made. This results in a more thorough and complete initial adjudication process, whereby decisions are based on a fully developed case, which consequently reduces the number of appeals and their associated costs. As with initial adjudication, it also appears that the appeals process might be made more rigorous by the greater use of independent medical examiners. These represent areas for potential improvement of CPPD.

3.7 Program Abuse and Reassessments

The evaluation considered the question: What are the levels, if any, of misuse and abuse of disability benefits (e.g. extent of receipt of benefits by those who are not entitled to them)?

3.7.1 Program Abuse

Issues of system and program abuse have been prominent in discussions of CPPD, particularly in light of the rapidly increasing claim rate for CPPD benefits. While a few stakeholders who were interviewed for this evaluation suggested that there was both some over-use and abuse of CPPD, the majority indicated that there was little concrete evidence to suggest widespread abuse. This view is consistent with past administrative audits which have, as a rule, found that CPPD fraud is rare.

Where abuse was suggested by these audits, it was generally indicated to be unintentional misuse or questionable legal practices. Such reasons as acceptance of overpayments, or lack of awareness of the responsibility to report payment of duplicate benefits, or failure to report a return to work, were suggested as the principal reasons for most such problems.

Some LTDI providers, however, alleged that there are abuses of CPPD in that inappropriate claims are being accepted, which coupled with a lack of reassessments, results in a number of inappropriate claimants continuing to draw benefits. A requirement for regular reassessments, in combination with a well-developed system of information-sharing and co-ordination would, according to LTDI key informants, provide a solution to this problem.

Given the perception that some abuse of the CPPD system does exist, coupled with the desire of governments to administer programs in a cost effective and efficient manner, and given the recent suggestion that as many as 3% of cases may have been granted unwarranted benefits,¹²³ it would be appropriate for systematic reassessments to become an integral and well-established part of the CPPD program. Recent initiatives in reassessment provide useful directions as noted below.

3.7.2 Reassessments

The evaluation considered the questions: How effective are reassessment procedures for determining whether persons receiving disability benefits should continue to receive them? Should the eligibility requirements of disability claimants be monitored more rigorously? What steps are needed to verify continued disability of recipients of CPP Disability pensions on a cost-effective basis? Are resource constraints preventing efficient and cost-effective monitoring activities? Are sufficient resources being allocated to re-assessments, rehabilitation and fraud investigation?

CPPD beneficiaries' files were not reassessed in a systematic manner in the period leading up to 1993, the assumption being that, according to the CPP legislation, beneficiaries have disabilities which were severe and prolonged and would not be able to return to the workforce. There was also a re-assignment of resources to adjudication to reduce inventories of applications awaiting processing. Only occasional audits were conducted, which revealed various administrative errors and, very rarely, instances of fraud. Since 1993, however, reassessments have been intensified under a special project, which has focused on beneficiaries who have a high probability of being gainfully employed. Under this innovative and aggressive project, reassessments have been specifically targeted at those beneficiaries most likely to have regained capacity, which has resulted in a relatively high level of benefit terminations.

Between May 1993 and November 1995, 12,055 cases were reviewed, of which 4,911 resulted in the termination of benefits with an estimated annual saving of $45.2 million in benefits and the identification of a further $14.7 million in overpayments to be recovered.¹²⁴

Concerns have been voiced that a lack of resources has limited the reassessment component of CPPD administration. A report prepared by the CPPD reassessment staff has argued for a more comprehensive and extensive program of reassessments, projecting a potential for 4,000 terminations and savings of $40 million annually, and even higher cumulative impacts in the future.

3.7.3 Conclusions on Abuse and Reassessments

While abuse per se and error appears to be modest, the reassessment feature of the CPPD program seems to be an important one, worthy of significant expansion, as will be noted in Section 4.

3.8 Rehabilitation And Work

3.8.1 CPPD and Rehabilitation

Some specific evaluation questions addressed in this section are: Do current provisions for disability benefits support rehabilitation and return to the labour force? What are the rehabilitation efforts under CPP? What is the relationship between the rehabilitation efforts under CPP and those under complementary Provincial and Federal programs (e.g. VRDP)?

Just as CPPD has not traditionally focused on reassessment, neither has rehabilitation been a focus of CPPD. Rehabilitation efforts have historically been lower at CPPD than in most other countries, partly because of the assumption that CPPD beneficiaries may well be permanently disabled. Rehabilitation is a primary concern for other disability insurance providers such as WCBs and LTDIs. However, some innovative efforts in this area have begun within CPPD in the past few years. Also, there would be a need to harmonize any enhanced CPP efforts in the area of rehabilitation with complementary provincial programs. Of interest also would be the question of an equitable sharing of the rehabilitation costs for beneficiaries of the CPPD and other programs.

The National Vocational Rehabilitation Project: In response to the need for a rehabilitation component of CPPD, the CPP National Vocational Rehabilitation Project (NVRP) was initiated in 1993. Its objectives were:¹²⁵

• to return CPPD clients who successfully complete vocational rehabilitation to substantially gainful employment;
• to return these clients to their pre-disability job with their pre-disability employer; modified work with their previous employer; a comparable position with a different employer; or other appropriate work with any employer; and
• to promote independent living that: improves quality of life and self-esteem; allows for a contribution to the economy; and alleviates the need for CPPD benefits.

Participation in the NVRP is voluntary¹²⁶, and CPPD pension beneficiaries are not currently compelled to participate, even if identified as clients with significant potential for returning to work following rehabilitation. Clients continue to receive CPPD benefits while they are participating in the project.

The current selection criteria for participation in the NVRP are as follows:

• the beneficiary must be under the age of 50;
• the beneficiary must not have been granted benefits under the provision of Bill C-57 permitting retractive claims;
• there must be a reasonable level of education and transferable skills;¹²⁷ and
• the beneficiary must be medically stable and not be terminally ill.

The NVRP's services are provided through a network of 11 rehabilitation contractors (consultants) located in most provinces. The private rehabilitation consultants are responsible for case managing the rehabilitation process, including the assessment of the potential for rehabilitation, the development of a rehabilitation plan in collaboration with the client; and supporting clients through their rehabilitation plan. NVRP managers (CPPD staff) are responsible for client selection, the approval of rehabilitation plans, and the supervision of the work of rehabilitation contractors. The rehabilitation services are paid for by NVRP. Among other things these services may include high school upgrading, short-term on the job training, formal education programs, physical conditioning and the provision of assistive devices. The NVRP undertook some communication efforts such as the publication of a project brochure, and liaison with non-government organizations to publicize its services.

The NVRP, with a budget of $6 million, has provided rehabilitation services to about 623 CPPD beneficiaries who were identified as having a good potential for return to work. The NVRP was originally a three year project (1994-96), but was extended through to the end of fiscal 1996-97.

