Income Research Paper Series
The effect of government transfer programs on low-income rates: a gender-based analysis, 1995 to 2016

View the most recent version.

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.

Archived

This page has been archived on the Web.

by Adriene Harding

Skip to text

Text begins

Abstract

This paper is a gender-based analysis of the effect of government transfer programs on low income in Canada between 1995 and 2016. It compares the low income situations of couples, unattached women, and unattached men. It addresses the difference in prevalence of low income among women and men by age, labour force status, education level, immigration status, Aboriginal group, and region. It also looks at how specific transfers, such as Canada Pension Plan (CPP) and Quebec Pension Plan (QPP) benefits or federal child benefits for example, affected the rate of low income. The results show that government transfers reduced low-income rates and mitigated the difference between women and men in low-income rates.

Introduction

Low-income rates have historically been higher for women than for men, although the gap has been decreasing over time. In 2016, 13.8% of girls and women and 12.2% of boys and men were below the low-income line^Note 1 in Canada, a difference of 1.6 percentage points. This gap was 3.1 percentage points in 1976.^Note 2

Women can be more vulnerable to low income as a result of demographic characteristics, such as marital / couple status, living arrangements, age, race, education level, and employment status (Lahey, 2010 and Smeeding, 2003). For example, senior women not living in a couple and female lone-parents and their families are at a particularly high risk of low income (Chung, 2004).

This paper compares the low income situations of couples, unattached^Note 3 women, and unattached men. For both practical and conceptual reasons, it is standard for studies of the distribution of income to examine income at the family level.^Note 4 The assumption is that financial resources are shared within the family, however, it is worth considering this assumption from a gender perspective. In the past, women were less likely to work than men and for many their income was therefore tied to their husband/partner (Moyser, 2017). The difference in labour force participation rates of women and men has decreased over time, but labour force participation and wages are still lower for women than for men (Drolet, Uppal and LaRochelle-Côté, 2016; Drolet, 2002 and Mumford and Smith, 2009). Women carry a disproportionate amount of household care, childcare, and elderly care, and these activities are also associated with lower labour force participation, part time employment, and less job stability (Moyser, 2017). Although the difference in household-level low-income rates between women and men is small, individual income levels are quite different: in 2016, the median income of women was $28,200, and the median income of men was $40,100.^Note 5 Therefore, studies of family income may hide other disparities between women and men.

Federal government tax and transfers have a redistributive effect that can reduce low-income rates, and reduce inequality between women and men and among people in different demographic groups. Individuals interact with the tax-and-transfer system through paying personal income tax, goods and services tax, and other taxes, and through receiving transfers. One of the goals of the tax-and-transfer system is to redistribute income from those with higher incomes to those with lower incomes. This is achieved through various government transfer programs, including tax credits, cash transfers, and non-cash programs, such as subsidized housing. Several studies have found a reduction of approximately 50% in low-income rates from taxes and transfers in many developed countries (Picot and Myles, 2005 and Smeeding, 2003).

Individual transfer programs, such as Old Age Security (OAS) and the Guaranteed Income Supplement (GIS), Canada Pension Plan (CPP) / Quebec Pension Plan (QPP) programs, child benefits, and Employment Insurance (EI), affect low-income rates and income inequality differently over time and in different magnitudes (Heisz and Murphy, 2016). Different transfer programs affect different family types—older families may receive OAS and CPP payments while younger families are more likely to receive child benefits or EI. Effects of transfers on seniors are particularly important in reducing their low-income rates (Fox and Moyser, 2018). Examining the effects of these transfers by family type also allows for an analysis of gender and low-income.

This paper is a gender-based analysis of the effect of government transfer programs on low income. It addresses the differences in incidence of low income among females and males of all ages. It describes the groups receiving benefits and how they are affected by transfers, taking into consideration factors such as age, marital / couple status, children, labour force status, education level, Aboriginal status, and immigration status. Finally, the paper looks at how specific transfers affect the rate of low income and what role these transfers have played in the relative income of women and men from 1995 to 2016. The government transfers included in the analysis are Old Age Security (OAS) and Guaranteed Income Supplement (GIS) benefits, Canada Pension Plan and Quebec Pension Plan (CPP/QPP) benefits, child benefits, social assistance benefits, Employment Insurance (EI) benefits, the Working Income Tax Benefit (WITB), and the Goods and Services Tax (GST) / Harmonized Sales Tax (HST) credit. The paper does not specifically examine smaller transfers or provincial programs, although these benefits are included in the analysis of the total benefits. Non-cash programs that would also have an impact on low income, such as subsidized housing, are not included in the sum of government transfers.

