Archived Content

Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please contact us to request a format other than those available.

Statistics Canada - Statistique Canada
Skip main navigation menuSkip secondary navigation menuHomeFrançaisContact UsHelpSearch the websiteCanada Site
The DailyCanadian StatisticsCommunity ProfilesProducts and servicesHome
CensusCanadian StatisticsCommunity ProfilesProducts and servicesOther links
Side menu bar Catalogue Number 75-001-XIE Table of contents Latest issue News from The Daily Latest data Survey information Back issues Feedback Studies Latest issue in PDF

October 2005
Vol. 6, no. 10

Perspectives on Labour and Income

Who's missing out on the GIS?
Preston Poon

The 2005 federal budget directed more spending to help low-income seniors by increasing Guaranteed Income Supplement (GIS) payments by roughly $2.7 billion (more than $400 per year for a single senior and almost $700 per couple for those receiving the maximum). The GIS was established in 1967 as an additional benefit to low-income seniors receiving Old Age Security (OAS). These programs, plus the maturation of the Canada and Quebec Pension Plans (C/QPP) and an increased use of private pension plans have reduced low income among seniors significantly over the past decade (Myles 2000). In 1980, roughly 1 in 5 seniors were in low income; by 2003 this had fallen to 1 in 15.1 According to the 2001 Census, seniors living in low income received two-thirds of their income from OAS and GIS benefits. An additional 20% came from C/QPP.

In order to receive the GIS, individuals must apply annually. For those already receiving the benefit, this can be done automatically by filing an income tax return.2 If one is not filed, a detailed income statement and application must be submitted to Social Development Canada (SDC). Individuals who lose eligibility because of an increase in income in the previous year (for example, an RRSP lump sum withdrawal) or a change in marital status are required to re-apply the following year if they wish to be re-considered. Eligible individuals unaware of this requirement to re-apply will not receive the benefit. Some may also miss out because they do not fully understand how eligibility is determined—for example, they are unaware that OAS should not be included in calculating their income. This is important since GIS recipients by definition have low income and since many provincial programs are linked to GIS receipt.

The importance of reaching eligible non-recipients was recognized in a House of Commons standing committee, which recommended that SDC address the situation (Canada 2001). In response, SDC increased outreach activities and simplified application forms. Also, in conjunction with the Canada Revenue Agency, SDC sought out eligible individuals using tax information. Early estimates of the numbers affected ranged from 220,000 to 380,000, while a more recent figure pegged the number at roughly 135,000 for 2002 (Ha 2003; Thompson 2005). Other than these estimates, little is known about these individuals.

Although income is an important indicator of a family's ability to maintain a given standard of living, equally important is wealth, especially since financial assets can easily be converted to cash in times of need (Morissette 2002). Low-income families with little wealth are potentially at risk should unexpected shocks such as sickness or divorce arise.

Two key issues are addressed here. First, to examine whether GIS families are more financially vulnerable than senior non-GIS families, the Survey of Financial Security (SFS) was used to compare the wealth of families receiving GIS and senior families not receiving GIS.3 Second, the question of who is not applying for GIS even though eligible is answered using the Survey of Labour and Income Dynamics (SLID). In addition, logistic regression models were used to determine which characteristics appear to be associated with whether an eligible individual applies for GIS or not (see Data sources and methodology).

Unattached seniors living alone most frequent GIS recipients

Seniors are found in a variety of situations, ranging from living alone to being a member of a family headed by an adult child (Chart A). According to the 1999 SFS, 1.1 million families had at least one GIS recipient, the largest group being seniors living alone (45%) followed by senior couples (24%). About 20% were non-senior headed families where the relationship of the GIS recipient to the major income earner was parent (60%), other relative (26%), or spouse/grandparent/sibling (14%). Among non-GIS families with a senior, the largest group was senior couples (43%) followed by unattached seniors (36%).

Median wealth of unattached GIS recipients one-sixth that of unattached non-recipients

In general, GIS families had lower median income, assets and net worth than their non-GIS counterparts. Median after-tax income was almost twice as high for non-GIS families (Table 1). The greatest differences in terms of wealth were found for unattached seniors where those not receiving GIS had roughly six times more in total assets, nearly seven times more in financial assets, significantly more homeowners, and median net worth nearly six times higher. Homeownership is especially important for seniors since the vast majority are mortgage-free, making direct shelter costs minimal.