The Results of the NVRP Evaluation:¹²⁸ The findings of an evaluation of the NVRP conducted in the summer of 1996 were as follows:

• NVRP was a success in demonstrating the practicality, as well as the cost-effectiveness and societal benefits of a rehabilitation function within CPPD;
• As shown in Exhibit 19, NVRP has resulted in a relatively high success rate of 41%, 160 of 391 candidates who have completed the project or have since withdrawn from the project. With most of the remaining 227 participants still participating in the project, the overall success rate is still to be determined.
• The rehabilitation of these 160 clients is expected to yield cost-savings of $4.6 million after three years, $15 million after ten years, and about $30 million by the time the rehabilitated CPPD clients reach the age of retirement, assuming they do not return to the CPPD disability rolls¹²⁹; and
• A client survey in support of the NVRP evaluation revealed that about 60% of project participants who successfully completed their rehabilitation found employment, with about two-thirds of those employed full-time at the time of the survey (May 1996)¹³⁰.

Exhibit 19 CPPD National Vocational Rehabilitation Project Numbers and Proportions of Successful, Unsuccessful and In-Progress Cases

Status	1993/94	1994/95	1995/96*	Total
Completed (Successful)	16	41	103	160
File Closed (Unsuccessful)	41	70	120	231
In-Progress	122	128	227	227
Under Appeal	0	3	2	5
Total	179	242	452	623

• The evaluation supports the rationale for a permanent rehabilitation function as part of the CPPD. Significant cost-savings are possible, even with the rehabilitation of a small portion of CPPD beneficiaries. Rehabilitation is also an effective caseload management mechanism used by most other providers of disability insurance.
  The existence of certain services from the Employment Insurance program, specifically Human Resources Investment Fund (HRIF) re-employment assistance, would not eliminate the need for a rehabilitation function as part of the CPPD. However, linkages between the two would enhance the opportunities for CPPD rehabilitation clients who are attempting to return to work once they have completed their rehabilitation program. In addition to the HRIF services, EI services available to all Canadians could assist CPPD beneficiaries to overcome employment barriers, through job counselling, placement services and access to labour market information. Whether the federal government continues to deliver these services, or they are transferred to the provinces, these services could provide a bridge to employment for CPPD rehabilitation clients, as well as others.
• In spite of substantial project success, less than one-third of clients reported having been successful in achieving their personal rehabilitation goals. This suggests that many participants viewed the NVRP as an opportunity to pursue career goals (an occupation of their choice), and did not fully understand that the goal of the NVRP was the provision of labour market entry-level skills to NVRP participants at reasonable cost ;
• The NVRP evaluation also found systemic barriers to the effective and efficient delivery of NVRP services:
  • the perception that rehabilitation is not clearly presented to beneficiaries as an option at the time of initial contact with program;
  • significant variations in the quality of services, and average costs per client;
  • concerns by some rehabilitation consultants of a lack of clear policies regarding the types of services covered, and delays in approving rehabilitation plans;
  • rehabilitation-related information is not collected in a systematic manner by the program to facilitate the selection of candidates; and
  • data limitations prevent an adequate estimate of the optimal size of a rehabilitation function as part of the CPPD.
• The NVRP evaluation could not provide a clear answer to the question as to whether rehabilitation should be mandatory or voluntary; enforcing the regulation would contribute to equitable treatment of beneficiaries, but might have significant program cost implications, while the threat of non-compliance would be inconsistent with sound vocational principles to encourage the most highly motivated to participate;
• Any permanent rehabilitation function as part of the CPPD should be accompanied by a number of changes designed to enshrine a "rehabilitation mission" as part of CPPD operations. This could include better communication to facilitate the rehabilitation processes and inform beneficiaries of the rehabilitation goals of the CPPD, beginning at the time of the initial application for the program. As well, significant improvements could be made in the area of client selection including:
  • more and better rehabilitation-related data collection, from the time of the initial application for CPPD benefits (e.g., the use of a Vocational Profile Questionnaire), and better follow-up processes to identify rehabilitation opportunities once the medical condition of beneficiaries have stabilized;
  • more cost-effective screening mechanism with basic criteria, which would rely on an improved information base for on-going reviews of rehabilitation potential of CPPD beneficiaries¹³¹;
  • stronger work incentives within the CPPD, and policies to encourage work-place experimentation (part-time work, longer job search periods¹³²);
  • cost-sharing agreements with other service providers (private insurance carriers, Workers Compensation Boards) are implemented for only a small portion of clients. There is a need for increased communication efforts to explain the NVRP to other service providers, and to negotiate more partnership agreements with them.

3.8.2 Rehabilitation "Market" in CPPD

A significant market exists for rehabilitation in CPPD. This is revealed by both stakeholders' opinions and by respondents to the 1995 Statistics Canada Survey of CPPD Beneficiaries. Such a "market" exists partly because of the age structure of the CPPD beneficiaries. CPPD benefits have been granted to an increasing number of younger claimants over the past few years (currently about 50% of CPPD beneficiaries are under the age of 55 and about half of these are under the age of 45). The potential for major cost savings cannot be underestimated, particularly as younger beneficiaries constitute a significant on-going financial obligation for CPPD, for benefits payable until normal retirement age.

Beneficiaries' Experiences and Perceptions: As expected, CPPD beneficiaries' experience with vocational rehabilitation has been limited, as evidenced in the 1995 Survey of CPPD Beneficiaries. Indeed, vocational rehabilitation is extremely uncommon for CPPD beneficiaries. Only 7.3% of CPPD beneficiaries surveyed in 1995 had ever participated in work-related training (vocational rehabilitation) since they started receiving CPPD benefits, a finding consistent with the historical absence of rehabilitation programs for CPPD.

According to respondents to the 1995 Survey of CPPD Beneficiaries, 69.3% reported being completely unable to do any work now or in the future. However, 30.7% reported that they were either unable to do any type of work now, but may potentially be able to work in the future, or were able to do some work, but may be limited in the kind or amount of work they can perform. This significant minority of CPPD beneficiaries - three in ten - felt that they had some potential for return to work and regular employment.¹³³

Most of those who wanted to work indicated that they required vocational training to return to work (76.3%), while 23.7% of them did not believe that they required any training. Part-time work and full-time work were equally preferred. Among the sub-group of those CPPD beneficiaries who may be able to return to work, but who are not currently working, 88.2% said that they would like to work, provided that the work could accommodate their disability needs.

Overall, a surprisingly large percentage of CPPD beneficiaries expressed a desire to work, and a matching readiness to undertake rehabilitation and training to be able to do so. As well, related findings noted earlier suggest that there is an extensive potential for rehabilitation and return to work among CPPD beneficiaries. For example, as was indicated using the 1991 HALS data, many people with disabilities and activity limitations comparable to those reported by CPPD beneficiaries survey, support themselves and their families through employment (36% of persons with any activity limitations in Quebec, and 45.6% of persons with activity limitations in other provinces reported that they worked full-time in 1991). And about 10% of CPPD beneficiaries were estimated to be capable of full-time work overall.