Some of the key findings of this paper:

• Transfers reduced low-income rates and mitigated the difference in low-income rates between women and men.
• OAS/GIS and CPP/QPP were the largest benefit programs and had the largest effect in reducing low income among the whole population.
• Low-income rates were highest among female lone-parents and unattached elderly women, and government transfers also had the largest effect on these groups.
• The low-income rate among unattached elderly women increased over time, and OAS/GIS and CPP/QPP became less effective in reducing low income over time for this group.
• Child benefits were increasingly more effective in reducing low income over time for women and men, and more so for women.
• Social assistance became less effective in reducing low income over time for women and men.

The first section of this paper explains the data and methodology. The second section provides a gender-based analysis of low income and the effect of government transfers by economic family type. The third section looks at the effect of government transfers on low income through an intersectional lens by considering how gender intersects with other identity characteristics. The last section describes specific transfer programs and their effects on the low-income rate of their targeted groups over time. An appendix includes detailed data tables.

Data and methodology

This paper uses data from the Survey of Consumer Finances (SCF), the Survey of Labour and Income Dynamics (SLID), and the Canadian Income Survey (CIS). The dataset includes data from different surveys in different years:

• 1995 to 1997: A combination of the Survey of Consumer Finances (SCF) and the Survey of Labour and Income Dynamics (SLID)
• 1998 to 2011: The Survey of Labour and Income Dynamics (SLID)
• 2012 to 2016: The Canadian Income Survey (CIS)

Previous research has shown that data from these three surveys can be used for a historical analysis.^{Note 6Note 7}

The SLID was a longitudinal survey, where the same individuals were followed each year. The SCF was a cross-sectional survey, where in each year different individuals were surveyed. The CIS is also cross-sectional. This paper uses data from all three surveys to provide a cross-sectional analysis of low income and government transfers over time. The limitation of a cross-sectional approach is that the paper cannot analyze the effect of a government transfer on an individual over time. It is not possible to know whether a transfer in one year affected low-income rates in subsequent years among recipients. Moreover, it is not possible to know whether recipients changed their labour market, purchasing, or other decisions in subsequent years as a result of the transfers.

The population sample for these surveys is drawn from all individuals in Canada,^Note 8 excluding residents of the Territories, residents of institutions, individuals living on reserves and other Aboriginal settlements, and members of the Canadian Forces living in military camps. There were 65,782 respondents in the 2016 CIS. All results are weighted to be nationally representative.

Start of text box

End of text box

The Low-income measure, after tax (LIM-AT)

The definition of low income in this paper is the Low-income measure, after tax (LIM-AT). The LIM-AT is calculated at the household level. This means that, if the total after-tax income of a household is below the LIM-AT threshold, each individual in that household is considered to be in low income, including the adults and the children.

The LIM-AT accounts for the size of the household and it also makes an adjustment to reflect the fact that larger households require more resources, but that each additional person does not require the same amount of resources as the last. The equivalence of scale, or adjustment factor, is the square root of the number of persons in the household.

The LIM-AT threshold for a one-person household is defined as 50% of the Canadian median adjusted household after-tax income. For households comprising more than one person, this LIM-AT threshold is multiplied by the adjustment factor (see Table 2).

The LIM is recalculated every year on the basis of the current median adjusted household income. The threshold for being in low income according to the LIM has risen over time because the median income has also risen over time. Hence, the LIM reflects current living standards.

The household level of analysis of the LIM-AT has both benefits and limitations. In this context, a household refers to one person or a group of individuals living in a dwelling (they can be related or not related). When a household is composed of a couple, a family, or other group that shares resources, it makes sense to measure low income at the household level. For example, in a household composed of a child and two parents, one parent may be the primary caregiver and may not have an income or may have a low income, but it would not generally be logical to say that the primary caregiver is in low income if the working parent has a high salary.

