Income Research Paper Series
Social assistance recipients in Canada: Documenting the filing patterns of recipients

Summary

By combining data from administrative files, such as the T5007 slip that provides information about receipt of social assistance (SA) with data on income tax returns (T1), this paper aims to document the rate of non-tax filing among those who have received social assistance by different sociodemographic variables. This paper also investigates the potential eligibility of non-filers for the GST and HST and Climate Action Incentive non-refundable credit.

Start of text box

Highlights

• Compared to the general population, social assistance (SA) recipients under the age of 65 are much less likely to file: men are almost twice as likely to be non-filers than women among this population in the under 65 years old group. Senior recipients, however, are less likely to not file and are more equal across genders.
• Temporary residents represent a small proportion of SA recipients (1.7%), but they are more likely not to file and that rate is growing (24.0% in 2019 and 41.1% in 2021).
• Prior interactions with the tax and benefit system influence current behaviour: recipients who had filed last year are around ten times more likely to file this year (5.9%) than intermittent filers who missed the previous year (56.6%) or never filers (63.8%).
• Northwest Territories, Alberta and Nova Scotia have lower non-filing rates than other jurisdictions.
• Provincial and territorial differences in filing rates among SA recipients may be affected by differences in the demographic composition of SA recipients targeted by regional SA programs. For example, presence of an SA program targeting seniors (who are more likely have a filing history and must file for other benefits like the Guaranteed Income Supplement) would naturally lead to a lower non-filing rate in this area.
• In the 45 to 64 years old age group, over one in five filing SA recipients (22.0% in 2021) are filing with the help of the Community Volunteer Income Tax Program or similar volunteer programs.
• GST and HST credits for the 330,000 non-filing social assistance recipients in 2020 left unclaimed or received late are estimated to have been 119.3 million dollars. Unclaimed payments for the Climate action initiative payments are estimated at 68.7 million.

End of text box

Introduction

In 2017, the Canada Revenue Agency (CRA) estimated that more than one in ten individuals did not file their taxes (Office of the Auditor General of Canada, 2022). Non-filing individuals may be precluded from obtaining certain government benefits that are managed through the tax system. According to the Office of the Auditor General of Canada’s Access to Benefits for Hard-to-Reach Populations (2022) report, although the CRA had taken steps to reduce the non-filing rate, a significant number of people were still not filing and not receiving the benefits they were entitled to claim. Many of those were from the bottom of the income distribution and could be expected to be entitled to receive benefits.

In this paper, the non-filing rate of recipients of social assistance (SA) is documented. Specifically, this study looks at individuals receiving a T5007 Statement of Benefits slip^Note  with an amount in box 11 - Social assistance payments or provincial or territorial supplements – in any of the years 2019 to 2021. Within this population, the analysis focuses on the characteristics of individuals who do not file a tax return. Even if recipients are not required to file a tax return unless they have taxes owing, many may decide to file in order to claim certain benefits or refundable credits also managed through the tax system. The paper further examines how many of the remaining non-filing SA recipients could be expected to be entitled to two such government programs: the Goods and Services Tax and Harmonized Sales Tax credits (G/HSTC), and the Climate Action Incentive (CAI).

By combining data from different administrative files including the T5007 slip and the T1 Income tax and benefit return (T1), this paper presents the rate of non-tax filing (or “non-filer rate”) between 2019 and 2021 among those who have received social assistance as well as describes the characteristics of recipients who have not filed a tax return for those years.

This analysis thus answers three research questions:

• Of those receiving social assistance payments, how many filed a T1 and how many did not?
• What are the characteristics of the individuals who received social assistance but did not file a tax return?
• Of those who did not file a T1, is it possible to roughly predict how many might also be eligible to receive G/HSTC or CAI? If so, how large were the unclaimed benefits missed by recipients who did not file?

This paper starts with a description of typical social assistance programs for which T5007 are issued and describes the dataset constructed for the analysis. Then, a short presentation of the methodology is followed by the non-tax filing rate results. Finally, as an illustration of some of the downsides of not filing, this paper investigates the potential eligibility of non-filers for G/HSTC and CAI and is capped by some concluding remarks.