Vocational rehabilitation was of interest to a substantial minority of CPPD beneficiaries who responded to the 1995 Statistics Canada Survey of Beneficiaries. About 25% of CPPD beneficiaries indicated an interest in having vocational rehabilitation, 38.7% indicated no interest at all and 36.3% reported being completely unable to take training.

These results may be explained by the fact that the historic CPPD model appears to have captured many beneficiaries who, once in receipt of benefits continue to collect a pension, even though their capacity may improve, or other factors (e.g. new technology) make a return to substantial gainful employment possible. Improving this aspect of the program, so that beneficiaries would not be "locked into" CPPD would appeal to many. Those favouring such changes would include those concerned with program costs, and also social advocates who argue that CPPD should facilitate the independence and development of individuals with disabilities. Thus, the considerable potential for significantly expanding rehabilitation efforts may provide important opportunities for CPPD program refinements and more efficient program delivery.

3.8.3 Issue of Timing in Rehabilitation

Unlike QPPD, the definition of disability under CPPD does not require that the impairment or disability be permanent. Rather, a "one year" test is applied to determine whether a claimant is entitled to receive disability pensions. Thus, claimants with a potential for recovery are not systematically excluded from the program, as seems to be the case under QPPD. Moreover, the fact that consideration has been given to factors such as the age, experience and education of claimants in the adjudication process suggests that many beneficiaries should have potential for rehabilitation, if properly re-trained.

The current CPPD rehabilitation process starts if at all, well after benefits have been granted rather than being part of an "up-front" decision-making process. This timing results in two inter-related problems: (1) the client will not likely demonstrate any rehabilitation potential if he thinks to do so might adversely affect his eligibility for a benefit; and (2) according to rehabilitation specialists, the optimum time to start vocational rehabilitation is immediately after the onset of disability. There is typically a considerable administrative lag between the date of onset of a disability and the receipt of CPPD benefits. With respect to this second point, these concerns were noted widely by stakeholders interviewed in the course of the evaluation.¹³⁴

Given the somewhat lengthy application, appeal and approval process associated with the granting of CPPD benefits, rehabilitation efforts are almost never started within the optimum time for the most successful outcomes for clients. However, it is well recognized that rehabilitation is most effective when it is started directly after the onset of disability. Thus, given the CPPD process, rehabilitation efforts, beginning after CPPD benefits have been granted, are often too late to provide the maximum benefit for disabled people.

This suggests that new program concepts and designs are needed to identify opportunities for early interventions for effective rehabilitation. This would also require new kinds of inter-program linkages to allow for the timely identification of candidates for rehabilitation; for example, with UI sickness benefits. But there is a need to determine what should be CPPD's responsibility in the area of rehabilitation, in light of provincial jurisdiction in this field as well as employers' responsibilities in the case of "on the job"-induced disabilities.

3.8.4 International Comparisons in Rehabilitation

Other countries offer rehabilitation programs in conjunction with their PDI programs, and broader income security programs for people with disabilities. These programs often have extensive sickness benefits which allow for up to one year of benefits before recipients are transferred to disability pensions. They also usually provide benefits for partial disability, either as part of the main disability benefits program or under a separate program. Thus, rehabilitation is often more closely linked to sickness benefits (such as in Sweden) or partial disability (such as the Netherlands) rather than the main disability benefits system.

CPPD also has less effective linkages with short-term sickness benefits than are found among other major trading partners. In Canada, up to 15 weeks of unemployment insurance (UI) benefits are available to a person who leaves a job because of illness or injury. Although the technical waiting period before eligibility for CPPD corresponds to this period so that, theoretically, UI sickness benefits could serve as a bridge to CPPD benefits, this period is quite short, especially given the stipulation that disability must be expected to last at least one year, and given the length of time that can be taken for CPPD adjudication and appeals.

Most other countries studied (with the exception of the U.S.) provide sickness benefits for one year and have established mechanisms to provide a smooth transition from sickness to disability benefits. Some countries (for example, Germany and Sweden) make use of the sickness benefits system to identify potential candidates for rehabilitation before they enter the disability rolls, something that is not currently done under CPPD.

3.8.5 Conclusions Regarding Rehabilitation

The potential gains from rehabilitation programs in Canada are unknown. They may be lower than in countries offering benefits for short-term or partial disability, but could be substantial, especially considering the low and declining average age of CPPD beneficiaries.

Nonetheless, CPPD may benefit from the experience of other countries such as Germany and Australia, particularly where rehabilitation efforts can result in outreach to individuals soon after the onset of disability. As will be noted below, this objective of appropriate timing for rehabilitation requires additional consideration of the way in which CPPD is linked to other programs such as WCBs and UI sickness, which are more likely to deal with individuals with disabilities shortly after the onset of disability.

3.9 Program Costs And Funding

Some specific evaluation questions addressed in this section are: What are the comparative costs of CPPD and LTDI coverage? Why have the expenditures on disability benefits increased significantly in recent years above those projected in the CPP Fourteenth Actuarial Report? What were the causes for the increase in CPPD benefit payments? Is this recent increase in CPP Disability payments a permanent or temporary trend?

3.9.1 Costs and Cost Increases, 1991-94

As noted earlier, a major concern with CPPD in the period 1991-94 was its rapidly increasing caseload, and its overall costs. Thus, nominal program costs rose from about $1.7 billion in 1990 to about $2.8 billion in 1995. Elsewhere it has been noted that these cost increases impact on the fiscal stability of the CPP fund, adding to the call for an increase in contribution rates to stabilize the funding of the plan.¹³⁵

These costs are typical of a disability cost spiral that has affected most developed nations. For example, nominal PDI costs for the United Kingdom increased by 279% from 1984 to 1994; Australian nominal costs increased by 246% in the same period; Swedish PDI costs increased 139% in 1983-92; New Zealand PDI costs were estimated to be up approximately 154% in the same time frame. In contrast modest increases were reported only for Germany (35% increase, 1985-94) and the Netherlands, which also reported small increases.¹³⁶

Indeed, even the LTDI sector, with its strong profit-orientation, and powerful tools of reassessment and rehabilitation, has faced significant cost increases in recent years. For example, between 1990 and 1993, LTDI group policies in Canada increased their coverage of the workforce (persons) by 3.4%, but the total dollar value of claims paid by LTDI companies increased by 33.9%, far in excess of the inflation rate over this period.¹³⁷

3.9.2 CPPD Cost Comparisons With LTDI

Some evaluation questions considered included: What are the comparative costs of CPPD and private insurance coverage? What are the implications of long-term CPPD for private insurance coverage? In the absence of CPPD, what would have been the likely effects on private long-term (LTD) insurance contribution rates needed to provide the same protection?