Calculating the effect of transfer programs on low-income rates

In this paper, the effect of government transfer programs on low-income rates is assessed by calculating a household's income without a particular transfer, and then determining, on the basis of this income, whether they would have been in low income or not. This hypothetical low-income rate is then compared to the real low-income rate. The analysis is simple: it subtracts the value of the government transfers from the household's after-tax income. If this new hypothetical income is below the LIM-AT threshold, the household would have been in low income, and all individuals in the household (adults and children) are considered to be in low income.

This methodology does not account for behavioural effects, i.e., it does not account for changes in labour market or social decisions in response to government transfers. Previous work has used this methodology, which looks at the "direct effect" of transfers, and found that government transfers reduced low-income rates considerably (Heisz and Murphy, 2016; Picot and Myles, 2005; Picot, Morissette and Myles, 2003; Smeeding, 2003; and Blank and Hanratty, 1993). However, researchers have also addressed methodology limitations and have found an overestimation of the effect of transfers on the low-income rate when not accounting for behavioural effects (Kim, 2000). According to Picot and Myles (2005), among seniors, behavioural effects overwhelmed the direct effects measured in this kind of analysis. In other words, if seniors did not have public pensions to rely on, they would work more or save more income during their working years; this would partially offset the income they did not receive from these pensions. Picot and Myles found that the magnitude of these behavioural effects on low-income rates was smaller among working-age adults and their children (Picot and Myles, 2005). Ben-Shalom, Moffitt and Scholz (2011) found that the overall magnitude of behavioural effects was quite small for American social insurance programs. This included old-age security benefits, which is in contrast to the findings of Picot and Myles.

An additional limitation is that this methodology is not a simulation of a tax-and-transfer model. The same low-income threshold is used to calculate the low-income rate both with transfers and without them. In calculating the low-income rate in the absence of transfers, this methodology does not recalculate the low-income threshold according to the new, lower, median income. Furthermore (assuming no change in income taxes), because people would have lower incomes, many people would be in different income tax brackets and would pay less tax. This is also not taken into account.

Despite these limitations, the methodology used in this analysis is a useful accounting approach to approximate the importance of government transfers, and for comparing the outcomes of women and men, and different demographic groups. It is also a useful approach to understand to whom each of the transfers is going.

Statistical significance of the results

The data used in this paper is weighted to represent the population of Canada. Statistical significance testing analyzes the sample size and weights to identify the sampling error to determine whether the data is likely to correctly represent the real population. Bootstrap hypothesis testing was used to determine whether the differences between percentages for women and men were statistically significant. In the tables and charts in the paper, a star is placed beside the percentages for men if they were found to be statistically significantly different from the value for women, at the 95% confidence level.

The effect of all government transfers^Note 9 on low-income rates

Chart 1 compares women and men and shows that the low-income rate was higher among women than among men in all years from 1995 to 2016. The low-income rate for women increased from 12.6% in 1995 to 13.8% in 2016. In comparison, the low-income rate of men increased from 11.5% in 1995 to 12.2% in 2016.

A range of government transfers have helped Canadians of all ages over time. Chart 1 also shows that during the period of 1995 to 2016, government transfers reduced the low-income rate of men and women, although this effect has decreased over time. In 1995, the low-income rate of women would have been 30.7% without government transfers versus 12.6% with all transfers, a difference of 18.1 percentage points. In 2016, the low-income rate of women would have been 30.2% without transfers; this rate was 13.8% with the transfers. This is a difference of 16.4 percentage points.

As with women, government transfers reduced low-income rates of men over the 1995-to-2016 period.

Government transfers also mitigated the disparity in low-income rates between women and men. In 2016, without government transfers the difference between the low-income rates of women and men would have been 3.8 percentage points. With the actual transfers, the difference was 1.6 percentage points. The differences in low-income rates between women and men with and without transfers were statistically significant for all years from 1995 to 2016 (see Table A1 in the appendix).

The effect of transfers by economic family type

Women in different family situations can be at a greater risk than men in the same situations to fall below the low-income threshold. In particular, women with children and women who are not in a couple are at a much higher risk of low income.