Social assistance programs

Social assistance programs encompass a wide variety of support mechanisms mainly provided by provincial, territorial, and Indigenous governments but also by some private charitable institutions. For most of them, they are considered income support programs of last resort. Typically, a payment is made to individuals with low family incomes or in straitened financial circumstances or urgent need to meet their basic necessities such as food, shelter, and clothing. Benefits are not taxable but are to be reported in the income tax return. This general social assistance heading includes several financial assistance packages such as provincial income supports, income assistance for persons with disabilities, and more, depending on the province or territory (Maytree Foundation, 2021). Although these types of programs extend across Canada, the eligibility rules, benefit levels and administrative processes differ from one jurisdiction to another. The data used in this paper show only the total amount of social assistance paid to the individual and not the specific program(s) under which the benefits were allocated. SA programs and eligibility criteria vary among provinces and territories. Therefore, the demographic composition and amount of benefits received by SA recipients can also vary depending on the jurisdiction.

Data

This section describes the database constructed for the analysis, bringing together the different sources of data. The T5007 administrative file is the primary source defining the universe of persons of interest.

T5007 Statement of benefits

The supplementary T5007 database includes information on individuals who have received a T5007 slip for any of the years 2019 to 2021. The T5007 slips were sent to recipients at year end and filed directly with the CRA by provincial, territorial, or municipal agencies or boards. These data are then used by Statistics Canada for various analytical and statistical purposes such as this project. For each slip, the database includes data elements, including social assistance amounts and Workers' Compensation Benefits that recipients have received in the year and the province in which they lived. No other additional personal attributes were retrieved from the dataset, like the sex, age, or the presence of children of the recipient, as this information is not on the T5007 slip. The T5007 will be restricted to recipients of social assistance payments (as reported in box 11 on the slip) and therefore constitute the population of analysis for each of the years.

T1 Income tax and benefit return

The T1 Income tax and benefit return (T1) data holds information on the circumstances of individuals who filed their tax return by the end of December of the year following the year of social assistance receipt. The main attribute derived for the analysis is the presence on the T1 file in the same year which permits the calculation of the filing and non-filing rates for all T5007 recipients. From presence in earlier years, one can also infer filing habits and experience or familiarity with the tax system. For the subset of tax filing recipients, additional data are also used such as receipt of CAI, amounts received, date the tax return was filed, and whether the tax return was prepared by the filer on their own or with the help of somebody else such as a community volunteer.^Note 

Goods and services tax and Harmonised sales tax credit

The Goods and services tax and Harmonised sales tax credit (G/HSTC) data available at Statistics Canada are an extract of the administrative files and include amounts paid under the G/HSTC program to individuals and couples, generally four times a year.^Note  The G/HSTC database is used to establish for the subset of T5007 recipients who had filed, the number who were entitled to the credit after they file and the credit amount.

Linkage Control File and T1 Ident

The data from the Linkage Control File (LCF) is used to obtain, for all social assistance recipients, the sex and date of birth (the latter serves to compute the age of recipients). The T1 Ident database includes the marital status of the recipient, and the presence of dependent(s) aged under 18 years old. It also identifies individuals who have not filed a tax return at one point in the last 40 years. The T1 Ident and the LCF database are only used to obtain those specific attributes to classify the recipients for descriptive statistics on filers and non-filers.

Methodology

This section presents the technique to integrate the data sources listed above defining the population of social assistance recipients as well as the calculation of the tax filing rates.

Data linkage

First, by using the social insurance number (SIN) as the linkage key between the T5007, T1, G/HSTC, LCF and T1 Ident, information on the amount of social assistance (SA) benefits the recipient received, the various socio-demographic data on the recipient and information on recipients who did not file a tax return were able to be brought together. Second, to be able to link these files, slips needed to be excluded from the original T5007 dataset, such as slips which did not have a valid social insurance number (SIN) or that only presented Worker’s Compensation Benefit and no social assistance amount. Some observations in the dataset had more than one slip containing SA benefits, so they were aggregated to a single record with a total amount. Once linked, individuals under the age of 18 years old were excluded from the analysis. Lastly, all SINs are removed from the constructed files as a standard procedure to preserve anonymity and privacy of Canadians (names and addresses are not kept either). Following the data integration steps, the population of analysis now consists of recipients 18 years of age or older who received social assistance benefits ($1 or more). Within the linked dataset we can now identify the sex, age, presence of children, marital status, location of residence and previous filing behaviour of social assistance recipients who may or may not have filed a T1.