The other family types exhibited similar but smaller differences. Median assets and net worth were roughly three times higher for non-GIS senior couples and other senior-headed families compared with their GIS counterparts. Not surprisingly, differences were smallest for non-senior headed families. Since seniors in these households were not the major income earners, they had less impact on the family's financial situation.

GIS families were more vulnerable financially

Results were similar for financial wealth. Unattached GIS recipients had median financial wealth one-sixth that of their non-recipient counterparts (Table 2). The differences for senior couples and other senior-headed families were similar, roughly one-third. The idea that GIS seniors in non-senior-headed families may have been sheltered was again found here. The difference between the median levels of financial wealth was not significantly different for non-senior-headed families.

As noted, median after-tax income was lower for GIS families than for their non-GIS counterparts. Indeed, for senior-headed GIS families (unattached, senior couples, and other senior-headed), income levels were significantly lower. For all three family types, virtually all were found in the bottom two-thirds of the after-tax income distribution, with the largest portion in the lowest third—roughly 60% of unattached and other senior-headed families and over 80% of senior couples. Once again, the picture was not as clear for non-senior-headed families where nearly 30% of the GIS families were found in the top third of the income distribution.

Those in low income with little to no financial wealth are often the most financially vulnerable (Morissette 2002). And GIS families were primarily the ones found to be in the bottom third of both the after-tax income distribution and the financial wealth distribution—nearly 39% of unattached GIS recipients, 36% of other senior-headed GIS families, and just over half of senior GIS couples.

GIS families less able to handle an unexpected major expense

Another way to examine vulnerability is to look at the capacity to handle an unexpected expense of $5,000. For those with moderate or high levels of income or financial wealth, the expense could be managed with savings or by selling assets. Significantly more GIS families reported that they either would not be able to manage or would have to borrow to deal with such an expense (Table 3). This reinforces the finding that among senior families in general, GIS families were relatively more vulnerable than their non-GIS counterparts. Roughly half of unattached GIS recipients and senior GIS couples reported they would have to look beyond their own resources to manage the expense; the proportion was even higher for other senior-headed families (61%). Just over one-quarter of unattached non-GIS recipients and one-third of senior non-GIS couples would have difficulty. Interestingly, the majority of non-senior-headed families, regardless of the presence of a GIS recipient, would have to borrow or could not manage it.

In terms of spending relative to income, the majority of GIS families spent their family income or more. Spending more than one's income is not in itself troubling; rather, the rate at which a senior dissaves is important. In all cases, compared with their non-GIS counterparts, a greater proportion of GIS families had spending levels equal to or greater than their income.14

Given the financial vulnerability of the population covered by GIS, one would expect that those eligible would apply for it. Is this the case and what are the characteristics of eligible seniors who do not apply?

Most individuals eligible for GIS received it...

Of the nearly 3.6 million seniors covered in SLID in 2000, about 1.3 million received the GIS while approximately 206,800 eligible individuals did not.15 The theoretical annual cost of payments for these eligible non-recipients was roughly $300 million.

Overall, 86% of those eligible for GIS actually received it (Table 4). Although women made up a larger portion of the eligible non-recipients, they had a significantly higher take-up rate than men: 88% versus 83%. Among the other statistically significant differences: the Atlantic provinces had a higher take-up rate than the provinces west of Quebec. Higher take-up rates were also evident for individuals with the lowest level of education, those in poor health, and those receiving higher payments. Lower rates were noted for those reporting their major activity as working, and those living in a home owned by a household member.

...but only 41% of those who needed to apply actually did so

Given that take-up rates include a large portion of individuals whose benefits are automatically renewed through the income tax system, the application rate may provide more insight. Overall, only 41% of those who needed to apply actually did so. Among the statistically significant differences in application rates, however, two characteristics stood out: age and annual GIS payment. The youngest seniors were by far the most likely to have applied—64% compared with 23% for those 70 to 79 and 16% for those 80 and over. In terms of payment, the application rate was highest for those getting $2,000 or more—53% compared with 33% for those receiving less than $500.