This analysis looks specifically at economic family type. A household may include multiple economic families. Low income according to the LIM-AT is based on household size, which is one person or a group of people living in a dwelling. An economic family is a group of two or more people related by blood, marriage, common-law partnership, adoption, or a foster relationship, living in a dwelling. Therefore when disaggregating low-income rates into economic family types, it is important to be aware that households may also contain individuals who are not related to the economic family(ies) in that household. The fact that a certain economic family type is more likely or less likely to live in a household with other economic families or with individuals to whom it is not related could affect comparisons of low-income rates among the different economic family types.

In the analysis in this paper, for elderly and non-elderly unattached people and couples, there are no other members in the economic family. In other words, they do not live with any other relatives. For female lone-parent families, male lone-parent families, and couples with children, these economic family types include the parent(s), children, and could include other relatives in the household. Any of these economic families could be living in a household with other unrelated economic families or individuals.

Chart 2 shows low-income rates by economic family type. Being unattached^Note 10 increases the risk of being below the low-income threshold for both women and men. In 2016, unattached elderly^Note 11 women had a low-income rate of 34.3%, and unattached non-elderly women had a low-income rate of 30.0%. In comparison, unattached elderly men had a low-income rate of 32.5%, and unattached non-elderly men had a low-income rate of 25.8%. Couples had the lowest rates of low income among all economic family types: 8.7% for elderly couples and 7.5% for non-elderly couples. The low-income rates of non-elderly unattached women and non-elderly unattached men were statistically significantly different with and without transfers. The difference between elderly and non-elderly unattached women and couples was statistically significant with and without transfers. The low-income rates of elderly unattached women and elderly unattached men were only statistically significantly different for the rates without transfers.

The largest difference between women and men was observed when comparing lone-parents. In 2016, the low-income rate was higher among female lone-parent families than among any other family type, at 39.7%. In comparison, male lone-parent families had a low-income rate of 13%. These low-income rates apply to the lone-parent as well as this parent's children. The low-income rates of female lone-parent families and male lone-parent families were statistically significantly different both before and after transfers, as were the rates of female lone-parent families and couples.

Previous research supports this finding that marital / couple status and living arrangements affect low income. Research looking at low-wage workers found that 14% to 19% of married women were in low income and that 35% to 80% women not in a couple were in low income. Among both low-wage workers and the total population, unattached people, people living alone, and lone-parents were found to be most vulnerable to low income (Chung, 2004). A much greater percentage of lone mothers are in low income compared to lone fathers. Differences in the overall characteristics of lone mothers compared to lone fathers may account for a part of these differences in low-income rates of lone-parent families.

In addition to having the highest low-income rates, the largest effect of government transfers was observed for elderly people and female lone-parent families. In a subsequent section, it will be shown that the retirement transfer programs had an important impact on low income among seniors. Chart 2 shows that, for unattached senior women, the low-income rate without transfer programs would have been 77.1% while it was 34.3% with transfers, a difference of 42.8 percentage points. Similarly, for unattached senior men, the difference in low-income rates without and with transfers was 38.9 percentage points, and for senior couples the difference was 44.6 percentage points. Due to government transfers, senior couples had the lowest low-income rate of any economic family type.

The effect of transfer programs was also large for female lone-parents. Without the transfers, the low-income rate for female lone-parents and their families would have been 61.9%. With the transfers, the low-income rate of these households was 39.7%, a reduction of 22.2 percentage points.

The effect of government transfers on low-income rates by socio-demographic characteristics

Low-income rates also vary according to demographic characteristics. Age, education, being employed or unemployed, or living in a rural or urban area may have an influence on low-income rates. Similarly, those with different socio-demographic characteristics may receive a greater or lesser amount in government transfers. This section will look at the effect of government transfers on the low-income rates of different groups. All government transfers are included in this section.^Note 12

Age

Table 3 shows that girls aged 4 years old and younger as well as elderly women of 85 years and older are more at risk than other girls and women of being in low income. In 2016, the low-income rate of girls from birth to age 4 was 18.9% while it was 24.5% for women aged 85 years and older.