Filers, non-filing rate and late filers

A 'non-filer' is defined in this article as an individual who is found in the T5007 file but is not present in the T1 generated by December 31st of the following year. Thus, a recipient that received SA benefits and filed a T1, either to report earnings or other income or claim eventual benefits or tax credits will be found in both the T5007 and T1. It is important to note that an SA recipient could file even later than the end of December and still be entitled to benefits, often including retroactive payments. Therefore, this non-filer status is defined at a point in time and is not necessarily a permanent state.

Table 1 shows the number of observations in the T5007 that constitute the population of analysis, i.e., individuals who received social assistance. The slips present in the T5007 file varies from 2.83 million to 2.89 million observations per year. Of those observations, 2.67 to 2.73 million observations had $1 or more of social assistance payments.^Note  Then, some T5007 slips were excluded from analysis because they have either no SIN or invalid SIN. Finally, other slips were excluded because there was inappropriate age data. Table 1 presents between 1.97 and 2.06 million individuals per year in receipt of social assistance, of whom 290,930 to 313,720 did not file a T1. Table 1 shows that there are between 28.7 and 29.3 million filings of the T1, from 2019 to 2021, identifying 1.68 to 1.75 million returns within our population of interest.

Results

This section focuses on the results of the analysis, presenting answers to the research questions defined previously. Chart 1 illustrates the non-filing rate in 2021 for our social assistance recipient universe (T5007) by age and sex, contrasting it with the non-filing rate of the general population. By using data on the population estimates by age and sex at the end of the year and using total filers from the full T1 data, an implicit non-filing rate among the general Canadian population was computed and placed in the figure. The chart shows that, regardless of age, men have a higher non-filing rate than women. Also, compared with the general population, the SA recipient population has a higher non-filing rate for individuals between the ages of 18 and 65. After 65 years old, this situation is reversed. Also, SA recipients’ non-filing rates are higher, especially for men, between the ages of 18 and 65. Then, the non-filing rates decrease for those over 65 years old. The non-filing rates reach almost 30% for men in their late thirties receiving social assistance, while this rate is less than half as high for men in the general population.

Data table for Chart 1

Data table for Chart 1 Table summary
This table displays the results of Data table for Chart 1 Social assistance recipients, Men, Women and General population, calculated using units of measure (appearing as column headers).

	General population		Social assistance recipients
	Men	Women	Men	Women
Note: Population estimates January 1st of the following year, by sex and age. The T5007 population is restricted to recipients who received $1 or more of SA benefits. Sources: Statistics Canada, Population estimates by age and sex, 2022 and T1 and T5007, 2021.
Age	percent
18	27.4	25.2	24.1	25.4
19	22.0	19.4	21.4	20.9
20	19.3	16.0	21.8	18.0
21	19.8	15.6	22.0	16.2
22	18.7	13.5	22.2	15.5
23	16.5	10.9	23.6	15.3
24	15.2	9.7	24.1	15.2
25	13.9	8.2	23.7	13.6
26	13.6	7.4	24.8	14.3
27	12.6	6.4	25.3	13.6
28	13.1	6.2	26.2	13.8
29	12.0	5.1	27.3	14.1
30	12.8	6.1	26.9	14.3
31	11.8	5.3	28.3	13.9
32	12.3	5.6	28.4	13.8
33	11.8	5.6	28.6	13.8
34	11.9	6.0	28.0	13.5
35	12.1	6.3	28.9	14.0
36	12.0	6.8	28.6	13.7
37	11.0	5.9	29.3	14.4
38	11.0	6.5	29.2	14.2
39	10.3	5.9	29.5	14.5
40	10.8	7.1	28.3	14.6
41	10.3	7.1	29.3	14.6
42	10.1	6.9	28.3	14.2
43	10.6	7.9	27.9	15.0
44	9.9	7.5	28.3	15.5
45	10.6	8.0	27.8	14.7
46	10.8	8.3	28.0	15.1
47	10.8	8.7	27.1	15.2
48	11.0	8.5	28.2	15.2
49	10.5	8.3	27.8	15.1
50	11.0	8.3	26.9	14.8
51	11.1	8.4	26.9	15.3
52	11.5	8.5	25.9	14.5
53	10.8	8.2	25.7	15.0
54	11.9	8.9	25.1	13.4
55	11.4	8.7	24.4	13.0
56	11.8	8.6	23.8	13.2
57	10.2	7.2	23.4	12.7
58	10.0	7.2	22.2	11.8
59	9.5	6.5	20.6	11.1
60	10.2	7.5	20.3	11.2
61	9.1	6.4	19.1	11.0
62	8.9	6.3	17.9	10.8
63	8.6	6.2	15.8	8.9
64	6.2	4.7	11.3	6.7
65	6.3	4.8	8.4	4.5
66	6.3	4.6	6.1	3.6
67	5.2	3.9	5.1	3.0
68	5.7	4.4	4.6	2.7
69	4.3	3.3	4.1	2.5
70	6.0	4.9	3.9	2.7
71	5.3	3.9	3.9	2.6
72	4.7	3.6	3.5	2.5
73	5.1	4.1	3.8	2.7
74	3.2	2.3	4.0	2.9
75	3.5	2.5	3.6	2.9
76	7.6	6.7	3.5	2.9
77	5.7	4.2	3.7	2.8
78	5.6	4.3	3.7	3.2
79	5.8	4.3	4.0	3.2
80	7.8	7.0	3.5	3.4
81	6.6	5.0	3.7	3.3
82	8.3	6.2	4.2	3.7
83	7.4	6.3	4.4	3.4
84	9.9	7.2	4.0	3.4
85	10.3	8.0	4.7	4.0
86	11.5	8.5	4.9	4.4
87	12.3	9.3	4.9	4.2
88	14.0	11.7	4.6	4.6
89	14.9	11.3	5.1	4.9
90	17.3	13.9	6.8	4.8