The Longitudinal Administrative Databank (LAD) yielded similar results to SLID (Table 5).16 Total eligible non-recipients in LAD numbered 195,000 and the annual cost was $275 million. Distributions by sex, age, and region were also comparable between the two sources. The LAD overall take-up rate was 87% (89% for women and 84% for men). While trends for application rates were also similar, the rates from LAD were lower—an overall application rate of 31%. Again, higher application rates were found for younger seniors and for those eligible for the highest payment.

Likelihood of applying related negatively to age, positively to payment amount

To further validate the comparison between recipients who had to apply and eligible non-recipients, a logistic regression was run to predict the likelihood of not applying (Table 6). The regression covered all the characteristics in Table 4 for 143,600 recipients and 206,800 eligible non-recipients. Only two variables were significant: age and annual GIS payment.17

In general, the likelihood of not applying for GIS when eligible increased with age and decreased with payment amount (Chart B). Eligible persons aged 80 and over were a third to a half as likely to apply for the benefit as persons 65 to 69. Interestingly, the majority of those 70 and older were likely not to apply regardless of the payment amount; that is, the age effect superseded the payment effect. The age effect may be partially explained by length of time since the last contact with SDC. Those 65 to 69 would have communicated with SDC within the previous five years to apply for OAS, and thus would have a fresher understanding of the GIS program. Attitude toward the government could also be a factor: older seniors may assume that the government will automatically provide them with whatever they are eligible for.

The payment effect is not surprising given that the incentive to apply increases with the amount to be received. At the 95% confidence level, the lowest and third lowest payment groups were significantly less likely to apply than the highest payment group; at 90%, this was true for all three of the lowest payment groups.

Only one-quarter of eligible persons 66 and over who had to apply actually did

To further examine the age effect, those who turned 65 in 2000 were excluded. This ensured that those remaining in the sample were not individuals who had the opportunity to apply for the GIS at the same time they applied for OAS. This further constraint reduced recipients from 143,600 to 68,900, and eligible non-recipients from 206,800 to 202,800. Since the majority of eligible 65 year-olds ended up receiving GIS, it appears that the problem of take-up arises later when individuals are required to re-apply.

The logistic regression results were similar to the previous model (Table 6). Again the two older age groups were significantly different from the 66-to-69 group, and the 70-to-79 group remained not significantly different from the 80-and-over group. Given that removing the 65 years-olds had the greatest impact on the recipient group, it is no surprise that the likelihood of not applying increased substantially for this age group and had virtually no effect on the 80-and-over group.

The effect of removing the 65 year-olds was that those in the younger age group also ended up more unlikely to have applied—a 73% probability of not applying at the lowest payment level compared with 44% when the 65 year-olds were included (Table 7). Even at the highest payment level, a slight majority were likely not to apply. This result is unexpected since those 66 to 69 would have communicated with SDC during the previous five years and should have had the most up-to-date knowledge of the program.


Established in 1967 by the federal government to assist seniors with little or no income other than Old Age Security, the Guaranteed Income Supplement has played an integral part in reducing low income among seniors. Today, of the 4.2 million seniors who receive OAS, roughly 1.6 million also receive GIS.

Senior families generally have lower after-tax income than non-senior families, but when wealth or net worth is examined, they appear to be better off than non-senior families (Williams 2003). Nonetheless, among families with seniors, GIS families appear to be the least well off. They had lower median incomes, lower median assets, and lower net worth. They also had lower financial wealth and were, relatively speaking, more vulnerable than their non-GIS recipient counterparts. Furthermore, they were less able to handle a major unexpected expense.

The majority of GIS families had spending levels either equal to or above their incomes. This in itself is not necessarily troubling; according to the life-cycle hypothesis, dissaving is a natural part of the accumulation and dispersal of wealth over a lifetime. However, significantly more GIS families than non-GIS families were dissaving.

In order to receive GIS, individuals are required to apply and renew annually. This is done either automatically through the tax system or through a paper application process with SDC. Existing clients are for the most part renewed automatically, but clients applying either for the first time or after losing eligibility are required to do so directly with SDC. For a number of reasons (not understanding eligibility requirements, language barriers, unaware they must apply, diminished cognitive abilities), many of these individuals end up not applying.