There are significant differences between CPPD and LTDI which make comparisons difficult. Some of these differences, which must be kept in mind, are:

• LTDI provides coverage only for those who are employed, while coverage under CPPD begins only after contributions have been made for at least 2 years and continues up to the age of 65 after leaving the labour force;
• LTDI usually provides benefits for two years, on an own occupation basis, and on "any occupation basis" after that, while CPPD provides benefits on an any occupation basis;
• LTDI has a waiting period of from 1 to 12 months, usually 4-6 months, during which benefits are not paid, even though the beneficiary has been deemed to be eligible, while CPPD pays benefits after a four-month waiting period from the date of disability. With LTDI, benefits in the waiting period are usually paid by an uninsured or insured sick leave, short-term disability, or salary continuation scheme, or by UI Sickness;
• LTDI usually provides benefits in the range of 40% to 80% of earnings, usually around 70%, depending upon whether the scheme is structured to provide income on a taxable or non-taxable basis, rather than on the CPPD flat-rate-plus-earnings related formula;
• LTDI benefits are usually reduced by the amount of CPPD benefits. Consequently, the actual LTDI benefit may be below the percentage indicated, because that percentage is inclusive of CPPD;
• LTDI almost never provides benefits for the children of beneficiaries, as CPPD does; and,
• as well, LTDI is very rarely indexed, unlike CPPD.

Because CPPD does not provide substantial income replacement for those with high pre-disability earnings, there will always be a significant market for LTDI. The market for LTDI will be further strengthened by its provision of coverage on an own occupation basis, and its ability to integrate with other short-term earnings replacement plans such as Unemployment Insurance sickness.

Because of the great variation from one LTDI plan to another in the level of coverage and the waiting period, as well as other provisions, it is not possible to give one figure to indicate the effect on LTDI rates of CPPD. LTDI premiums in the aggregate are undoubtedly much lower than they would otherwise be because CPPD is the first payer.¹³⁸ CPPD's full inflation indexing accentuates the difference because each year the increasing indexed CPPD benefit is subtracted from a typically fixed LTDI guarantee.

3.9.3 CPPD and RPPs

Disability coverage under Registered Pension Plans (RPPs) is not a significant source of disability income. Some defined benefit plans provide that, in the event of disability, the accrued pension benefits will be paid as a disability pension. They relate the amount of benefit to the length of service, and so usually provide very small pensions. (While it is theoretically possible to relate benefits to the total years of actual service, plus potential service to retirement, few plans do so.)

The usual disability benefit now provided in defined benefit plans is the continued accrual of pension benefits while disabled without the necessity of the employee making contributions. This is considered to have a very low cost because of the expected impaired life expectancy of the disabled beneficiary. In defined contribution plans such a benefit is usually provided by the purchase of LTDI in an amount sufficient to cover the employer and employee contributions to the RPP in addition to the LTDI income paid to the beneficiary. Thus, CPPD has had almost no effect on the disability benefits provided by RPPs.

3.9.4 Net Costs of CPPD in the Tax Transfer System

What proportion of these benefits is recovered through the tax system and/or lower complementary program costs?

The interaction effects of the CPPD and QPPD and complementary federal programs like Guaranteed Income Supplement (GIS), Spousal Allowance (SPA), and with the Federal-Provincial income tax system, have implications for the net cost to Federal-Provincial governments of the CPPD/QPPD. Some of these more important interaction effects are as follows:

• CPPD/QPPD benefit payments could reduce the Guaranteed Income Supplement (GIS), Spousal Allowance (SPA) and widowed SPA¹³⁹, and the potential benefits from the Child Tax Credit (CTC), the GST credit and Provincial tax credits;
• CPPD/QPPD benefits reduce the cost of provincial programs such as WCBs (in some provinces) and PSA payments;
• a proportion of CPPD/QPPD benefits is recovered through Federal-Provincial income taxes; and
• employee CPP/QPP contributions (part of which go to pay for CPPD/QPPD benefits) are eligible for the 17% Federal non-refundable marginal tax credit and corresponding Provincial tax credit in Quebec; employer contributions are generally treated as deductions for corporate tax calculations.

The net cost of the CPPD/QPPD is therefore the gross benefit minus all induced increases in income taxes and induced reductions in the cost of other programs.

The Simulation-Tabulation (SIMTAB) model¹⁴⁰ of Strategic Policy, HRDC, was used to estimate recoveries through higher federal-provincial income taxes and lower costs of other federal government programs, from the provision CPPD/QPPD benefits, as well as the impact on net government revenues. These simulations for the 1996 calendar year are based on current Federal and Provincial tax rates (Exhibit 20).

Exhibit 20 Effects of the CPPD/QPPD on Government General Revenues, 1996 (Millions of Dollars)

Programs	Returns to Federal Govt.	Returns to Provincial Govt.	Returns to Fed./Prov. Govts.
SPA/GIS	60	-	60
Income Taxes	355	253	608
Other Program	19	-	19
Total	434	253	687
Less CPPD/QPPD Credit Net Effect on Govt.	(248)	(191)	(439)
General Revenues	186	62	248

The estimates show $434 million of CPPD/QPPD benefits were returned to the federal government in higher tax revenues ($355 million), lower SPA/GIS expenditures ($60 million) and lower costs of other programs ($19 million). This would represent about 14% of the CPPD/QPPD expenditures in the fiscal year 1996 of a projected $2.7 billion for CPPD and $415 million for QPPD. Another $253 million was returned to provinces in higher income tax revenues, for a federal-provincial total of $687 million of lower federal program costs or federal-provincial tax recoveries (about 22% of CPPD/QPPD expenditures). On the other hand, CPPD/QPPD contributors would receive $439 million in government non-refundable tax credits¹⁴¹ in 1996 in respect of CPPD/QPPD portion of contributions. SIMTAB does not estimate the impacts relating to the employer portion of these contributions.

It should be noted that, in a number of ways noted below, these calculations underestimate net savings to governments and the private sector from the existence of CPPD/QPPD benefits.

Other Savings: These cost estimates do not take into account the induced reductions (savings) in PSA payments through benefit programs like Ontario's GAINS-D, and (where offsets are applicable) WCB disability benefit programs which would further increase the benefits to employers of CPPD/QPPD. Similarly, these calculations do not take into account the business tax costs attributable to CPPD/QPPD which would have to be deducted from these recoveries.

As well, the possibility that without the CPPD/QPPD, more would be spent through other vehicles (private LTDI benefits) for protection against disability, was not taken into account. Neither was the lower cost of private disability insurance settlements considered because of CPPD/QPPD offsetting practices of private insurance (the reduction in LTDI benefits in proportion to CPPD/QPPD benefits).