Beyond age and sex characteristics, it is important to properly assess which groups are more likely to be non-filing among SA recipients. Table 2 presents some descriptive statistics by sex, age group, the presence of children, temporary resident status, filing status, mailing address province or territory on the main T5007 slip and by social assistance benefit range. The table presents the proportion among the SA population in 2021 and the non-filing rates in each group from 2019 to 2021. For example, the 2021 distribution shows that there were more women (52.5%) receiving social assistance benefits than men (47.5%). In terms of non-filing, results show that, within the SA population, men are less likely to file than women. In fact, around one in five men (19.7%) but one in ten women (10.3%) who received benefits did not file a tax return in 2021.

Youth 15 to 24 years old represented less than one in ten SA recipients (8.2%) while the 25 to 64 years old group represented two in three recipients (66%) in 2021 and the 65 and older age category represented one in four recipients (25.8%). Among those groups, the 65 years old and over recipients are the most likely to file. Indeed, in 2021, only 4% of them did not file a T1 whereas in the other age groups, it was closer to 20% of non-filers.

These results could be partially explained by filing incentives. Indeed, low-income seniors are also potentially eligible for the federal Guaranteed Income Supplement, which also has a filing requirement. It may also be the case that there are filing requirements to be eligible for certain provincial or territorial senior benefits that are reported on the T5007 as social assistance. These filing requirements to access other programs or credits might increase the incentive to file among older recipients.

Table 2 also shows that nearly 86% of the SA population had no presence of dependents under 18 years old in 2021. These recipients with dependents were substantially more likely to file a tax return. In 2021, one in seven recipients without children (16.5%) did not file a T1, while 4.6% of recipients who had children were non-filers. Again, for this population, the presence of substantial incentives might be influencing behaviours since filing is a requirement to obtain the Canada Child Benefit to which this group of our SA population would typically be entitled.

As for temporary residents – these represent less than 2% of SA recipients – the non-filing rates were higher and did show sizeable growth over the three years of this study: from one in four (24.0%) in 2019 to two in five (41.1%) in 2021. This may be driven by pandemic-related changes to various SA programs admissibility criteria or to the composition of temporary residents.

Among the SA population, nearly 99% of the recipients had filed at least one tax return previously. Their behaviour in the year just preceding the receipt of SA was a good indicator whether they would file this year: in 2021, 56.6% of previous-year non-filers did not file again, but previous-year filers were almost ten times more likely to file (5.9% not filing). For the more than 1% of the recipient population who had never filed a tax return before 2021 – the year in which they received SA – almost two-thirds (63.8%) remained non-filers.

In terms of geographical representation more than two in five recipients are from Ontario and around one in six SA recipients are from each of Alberta, British Columbia and Quebec. Less than one percent of recipients are from the Northwest Territories, Nunavut, Prince Edward Island or Yukon. This distribution does not mirror the distribution of the population (e.g., Quebec had 14.4% of social assistance recipients in 2021 but 22.4% of the population on January 1, 2022) reflecting different eligibility profiles for each province or territory’s programs.