In 2000, only 41% of those who needed to apply actually did so. Excluding 65 year-olds, this percentage decreased to 25%. Two factors were associated with whether an individual applied: age and payment amount. Age was negatively related and payment amount positively related, although the negative age effect superseded the payment effect. The most troubling finding was that the probability of a relatively young senior (66 to 69) eligible to receive a benefit of $2,000 or more actually applying was roughly only 50%; for those in the oldest age group it was only 24%.

In response to the Standing Committee on Human Resources Development and the Status of Persons with Disabilities, SDC addressed the issue of eligible non-recipients. Their plan included contacting potential GIS recipients using tax information. The effect on the take-up and application rates will become apparent in the coming years.

Data sources and methodology

The Survey of Financial Security (SFS) collects information from households on their income, education, employment, assets and debts. It thus provides information on the net worth (wealth) of Canadian families. Excluded are those living on Indian reserves and crown lands, residents of the territories, members of religious and other communal colonies, members of the Armed Forces living in military camps, and those living in institutions and residences for senior citizens.4

The Survey of Labour and Income Dynamics (SLID), is a longitudinal survey composed of six-year panels. A new panel is introduced every three years, so two panels always overlap. Each panel consists of roughly 15,000 households—about 30,000 adults—and covers all individuals in the 10 provinces, excluding persons living on Indian reserves and residents of institutions.5 The combined overlapping sample for 1999 to 2001 was used in this study.

The Longitudinal Administrative Databank (LAD) is a longitudinal 20% sample created from the T1 Family File (T1FF). Once selected, individuals are in the sample whenever they appear in the T1FF. As well, part of each year's sample includes a selection of individuals who appear for the first time, making the sample current and cross-sectionally representative. In 2000, LAD included nearly five million individuals.

The SFS was used to examine the assets, debts and wealth of families with senior members. The senior population (65 and over) was split into four family types: unattached seniors living alone, senior couples living alone, other senior-headed families, and non-senior headed families with a senior resident.6 Median amounts were calculated for total assets, financial assets, non-financial assets, debts and wealth.

As well, a concept of financial wealth similar to that used in Morissette (2002) was examined: Financial wealth = non-pension financial assets + liquid non-financial assets (such as vehicles) + RRSPs + RIFs - debts (except mortgages). Combined with income, this concept provided information on the relative vulnerability of GIS families compared with their non-GIS counterparts.

To determine the number of eligible GIS non-recipients in 2000, SLID was used.7 The senior population (65 and over) was divided into four groups based on the SDC marital categories: single, married to a non-pensioner, married to a pensioner, or married to an Allowance recipient. One criterion for GIS eligibility is that individuals must be receiving OAS; thus all non-OAS recipients were classified as ineligible for the GIS.8 Income as defined for the GIS program was then calculated for each record based on 1999 income, and family level cut-offs were then used to determine eligibility in 2000. 9,10 Records were checked to see if GIS was received in 2000 and the results classified into three groups: not eligible, eligible and receiving, and eligible but not receiving.11 Theoretical payment amounts were calculated for eligible non-recipients while actual payment amounts were used for recipients.

The take-up rate is GIS recipients in 2000 as a percentage of those eligible for GIS in 2000.

The application rate is GIS recipients in 2000 who did not receive GIS in 1999 as a percentage of GIS recipients in 2000 who did not receive GIS in 1999 and eligible non-recipients.

The GIS recipients in 2000 who did not receive GIS in 1999 were assumed to represent those who applied for GIS in 2000. (They were not automatically renewed since they were not paid in 1999.) The eligible individuals in 2000 who were not receiving GIS in 1999 represented all those who could have applied for GIS in 2000.

Two logistic regressions were run to examine the characteristics associated with whether an eligible individual applied or not.12 The first regression included all eligible individuals who had to apply, while the second included all eligible individuals who had to apply except those who were 65 in 2000. This was done since those turning 65 in 2000 had the opportunity to apply for GIS at the same time as they had applied for OAS. As such, they may have done so in response to being contacted for OAS.13

The Old Age Security Program

The Old Age Security Act of January 1, 1952, replaced legislation from 1927 requiring the federal government to share the cost of provincially run, means-tested, old age benefits. Benefits now include the basic Old Age Security (OAS) pension, the Guaranteed Income Supplement (GIS), and the Allowance. To be eligible, applicants must be Canadian citizens or legal residents of Canada on the day prior to approval of their claim. Benefits are adjusted quarterly (January, April, July and October) to reflect cost-of-living increases measured by the consumer price index. Old Age Security benefits are taxable while the GIS and Allowance are not.