Notwithstanding these omissions in the calculation of the estimates, these findings nevertheless indicate that the contributory CPPD/QPPD provides a significant return to government general revenues through either higher tax revenues or lower complementary program costs. The gross cost of the program (here estimated at $2.8 billion for CPPD and $415 million for QPPD in 1996) must take into account what is eventually recovered or saved by the governments, $687 million in 1996.

An approximation of the net effect on government net general revenues is $248 million after the CPPD/QPPD tax credit is taken into account. Therefore, induced government tax recoveries and the reduced costs of other programs are estimated at 8% of CPPD/QPPD gross program expenditures in 1996.

3.9.5 Containment of Costs

Funding of CPP (and thus CPPD) has been a matter of controversy in recent years, but beyond the specific mandate of this evaluation.¹⁴² Containment of costs, through various approaches, should be a concern for CPPD, within the framework of social goals for the program. And indeed, contribution rates are expected to increase significantly in Canada over the next few years, as the Canada Pension Plan fund currently operates with a very low level of capitalization and liabilities are expected to rise considerably.

Some administrative measures for cost containment and better outcomes for clients, as outlined below, focus on the refinement of internal operations (improved adjudication), and particularly on the strengthening of some initiatives such as those in reassessment and rehabilitation.

Other potential measures for cost containment have to do with more fundamental aspects of program design. Also, the interpretation of the appropriate role of CPPD by some provincial partners in income security, and the overall structure of program linkages, are key elements in any overall restructuring of the CPPD program.

3.9.6 Discussion

As is noted below, the accomplishments of CPPD are extensive in extending income protection to hundreds of thousands of Canadians (reaching close to 300,000 disabled persons and their households in 1996). This makes CPPD an important program in Canada's overall income security system.

Yet the program also appears to require significant improvements in such areas as inter-program linkages, adjudication practices, reassessment and rehabilitation. All of these improvements have potential to reduce costs, and thus improve the funding picture for CPPD and the CPP as a whole. These concerns are addressed in detail in Section 4 below.

---

⁴⁸ See: A Labour Market Analysis of CPP Disability Claimants, Abt Associates, January 1996.

⁴⁹ Gail Fawcett of the Centre for International Statistics of the Canadian Centre For Social Development examined the overall data quality of the 1995 CPP Disability (CPPD) Beneficiaries survey, and the comparability of the severity index constructed for this survey with that of the 1991 Health and Activities Limitations Survey (HALS). The particular concerns noted regarding the 1995 CPPD Beneficiaries Survey are a low survey sample size, possible strong selection bias (that some sub-groups constituted a disproportionately large sub-set of all respondents as compared to the beneficiary population) coupled with a high non-response rate, (54% compared to 86% for HALS 1991), and that some variables for the 1995 CPPD survey are not adequate for analysis. Only those indicators are reported in this report which are deemed appropriate for analysis. See Gail Fawcett, Report #1, Evaluation of the Severity Index Constructed for the CPP Disability Beneficiaries Survey, 1995 Report #2, Overall Analysis of Data Quality of the CPP Disability Beneficiaries Survey, 1995, May 1996 (unpublished), and technical note #6 at the end of this report. Similar considerations are noted regarding the HALS data, and discussed later in this report. The sample of QPPD beneficiaries in HALS was proportionately low vis-à-vis that for CPPD Beneficiaries Survey. This should be kept in mind where these data are examined.

⁵⁰ CPP Advisory Board, Report of the Committee on Disability Issues, December, 1994, p. 4 & 7.

⁵¹ See: Q.18 - Q.25 in CPPD Beneficiaries Survey, 1995 and Technical Note #6.

⁵² See: Statistics Canada Survey of CPPD Beneficiaries, 1995. However, these are indicative findings as to the income shares of CPPD beneficiaries because of the limitations of this survey (fee footnote 28).

⁵³ It should be emphasized that exact overlap cannot be estimated from the survey data, as it reports income for the entire 1994 calendar year: e.g. a particular beneficiary may have received two or more income sources at different times, e.g. one for the first 6 months of 1994 and the other for the second 6 months, etc. Data from the 1995 CPP Disability Beneficiaries Survey, Canada Pension Plan Advisory Board, 1994 Report of the Committee on Disability Issues, HRDC, 1994, p.20.

⁵⁴ But these individuals will receive CPP retirement pensions, which they would not if no CPP contributions had been made , although there may be a partial 50% Guaranteed Income Supplement (GIS) offset and some provincial benefit offsets (e.g. the GAINS-A, a program in Ontario at that time).

⁵⁵ The Year's Maximum Pensionable Earnings is the ceiling for earnings replacement under the CPP. It is historically an amount equal to the average industrial wage.

⁵⁶ 1964 White Paper on the Canada Pension Plan, pp. 15-16.

⁵⁷ The Year's Maximum Pensionable Earnings (YMPE) is historically an amount approximately equal to the average industrial wage.

⁵⁸See Torjman (1988); CPP Advisory Board (1994), p .20-22.

⁵⁹ It does not include any ancillary benefits such as dependants' extra costs or severe disablement allowances. Where a combination of means-tested and insurance-based programs exist, only the program based on insurance principles was included.

⁶⁰ CPPD beneficiaries' opinions about aspects of the CPPD program other than earnings replacement were mixed: about half of the beneficiaries (49.3%) indicated that the program should be changed to serve better the needs of those who received benefits, and half of the beneficiaries (50.7%) suggested no changes were needed. All things considered, beneficiaries were most likely to suggest improvements to CPPD benefits by raising levels of payments (55.7%) and by covering costs of disability-related supports and services (16.7%). A smaller proportion of beneficiaries suggested other changes.

⁶¹ The Low Income Cut-Off Line is determined by Statistics Canada on a periodic basis.

⁶² In contrast, for Canada a whole, the incidence of households living below the Low Income Cut-off Line was 14.5% in 1993, Statistics Canada Cat. No. 13-207, Table 66.

⁶³ Eligibility for PDI programs is based on a set of medical and non-medical criteria. These are described in detail in Section 3.3.2.

⁶⁴ The "A one-third of the contributory period" rule also meant that before 1993, younger workers with disabilities needed fewer years of contributions to qualify for QPPD benefits than older workers under 60 years of age.

⁶⁵ The "2 of 3 years" rule and the "5 of 10 years" rule are often termed the "recency-of-work" tests, as they seek to ensure the recent labour force attachment of disability claimants.

⁶⁶ See Technical Note #3.

⁶⁷ However, the persisting gap in program coverage among females between the CPPD (84%) and the QPPD (80%) is more difficult to explain. This difference could be attributed to either demographic and sociological differences (including labour-force participation), or to a slight over-estimation of program coverage by CPPD among females. It should be noted that estimations of the eligible population for QPPD are "actuals" obtained from a data base which included the complete earning history of contributors, while CPPD's are simulations from the Office of the Superintendent of Financial Institutions.