The non-filing rate among social assistance recipients is higher in Yukon (19.9%) than in any other Canadian jurisdiction in 2021. Other high non-filing rates appear in Manitoba, Ontario, British Columbia, and New Brunswick while the Northwest Territories, Nova Scotia and Alberta have lower non-filing rates. One notable ranking is Quebec where 13.6% of recipients do not file, relatively close to the Canadian average (14.7%). In other research, when considering other population groups, Quebec’s non-filing rate was rather lower than the Canadian rate.^Note  There was a decrease from 2020 to 2021 in non-filing rates for the SA population in all provinces and territories.

Considering the size of SA benefits, nearly one in five recipients (18.1%) received less than $1,000 in 2021, and this low-benefit group had the lowest non-filing rate (5.9% in 2021) while non-filing rates for groups receiving higher benefits went rapidly up, reaching 26.0% for persons receiving $8,000 to $8,999 and declining thereafter as benefits were higher. These observations prompted an examination by age group of the composition of each benefit range to better understand the differences.

Looking at the proportion by age group among filing and non-filing recipients in each social assistance benefit range, Chart 2.1 and Chart 2.2 shows how the distributions were heterogenous across filing status in 2021. The left graph shows the composition of non-filers and the one to the right presents the composition of filers. Each bar of the graph represents the composition in percentage by age group within the range benefit specified by the vertical axis. The further the bar is in the chart towards the bottom, the smaller the amount of benefits was. For example, 30.3% of non-filers who received between $1 and $999 are 18 to 44 years old, while 54.3% of non-filers who received benefits in the same range are 65 and older. Although, 4.8% of filers who received between $1 and $999 are 18 to 44 years old, while 91.6% are 65 to 90 years old. Chart 2.1 shows that there is more than one in two non-filing recipients of less than $1,000 in benefits who are between the ages of 65 and 90, while it’s almost 92% of in the case of filing recipients of that benefit level. This proportion among the older group decreases as benefits increase for both non-filers (Chart 2.1) and filers (Chart 2.2). There is a clear compositional effect in that older T5007 recipients do not receive the SA amounts typical of younger recipients.

Data table for Chart 2.1

Data table for Chart 2.1 Table summary
This table displays the results of Data table for Chart 2.1 Non-filers, 18 to 44 years old, 65 to 90 years old and 45 to 64 years old, calculated using units of measure (appearing as column headers).

	Non-filers
	18 to 44 years old	45 to 64 years old	65 to 90 years old
Sources: Statistics Canada, T1 and T5007, 2021.
Social assistance benefits amount	percent
$30,000 and over	49.8	46.6	3.6
$15,000 to $29,999	44.8	53.6	1.6
$10,000 to $14,999	51.0	47.2	1.8
$7,500 to $9,999	57.5	40.9	1.7
$5,000 to $7,499	59.5	37.4	3.0
$4,000 to $4,999	60.2	35.3	4.5
$3,000 to $3,999	58.7	33.4	7.9
$2,000 to 2,999	55.9	32.5	11.6
$1,000 to $1,999	54.6	29.6	15.9
$1 to $999	30.3	15.4	54.3

Data table for Chart 2.2

Data table for Chart 2.2 Table summary
This table displays the results of Data table for Chart 2.2 Filers, 18 to 44 years old, 65 to 90 years old and 45 to 64 years old, calculated using units of measure (appearing as column headers).

	Filers
	18 to 44 years old	45 to 64 years old	65 to 90 years old
Sources: Statistics Canada, T1 and T5007, 2021.
Social assistance benefits amount	percent
$30,000 and over	45.7	50.0	4.3
$15,000 to $29,999	47.3	50.5	2.2
$10,000 to $14,999	50.7	45.8	3.5
$7,500 to $9,999	47.4	46.9	5.7
$5,000 to $7,499	47.1	42.5	10.3
$4,000 to $4,999	45.6	35.4	19.0
$3,000 to $3,999	36.0	25.7	38.3
$2,000 to 2,999	25.3	17.0	57.7
$1,000 to $1,999	20.0	12.9	67.1
$1 to $999	4.8	3.6	91.6

Since older recipients are eligible for other credits and transfers, like the Old Age Security pension (OAS) and/or the federal Guaranteed Income Supplement (GIS), social assistance may be inaccessible by program design in certain jurisdictions and amounts may decrease. While the proportion of recipients receiving benefits between $1,000 and $14,999 are predominantly recipients between the ages of 18 and 44, the non-filers who received more than $15,000 in benefits are typically older (45 to 64 years old). For filers, as the social assistance benefits increase, the more likely the recipients are aged between 18 to 64 years, specifically, recipients who are between the age of 45 to 64 years old receive higher amount of benefits (more than $15,000) comparing to the group of 18 to 44 years old.