Old Age Security pension
In 2004, the federal government paid out roughly $21.9 billion in OAS benefits. For July to September 2005, the maximum was $476.97 per month. Persons 65 or older are eligible to receive the full benefit if they have resided in Canada for at least 40 years after age 18. In other cases, depending on specific residency status, age and valid immigration standing on or prior to July 1, 1977, an individual may still qualify for a full pension (more information on the Social Development Canada Web site). Those not qualifying for a full pension but residing in Canada for a minimum of 10 years may still be eligible for a partial pension—one-fortieth of the full pension for each full year in Canada after age 18. As well, special measures may apply to immigrants from countries that have a social security agreement with Canada.

Guaranteed Income Supplement
Of the 4.2 million OAS recipients in June 2005, about 1.6 million also received the GIS, which is payable to OAS recipients with low or no other source of income. GIS recipients must re-apply annually by filing an income tax return or completing an income statement and application. The payment year is July to June. For July to September 2005, the maximum benefit for a single person, or a married person whose spouse did not receive OAS or the Allowance, was $566.87 a month; for someone married to an OAS or Allowance recipient, the maximum was $369.24. GIS benefits are income-tested—that is, they are reduced depending on marital status and income level. For July to September 2005, the annual income ceiling for a single recipient was $13,608. For those with a spouse not receiving OAS, the ceiling was $32,976 while for those with a spouse receiving OAS, it was $17,760.

Allowance and Allowance for the survivor
In June 2005, close to 97,000 persons received the Allowance or the Allowance for the survivor. This is income-tested and paid to a person aged 60 to 64 who is the spouse or common-law partner of an OAS/GIS recipient, or whose spouse or common-law partner has died. It is designed to help survivors and couples living on only one OAS pension. In July 2005, the maximum Allowance benefit for the spouse or partner of an OAS recipient was $846.21 a month (equal to the maximum OAS plus the maximum GIS married rate), while the maximum benefit for a survivor was $934.24 a month. From July to September 2005, the annual income ceiling for the spouse or common-law partner of a GIS recipient was $25,392 and $18,624 for a survivor.

Sources: Social Development Canada: Old Age Security Payment Rates, Canada Pension Plan and Old Age Security - Monthly Statistical Bulletins, ISP Stats Book 2003.


  1. Low-income rates are based on 1992 after-tax, low-income cutoffs.

  2. This system of automatic renewal has been in place since 1999 with most GIS recipients being renewed in this way (HRDC 2002).

  3. The goal was not to compare the wealth of seniors and non-seniors; the differences between the two have already been well documented. According to Williams (2003), seniors had slightly more non-financial assets, lower levels of debt, and significantly more wealth than non-seniors.

  4. According to the 2001 Census, 287,000 seniors lived in collective dwellings (roughly 7% of all seniors).

  5. Seniors living in nursing homes and long-term care facilities are excluded. This was not expected to cause a large bias in the estimated number of eligible non-recipients (although it will have an effect on the take-up and application rates) since program staff in these residences are usually well informed about the programs available to seniors (refer to HRDC 2002 for more information on outreach programs). Also, the informal networks within these residences serve to efficiently spread information. Other exclusions such as homeless shelters, rooming houses, and other temporary accommodation will affect the results; however, the number of seniors in such dwellings was relatively small (roughly 7,000 to 8,000 according to the 2001 Census).

  6. The economic family concept was used—that is, persons living in the same dwelling and related by blood, marriage, common law, or adoption.

  7. Although the sample size for SLID was substantially smaller than for LAD, it was used because of the availability of significantly more explanatory variables. Overall, SLID results were confirmed with LAD.

  8. The Standing Committee's Report pointed out that some individuals who are eligible for OAS are not receiving it. Since receiving GIS depends on receiving OAS, the estimate here does not consider these eligible non-OAS recipients. Based on the 2001 Census, roughly 47,000 non-immigrant seniors over the age of 66 were not receiving OAS.