⁶⁸ In the American system, up to four quarters of coverage can be accumulated every year depending on the level of earnings. Less stringent eligibility criteria apply for people under the age of 31.

⁶⁹ See: A Literature Review on Public Disability Insurance Programs, a working paper for the CPPD Evaluation, SPR Associates Inc., March 1996.

⁷⁰ Yelin, (1986) and (1989) provides a particularly lucid discussion of this perspective; see also Berkowitz (1987), and the econometrically based debate around definition as represented in Leonard (1986), Parsons (1980a), Haveman and Wolfe (1984), Maki (1993), etc.

⁷¹ See: "A Statistical Comparison of the CPPD and QPPD Programs Using the Health and Activities Limitation Survey of 1991", Technical Analysis #1. Experience of the QPP Disability Program, A Comparison Case for the CPP Disability Program, a working paper for the CPPD Evaluation, SPR Associates Inc., March 1996.

⁷² Where a CPPD recipient's earnings level exceeds 25% of the average industrial wage ($8,850 in 1996), very strong medical evidence must be present to support a determination that this individual continues to be eligible for CPPD (See Section 4.6.1, Policy Directive 04/95-CPP-03: Medical determination of Disability Under CPP, Income Security Programs, HRDC, September, 1995.)

⁷³ Canada Pension Plan Act, Revised Statutes of Canada, paragraph 42(2)(a). It should be noted, that the CPPD/QPPD definition of disability differs in one extremely important way from definitions applied by the WCBs. The WCBs generally allow for partial disability - for example to provide a 70% pension on the grounds that an individual is 70% disabled. This difference represents an important issue for improved integration of CPPD with WCB programs.

⁷⁴ See: Policy Directive 04/95-CPP-03, Income Security Programs, HRDC, September 30, 1995.

⁷⁵ It is estimated that the demographic effect of this pre-retirement emphasis is such that QPPD caseloads would be about 25% higher if QPPD grants were proportionately awarded to the same age groups as CPPD. This estimate is based on an analysis substituting CPPD termination rates for recovery, death and retirement, to simulate growth in QPPD caseloads for 1984-93.

⁷⁶ Commentaires et observations ( Étude comparative des dossiers RPC/RRQ du 27 mai au 7 juin 1996) par Roger Dorion et Claude Sarra-Bournet du Régime des rentes du Québec., and Report on the CPP-RRQ Comparative Study of Disability Files, and appendices) ISP Branch, HRDC, prepared by Moyra Lauzière, September 6, 1996 (draft).

⁷⁷ The detailed results of this comparative analysis is set out in Technical Note #8, and Report on the CPP-RRQ Comparative Study of the Disability Files, and appendices, September 6, 1996 (draft).

⁷⁸ Policy Directive 04/95-CPP-03: Medical Determination of Disability under CPP, Income Security Programs, HRDC, September, 1995. These emphasise the primacy of medical criteria.

⁷⁹ HRDC, Income Security Programs Branch, CPP Disability, Comparative Results from a Study of CPP and QPP Decisions, August 1996, (draft), Appendix to the Report on the CPP/RRQ Comparative Study of the Disability Files study.

⁸⁰ See Canada Pension Plan Disability Benefits Caseload Growth Analysis Report (draft), June, 1996 and Canada Pension Plan Advisory Board (1994) Report. The grants and applications are set out on the left hand scale of Exhibit 15, while grants as a percentage of applications are illustrated on the right hand scale. Grants also follow applications with a lag (footnote 53).

⁸¹ But it is noted that a ratio of grants to applications is only a proxy indicator, because processing isn't instantaneous. Depending on backlog sizes and appeals, an application submitted in one year might not be processed until the next year, and could even be processed in the year after that because of administrative lags. Further, a denial at one level may turn into a grant further downstream.

⁸² Income Security Programs Branch, HRDC, unpublished statistics, 1996.

⁸³ This effect may be greater than the amount of program overlap alone would suggest, since individuals being case-managed by a WCB are more likely to be provided with rehabilitation opportunities - a feature which should result in a higher incidence of recovery and return-to-work. QPPD is 'second payer' to the WCB in Quebec.

⁸⁴ The proposition that disability rates are lower in Quebec is also supported by the fact that fewer people claim the federal Disability Tax Credit in Quebec, proportionate to the tax filing population. See: HRDC Inventory of Income Security Programs in Canada, January 1993, p.269.

⁸⁵ Comparisons between Quebec and other provinces are based on different size samples in the 1991 Statistics Canada Health and Activities Limitation Survey. See Technical Note #4.

⁸⁶ Social Assistance recipients would tend to have lower average income from employment and thus would be entitled to much lower CPPD/QPPD benefits.

⁸⁷ See: International Comparison of Public Disability Benefit Programs, SPR Associates, Section 2.

⁸⁸ Some of the difference may be attributable to the relatively relaxed adjudication process in place in Great Britain, prior to the introduction of a new system in April 1995, but other factors may also be at play.

⁸⁹ In 1995, this rate was about 2.3% for CPPD, but the rate of increase has been projected to fall in 1996. See: CPPD Benefits: Caseload Growth Analysis, Income Security Programs Directorate, Planning and Strategic Studies, Forecasting and Trend Analysis, HRDC, June 1996.

⁹⁰ Because characteristics of programs vary to a great extent from one country to another, Canadian programs are less comprehensive than other programs in such aspects as benefit level, compensation of partial disability, etc.

⁹¹ See: A Statistical Empirical Analysis of CPP Disability Benefits, Dr. Fred Lazar, SPR Associates, March, 1996; Canada Pension Plan Disability Insurance Benefits and Labour Supply and Well Being of Older Workers, Jonathan Gruber, Massachussets Institute of Technology and the National Bureau of Economic Research, August, 1996.

⁹² The impact of the aging 'baby boomers' as they enter a period of life more vulnerable to becoming disabled was also identified in a recent US study, Report on Rising Cost of Social Security Disability Insurance Benefits, Social Security Administration, February, 1996. In Canada, most of this effect which has been anticipated since the inception of the program is yet to come, according to the Chief Actuary, Office of the Superintendent of Financial Institutions.

⁹³ The proportion of CPPD female beneficiaries rose from 30% in 1981 to 42% in 1995, Statistics Related to Income Security Program, HRDC, March 1996, unpublished.

⁹⁴ Numerous authors have attempted econometric analysis of the economic aspects of disability insurance, giving rise to some controversy in the literature on disability. This debate is summarized in: Literature Review on Public Disability Insurance Programs, SPR Associates, March 1996, Section 5: "PDI and Labour Supply".

⁹⁵ Leonard (1986), pp. 66-67. This article provides a good summary of the evidence and arguments which support the existing argument on labour supply effects of disability insurance.