The previous chart suggested that older individuals received lower SA benefits than other age groups, while slightly younger taxpayers received higher benefits on average. Table 3 shows more clearly the average SA benefit amount received by age group, sex, and tax filing status. Whereas older recipients are known to receive lower benefits, within that same age group, non-filers were receiving more SA benefits on average than filers. This pattern is reversed for recipients aged 18 to 64. In addition, male recipients have lower average benefits than female recipients, except for older non-filer and filer recipients. Recipients between 45 and 64 years of age have highest average benefits, rounding up to between $9,900 to $11,500 for both non-filers and filers.

Learning a bit more about non-filers in the last few subsections, the following focuses on comparing the distribution of two groups of taxpayers (social assistance recipients and Canadian taxpayers) who filed their return with assistance from a community volunteer.^Note  Table 4 shows that just over one in ten SA recipients had their tax return filing done within a community volunteer program in 2019, while this proportion is less than two percent in the general population. The proportion of SA recipients helped with their tax return grew by more than three percentage points from 2019 to 2021 (32% increase over two years). For the Canadian tax filers in general, the use of volunteers grew by less than half a percentage point from 2019 to 2021.

Consistently over time, the use of the programs is highest in the SA recipients among those 45 to 64 years old, compared with other age groups. In 2021, the 65 years old and over group of the SA-receiving population was the least likely to participate in a clinic. While within the general population, that same older age group was more likely overall to get help to file their return. Although, regardless of the age group and the population, the proportion is increasing year over year.

Estimate of credits and incentives missed

The last objective of this paper is to evaluate whether some recipients of SA benefits would have been entitled to G/HSTC and CAI payments, had they filed their taxes, and what the amount of these unclaimed benefits could have been. The analysis focuses only on the G/HSTC credit and the CAI for potential entitlement; other programs are not evaluated in this paper.^Note  In the case of these two programs, the recipient is only eligible to receive the credits if they file their taxes. Therefore, any recipient who has not filed their taxes is automatically deprived of both amounts though they could file their return even later past the cut-off for the T1 file used in this paper (late December of the year following the income year) and receive amounts retroactively.^Note  This section considers the methodology used to estimate the potential entitlement and missed benefits.

Potential credit entitlement and estimate of amount not received by non-filers

First, it is not possible to know for certain what proportion of non filers would have been eligible to receive these credits if they had filed. However, by making certain assumptions about these recipients it is possible to estimate this proportion by extrapolating the amounts paid to similar filers. In order to estimate an individual's entitlement for the tax credit and the climate action incentive payment, it is assumed that the proportion of non-filers entitled to the tax credit would be about the same as that for the filers with similar characteristics.

The CAI eligibility is based on the marital status of the recipient, the presence of children and it is not awarded in all provinces/territories. The G/HSTC eligibility is based on the recipient’s family income, age, marital status, and the presence of children. Taking these eligibility criteria into account, these characteristics are used for similarity analysis: the province or territory, age, marital status, presence of children and SA benefit range.^Note  This means that a filer and a non-filer with similar characteristics (i.e., living in the same province as the filer, same marital status, sex, age range, have children or not and have the same range of benefits) could be assumed to have the same likelihood of being entitled to these credits.

Thereafter, to find the number of non-filers who could be entitled to credits, the proportion of SA-recipient filers who are receiving credits is multiplied by the observed number of non-filers in each similar group. This way, an estimate of the number of recipients who would have been entitled to the credits had they declared their income is produced. Likewise, to estimate the total amount of unclaimed benefits, the estimated number of non-filers who would have been eligible for either tax credit is multiplied by the average amount claimed by similar filers. For example, if the average tax credit amount claimed by filers for 2019 is $300 and the estimated number of non-filers that could have been entitled is 10,000, then the unclaimed tax credit amount would be $3,000,000. The estimate of total aggregate unclaimed amounts is the sum of the aggregate estimates for the G/HSTC and the CAI payment.