  9. For married or common-law couples, the combined income of the pensioner and the spouse or partner was taken into account. In some cases, income information was not available for 1999 (less than 4%). In these cases, 2000 or 2001 data were used.

  10. The cut-offs published by SDC are for those receiving the maximum OAS; for those not receiving the maximum, the cut-offs depend on the amount of the individual's OAS benefits. In general terms, the GIS for those receiving partial OAS benefits will be higher by an amount equivalent to the difference between the maximum OAS and their OAS. This was not accounted for in this analysis. However, partial OAS recipients make up only a small portion—less than 4% of domestic recipients in 2000.

  11. A number of assumptions were made to account for the difference in payment year (July to June) versus calendar year: An eligible non-recipient remained a non-recipient for the entire payment year; an individual receiving GIS in 2000 but not eligible based on 1999 income was classified as being not eligible and not receiving if they reported having GIS in 1999; an individual receiving GIS in 2000 but not eligible based on their 1999 income and reporting no GIS in 1999 was classified as being an eligible recipient who received an option (under certain circumstances such as a retirement, an individual can request to have an income estimate used rather than their actual income). These assumptions were not expected to have a significant effect on the results.

  12. Logistic regression estimates the probability of a particular outcome (here, not applying when eligible) as a function of several explanatory variables. The association between each explanatory variable and the outcome is examined while holding all other variables constant. To account for the complex survey design, bootstrap weights and SUDAAN version 8.0 were used.

  13. Since 1996, SDC has automatically sent out OAS application kits to individuals in advance of their 65th birthday. The kit provides information to clients about GIS and they are asked if they wish to apply for it.

  14. See Chawla and Wannell (2005) for a more detailed discussion of spending and saving.

  15. In SLID, imputations for GIS are done for seniors with no available tax information (approximately 21% in 2000). As such, a 100% GIS take-up rate is assumed, meaning an overestimation of the number of GIS recipients and an underestimation of the number of individuals eligible but not receiving. Of the 1.3 million individuals classified as having received GIS in SLID, 25% were imputed.

  16. Estimates produced using LAD may be considered more precise since they are generated using a larger sample size; however, since LAD is based only on tax information, the number of social demographic variables available for analysis is limited, and the population covered includes only those who filed tax returns.

  17. Interactions were included but none were found to be significant.


  • Canada. Parliament. House of Commons. Standing Committee on Human Resources Development and the Status of Persons with Disabilities. 2001. The Guaranteed Income Supplement: The Duty to Reach All. Sixth report. Ottawa. Internet:

  • Chawla, Raj K and Ted Wannell. 2005. "Spenders and savers." Perspectives on Labour and Income (Statistics Canada, Catalogue 75-001-XIE). March 2005 online edition.

  • Ha, Tu Thanh. 2003. "Senior seeks to sue Ottawa over unpaid supplement." Globe and Mail. January 31, 2003. p A7.

  • Human Resources Development Canada (HRDC). 2002. Response to the Sixth Report of the Standing Committee on Human Resources Development and the Status of Persons with Disabilities. Ottawa. Internet:

  • Morissette, René. 2002. "Families on the financial edge." Perspectives on Labour and Income (Statistics Canada, Catalogue 75-001-XIE). July 2002 online edition.

  • Myles, John. 2000. The Maturation of Canada's Retirement Income System: Income Levels, Income Inequality and Low Income among the Elderly. Analytical Studies Branch, Research Paper Series no. 147. Ottawa: Statistics Canada.

  • Thompson, Elizabeth. 2005. "Seniors losing out?" The Gazette (Montreal), January 20, 2005. p A16.

  • Williams, Cara. 2003. "Finances in the golden years." Perspectives on Labour and Income (Statistics Canada, Catalogue 75-001-XIE). November 2003 online edition.

Full article in PDF

Preston Poon is with the Income Statistics Division. He can be reached at (613) 951-4245 or

You need to use the free Adobe Reader to view PDF documents. To view (open) these files, simply click on the link. To download (save) them, right-click on the link. Note that if you are using Internet Explorer or AOL, PDF documents sometimes do not open properly. See Troubleshooting PDFs. PDF documents may not be accessible by some devices. For more information, visit the Adobe website or contact us for assistance.

Home | Search | Contact Us | Français Return to top of page
Date Modified: 2014-05-14 Important Notices