⁹⁶ See especially: Bound (1989); Breslaw and Stelcner (1987); Haveman (1985); Haveman, de Jong, and Wolfe (1991), Haveman and Wolfe (1984a, 1984b), Stern (1989), and Yelin (1986, 1989).

⁹⁷ Haveman and Wolfe (1984b), p .64.

⁹⁸ See: A Literature Review on Public Disability Insurance Programs, a working paper for the CPPD Evaluation, SPR Associates Inc., January 1996.

⁹⁹ See, A Statistical Analysis of CPP Disability Benefits, Dr. Fred Lazar, SPR Associates, March 1996. These results should be considered as indicative; more extensive econometric analyses would be required to confirm these findings. However, the results are also consistent with those of the related analysis provided by Abt Associates which demonstrates that individuals with poorer employment prospects are more likely than those with better employment prospects to make the transition from the labour market to CPPD. Any further econometric analysis should investigate the 'lag effects' in administrative decision-making, e.g., the time it takes to reach decisions once applications are received, as well as the administrative cumulative 'lags' following other potential causes of a change in demand for CPPD, e.g., legislative changes, effects of economic cycles, etc.

¹⁰⁰ Canada Pension Plan Disability Insurance Benefits and Labour Supply and Well-Being of Older Workers, Jonathan Gruber, MIT and NBER, August, 1996. Treating the 1987 rise in the flat rate portion of CPP disability (from $91.06 to $242.95) as a natural experiment, this study examined the labour supply response using a difference-in-difference framework. Persons living in the CPP provinces were considered to be the treatment group, while persons living in Quebec were considered to be the control group because QPP disability benefits were unchanged in 1987. The difference-in-difference analysis was incorporated into regression equations to control for factors such as age, education, marital status, spouse's age and number of children. The analysis also controlled for the potential replacement rate by calculating potential CPP disability benefits based on imputed earnings histories for males in the 45 to 59 age group.

¹⁰¹ The study by Gruber, 1996, planned to examine whether the health status of persons responding to the 1987 rise in CPP disability benefits was significantly different from the health status of CPP disability recipients before 1987. However, the available sample size was insufficient to undertake this analysis.

¹⁰² See, for example: Bound, 1989.

¹⁰³ See: A Statistical Empirical Analysis of CPP Disability Benefits, Dr. Fred Lazar, SPR Associates, March 1996.

¹⁰⁴ CPP Disability Incidence Study: Main Report, HRDC, 1995, pp. 73-97.

¹⁰⁵ See CPP Disability Incidence Study; the study identifies qualitatively a number of causes which might account for some unjustified awards in the past (before the revision of the medical adjudication standards in 1995) among them: (i) use of non-medical criteria, e.g., education levels attained; (ii) need to reduce a growing inventory of unprocessed applications, (iii) a 1988 administrative decision to recommend benefits to persons 55 years of age and over, if they were not able to carry on their usual occupation or some closely related work; (iv) the right of applicants to decline rehabilitation which might have enabled them to undertake gainful employment.

¹⁰⁶ Ibid, the CPP Disability Incidence Study is careful to note a number of limitations relative to this comparison, among these: (i) the HALS activity limitations are not the basis for awarding CPPD benefits and the CPPD definition of disability is more stringent than the HALS 'severe limitation' definition according to subject matter experts, (ii) it is a point in time analysis and is not longitudinal, - hence it cannot provide an explanation of the caseload growth over time. The HALS, CPP and Census data are for only 1991 and reflects only part of the most recent growth trend to 1994, which this study examines.

¹⁰⁷ See Grant Schellenberg, The Road To Retirement, Centre for International Statistics, Canadian Council on Social Development, Ottawa, 1995, pp. 89-91. While the author indicates that it remains to be seen whether CPPD is being used as financial bridge to retirement, the dramatic rise in recipients (and applicants) in the just before retirement age category (ages 54-65) for such pensions, suggest that this may be the case. This study also notes that there is possibly a similar use of provincial social assistance (PSA) and Unemployment Insurance benefits for the same reason, given the high rates and long duration of unemployment among older workers.

¹⁰⁸ By economic grants is meant the award of disability pensions to mildly or moderately disabled persons for economic reasons such as increased unemployment. Pre-retirement grants would refer to that category of economic grants directed to those mildly disabled persons in the 60-64 years of age bracket approaching retirement.

¹⁰⁹ This conclusion was based on a review of a random sample of 523 recent grants which found that about 3% of grants might have been prevented through a more careful application of the adjudication guidelines. See CPP Disability Incidence Study: Main Report, p.106.

¹¹⁰ See: CPP Random Review, Disability Operations Division, HRDC, January 5, 1996.

¹¹¹ Discriminant analysis is a multi-variate technique for the classification of qualitative data or prediction of membership in a class or group. The underlying mathematical model is similar to multiple regression. See: Statistical Package for the Social Sciences Professional Statistics, Release 6.0, 1993.

¹¹² It should be noted, of course, that actual eligibility for CPPD and QPPD is not based on the same factors as those used to predict eligibility in the above statistical analysis. See: "A Statistical Comparison of the CPP and QPP Disability Programs Using the 1991 Statistics Canada, Health and Activities Limitation Survey," Technical Analysis #1 in Experience of the QPP Disability Program, a Comparison Case for the CPP Disability Program, A Working Paper for the CPPD Evaluation, March 1996. Also see technical note #4 regarding limitations and reliability.

¹¹³ Where a disability response is an indication of a disability for the individual, or having a greater degree of impairing effect. See Technical Note #2 for details.

¹¹⁴ However, no attempt was made to weight the individual limitations as regards their relevance to employability, or to gauge the severity of the limitations. Nor was any adjustment made for the possibility that the 1995 CPPD beneficiaries might feel motivated to report limitations to support their receipt of benefits, while HALS respondents, not being involved in a program-oriented survey, would not be subject to the same degree of 'motivation'. This is further discussed in Technical Note #2 and #6 at the end of this report.

¹¹⁵ While CPPD recipients evidenced a mean number of 4.87 disability responses in 1991, disability responses increased to 5.97 in 1995. This analysis suggests that the extent of disabilities reported for CPPD beneficiaries was not lower, but higher in 1995 than in 1991. The two surveys, conducted somewhat differently, experienced somewhat different types of selection bias among those agreeing to be surveyed, and an over-representation of those with more severe disabilities for the 1995 survey than for the 1991 survey (See Technical Note #6). Neither of the samples has been demonstrated to be an adequately representative sample of CPPD beneficiaries, especially with respect to sub-group coverage by degree of disability (mild, moderate, severe).

¹¹⁶ It is possible that those persons with similar disabilities to CPPD/QPPD recipients have higher levels of education and experience than CPPD/QPPD recipients. The reader is also reminded that the survey definitions of disability differ from those for program eligibility.