Table 5 shows that for 2020 tax year,^Note  83.2% of the filing recipients of social assistance benefits also received the G/HSTC, while over half of them received CAI benefits (55.5%). Then, based on their characteristics and the patterns observed in similar filers, it is possible to estimate that 88.0% of non-filing SA recipients would have been entitled to the G/HSTC and around 56.3% of non-filers would have been entitled to CAI benefits, if they had filed a T1. The average amount received by filing SA recipients overall for the G/HSTC was $500 for 2020 and $440 for those receiving the CAI payment. Table 5 presents estimates for non-filers showing that the average amount of G/HSTC could be 420$ and 380$ for those entitled to CAI benefits.

The estimate of total unclaimed benefits is $188 million in 2020. An increase of 15.1% from 2019 to 2020 in the estimated total amount of unclaimed benefit is also observed, mainly driven by the increase of the G/HSTC and CAI average amounts received by filers. It can be noted that there was no apparent decrease in the estimate of total unclaimed benefits for 2020 when labour markets and income levels were disrupted in the first year of the pandemic (i.e., lower earnings and higher COVID-19 benefits). This pattern is different than seen for the Canada Workers Benefit – for which there was a 2020 dip – presumably because the last-resort population receiving was less likely to be eligible for employment-related amounts such as the Canada Emergency Response Benefit (CERB) and the Canada Recovery Benefit (CRB).^Note 

Concluding remarks

By augmenting data from the T5007 with characteristics from other administrative files and an indicator from an end-of-year version of the T1 file, it was possible to document the filing behaviour of those who were in receipt of social assistance. It was also possible to provide a reasonable estimate for a pair of credits to which these social assistance recipients may have been entitled had they filed their T1 Income tax and benefit return (T1). Each year between 2019 and 2021, 14.7% to 16.3 % of social assistance recipients did not file a T1. The results also showed that older recipients – typically recipients of lower social assistance amounts – were more likely to complete their income tax form presumably because of incentives such as additional benefits or because of program requirements. While recipients aged 25 to 64 who typically had higher social assistance benefits, they were less likely to file a T1 (40% in 2021). Also, among all SA recipients, one in five men did not file a T1, while this proportion is lower for women (one in ten).

There have been substantial efforts at getting vulnerable groups to file more often. Nearly 20% of T1 returns by SA recipients between the ages of 45 and 65 were prepared by a community volunteer program, a trend that appears to be increasing over time.

This paper also presented a preliminary evaluation of some of the estimated costs associated with non-filing for the roughly 300,000 recipients not filing in each of the years. Of those, it is estimated that around 5 out of 6 recipients could have been entitled to the G/HSTC and 56% or so of them could have been entitled to the CAI, had they filed a T1. Therefore, non-filing could have resulted in the loss – or late payment if they have eventually filed – of $163 million to $188 million in G/HSTC and CAI payments that could have been claimed by social assistance recipients.

They are a few limitations in this paper. First, it is assumed that eligibility and receipt amounts of G/HSTC and CAI of non-filers is the same as filers with the same characteristics, while it is impossible to accurately assess this information on an individual level from the data available. Second, when discussing non-filing, the paper is structured around analyzing the characteristics individually and we have seen there might be interactions to consider among people’s different attributes. For example, when discussing jurisdiction of residence or amount of SA benefits, the influence of age groups and the presence of seniors has great impact. It may be appropriate to revisit those aspects under a multivariate lens.

While this paper meets the original objectives, there are still further questions of interest, and additional studies may deepen our understanding. The use of administrative data has allowed to explore the non-filing issue for the social assistance recipients. Linking other databases identifying different vulnerable groups would certainly provide an enhanced picture of non-filing trends and the potential associated benefit and losses for other populations.^Note  Robustness tests under different specifications may also help interpret the results obtained.

While the study identifies socio-demographic characteristics correlated with non-filing, there is limited opportunity to identify specific barriers to non-filing for each vulnerable segment. It may therefore be useful to supplement this research with either direct interviews with these population groups or by synthesizing feedback from community volunteer program participants to better understand the mechanisms that may lead to non-filing.

References

Abdulla R. and Laporte, C. (2023). Who is filling a tax return? Results from the 2021 Census, Statistics Canada

Office of the Auditor General of Canada (2022). Report 1—Access to Benefits for Hard-to-Reach Populations. Government of Canada.

Zheng, C. (2023). The Filing Patterns of Wage Earners and Receipt of the Canada Workers Benefit. Statistics Canada. Catalogue no. 75F0002M. Income Research Paper Series.

Appendix