¹¹⁷ See administrative audits conducted by the Internal Audit Bureau: Disability Assessment Division and Appeal Operations (October, 1987); CPP Disability Program: Follow-up Audit (November, 1989); Income Security Audit (draft, February, 1995) and the 1993 Report of the Auditor General.

¹¹⁸ As noted in a companion report to this evaluation, however, important differences exist in adjudication criteria between the two programs, especially in the areas of the use of vocational criteria and the definition of "A prolonged" disability. See: Experience of the QPP Disability Program, a Comparison Case for the CPP Disability Program, SPR Associates Inc., March 1996.

¹¹⁹ Not all Redesign changes are necessarily positive. The move to regional services, for example, while positive for client access, is expected to produce challenges for the program in the area of maintaining equivalent standards of adjudication between regions. In particular, there is concern that inconsistent standards of adjudication may evolve across regions.

¹²⁰ The 1995 CPP Random Review, for example, provided a positive assessment of consistency of administrative procedures for 316 benefits paid out in 1994, but no data on the role of socio-economic factors in adjudication, or the extent of disability, or any indicators of employability of applicants. See CPP Random Review, Disability Operations Division, HRDC, January 5, 1996. Previous random reviews were also carried out in past years (1986, 1989, 1992).

¹²¹ CPP Disability Incidence Study, Main Report, HRDC, 1995, p. 47.

¹²² CPP Disability Incidence Study Main Report, HRDC, 1995, pp. 14-19.

¹²³ See, CPP Disability Incidence Study, HRDC, 1995. The issue of potential non-applicants who might prove to be eligible for benefits, were they to apply, was not examined.

¹²⁴ No information was provided on appeals that might have been successful against such terminations.

¹²⁵ See: Presentation of the Canada Pension Plan, National Vocational Rehabilitation Project, Income Security Programs Branch, HRDC, 1993 and program literature.

¹²⁶ Although participation in the NVRP is voluntary, vocational rehabilitation is supported by CPPD regulations which authorize the program to require beneficiaries to make reasonable efforts to return to the work force. See Part V Regulations, Sections 69(2) and 70(1).

¹²⁷ Defined as (i) literacy in one of English, French, or another language provided there is a nearby ethnic community providing employment opportunities in that language; (ii) cognitive ability to learn new materials. See CPPD Rehabilitation Project, definitions p.5.

¹²⁸ The detailed findings of this evaluation are set out in An Evaluation of the National Vocational Rehabilitation Project (Draft Preliminary Report), SPR Associates Inc., August 15, 1996.

¹²⁹ These number represent estimates of the dollar value of the benefits saved and should very quickly exceed the $6 million cost of the project. Moreover, these estimates do not include other forms of savings, such as increased CPP contributions, other revenues such as federal/provincial tax receipts, or reduced dependence on other income support programs.

¹³⁰ However, the NVRP has not established any system to measure longer-term employment outcomes of the project.

¹³¹ Such an expert system has already been tested and is operational within the CPPD administration, but it could use better information about the rehabilitation potential of CPPD beneficiaries.

¹³² At the present time CPPD benefits are only extended for a job search period of three months after completion of the NVRP, compared to nine months in the US system. Also, no partial benefits are granted for a minimal period to assure some minimal level of earnings, which the US system also provides. See Social Security Administration, Report on Rising Cost of Social Security Disability Insurance Benefits, February, 1996.

¹³³ See: An Examination of the 1995 Statistics Canada Survey of Beneficiaries, a working paper for the CPPD evaluation, SPR Associates Inc., March 1996.

¹³⁴ See: Report on Stakeholders' Views, a working paper for the CPPD evaluation, SPR Associates Inc., January 1996.

¹³⁵ Fifteenth Actuarial Report on the CPP, Office of the Superintendent of Financial Institutions of Canada, 1995 (page 3, main findings).

¹³⁶ These differ from the caseload growth trends in Exhibit 17 because of different cost effects in different countries, and the different time periods involved. The Netherlands reported no change in PDI costs as a percentage of GNP in the period 1980-90.

¹³⁷ Based on data from Survey of Health Insurance Benefits in Canada, 1990-93; figures were: persons covered in 1990, 5,080,900, in 1993, 5,253,549; claims paid, $905 million in 1990; $1,212 million in 1993.

¹³⁸ It must be remembered that, in the absence of CPPD, the increase in premiums would be much greater for LTDI plans providing a low level of benefits (e.g. 40%), than it would be for those providing a higher level such as 75%, always assuming that CPPD is the first payer. This is because the CPPD benefits reduce LTDI costs to a proportionately greater extent for those who would receive lower LTDI benefits.

¹³⁹ There would also be a minimal GIS effect for senior couples with a younger disabled spouse in the 60-64 age range, and an even smaller OAS effect for high income disabled persons turning 65 during the year and subject to the OAS clawback.

¹⁴⁰ SIMTAB is a micro-simulation package for analyzing the effects of tax-transfer programs in terms of the program costs, program interaction effects (at the federal level) and their distributional impacts. It is a static model and only takes into account the effects of CPPD on provincial and federal income taxes (and credits), and the costs of other federal programs. It does not take into account the provincial social benefit programs targeted to the disabled. The Federal disability tax credit was imputed but not explicitly taken into account. The CPPD/ QPPD benefit was examined for only 1996. The CPPD child benefit was not included in these simulations because of limited data in the 1992 Survey of Consumer Finances (SCF) data base employed for these estimations. This SCF data base was supplemented by other data bases, e.g. Statistics Canada 1991 Census, Consumer Expenditure, Revenue Canada "Green Book" and program data. SPA showed the greatest federal program savings; the GIS savings arise via family income effects from filing of joint returns when one of the spouse is over 64.
The number of CPPD/QPPD recipients reflects a 1992 usage rate of disability benefits, since the data for these simulations were based on the 1992 Consumer Finance Survey. Consequently the aggregate number of 1996 recipients generated by the SIMTAB model only reflect population growth from 1992 onwards, and are somewhat under-estimated. The 1996 pay-as-you-go (PAYGO) rate as set out in the Canada Pension Plant Fifteenth Actuarial Report (February, 1995), page 9 was 7.85%, of which the disability-related expense accounts for 1.46%. This proportion (1.46/7.85 = .186) was applied to the 1996 actual contribution rate in order to reduce the rate for the purposes of estimating the credit in the income tax system related to contributions for CPP disability benefits only. Thus the estimated contribution rate relating to the disability benefit was 1.04% out of the total 5.6% rate for 1996, and equal to the disability-related expense component of the PAYGO rate in 1996.

¹⁴¹ A credit on the disability portion of the non-refundable credit for CPP/QPP contributions.

¹⁴² See: Slater, David, Reforming Canada's Retirement Income System, Canadian Business Economics, Fall 1995, pp. 47-58.

Previous

Contents

Next