It is among seniors that government transfers had the greatest effect in terms of reducing low-income rates. For women aged between 65 and 69 years old, the low-income rate would have reached 48.8% without transfers compared to 12.7% with the transfers. For those aged 85 and older, the low-income rate without transfers would have been 67.4% compared to 24.5% with the transfers.

Low-income trends by age group were similar for girls and women and boys and men with some exceptions for seniors. However, these differences in low-income rates were only statistically significant for the 0-to-4 and 5-to-9 age groups for before and after transfers, and the 20-to-24 age group for without transfers only. For these age groups the low-income rates among girls and women were higher than for boys and men. Previous research found that the wage gap between women and men was larger for older women, and became larger among older age groups (Rodgers, 2006 and Drolet and Mumford, 2012).

Labour force status^Note 13

Chart 3 shows low-income rates for select labour force statuses. The largest low-income rate by labour force status was observed for women who were unemployed either part-year or full-year.^Note 14 Women employed all year had the lowest low-income rates.

Labour force status is the only demographic variable in this paper where men had statistically significantly higher low-income rates than women in some categories. For most categories, the low-income rates of women and men were within 1.6 percentage points. Unemployed men and men not in the labour force all year had higher low-income rates with and without transfers than women, although for unemployed women and men, the difference in low-income rate was not statistically significant.

With transfers, the low-income rate among men not in the labour force all year was 3.7 percentage points higher than for women. For those employed, women and men had similar low-income rates after transfers.

In 2016, government transfers reduced low income among women who were unemployed all year by a large amount, from 54.3% to 32%. The next largest effect was on women who were not in the labour force all year; transfers reduced low-income among this group from 47.1% to 25.3%.

Labour force participation, part-time or full-time status, and type of occupation are also correlated with low income, although these variables are not included in this analysis. Labour force participation is lower among women than men, although this gap has been decreasing over time. In 2015, this difference was 8.9 percentage points (Moyser, 2017). When looking at type of occupation, traditional occupations for women (e.g., services, sales, clerical) have lower wages, and women have on average fewer years of experience because they are more likely to take time off for family reasons (Chung, 2004).

Education level^Note 15

As seen in Chart 4, low-income rates are negatively correlated with education: the low-income rate is lower for women and men with more education. Women with less than a high school diploma had a low-income rate of 24.8% in 2016 while those with a university education had a low-income rate of 8.7%.

For less than a high school diploma, the difference between women and men was not statistically significant. For the other categories, the difference in low-income rates for women and men decreased for higher levels of education. For high school or partial postsecondary education, the gap was 2.3 percentage points. For non-university postsecondary, the gap was 1.9 percentage points, and for university educated there was no gap.

Transfers had a large effect on people with less than a high school diploma, reducing low income among women by 31.6 percentage points. Transfers reduced low income among this category for men by 26.1 percentage points. This group still had the highest level of low income, with 24.8% for women and 19.8% for men. The effect of government transfers on low-income rates actually increased the gap in low-income rates between men and women by a small amount for high school or partial postsecondary, and non-university postsecondary.

Immigration status^Note 15

In 2016, women who immigrated^Note 16 to Canada had a low-income rate of 14.9% while Canadian-born women had a rate of 13.2%. Low-income rates were 1.7 percentage points higher for immigrant women. As shown in Chart 5, the impact of government transfers were similar for both immigrant and Canadian-born women in 2016. Low-income rates were higher for women than men, and this gap was reduced by the transfers for both landed immigrants and Canadian-born individuals. For landed immigrants, this difference between low-income rates of women and men was not statistically significant with or without transfers. Picot and Lu (2017) used the Longitudinal Immigration Database and found that during the 2000s, half the immigrants in low income were in chronic low income (having family low income for five consecutive years or more). Future longitudinal analyses could examine the effect of government transfers on chronic low income.

Aboriginal group^Note 15

The CIS does not include individuals living on reserves or in the Territories. In addition, the weighted population in the CIS of First Nations people, Métis and Inuit is quite different than the population counted in the 2016 Census. This is likely due to the fact that the sample size of Aboriginal people in the CIS is small, particularly the sample size of Inuit people.

In 2016, First Nations women had considerably higher low-income rates than non-Aboriginal women. Métis women had a higher low-income rate than non-Aboriginal women, and the low-income rate for Inuit women was slightly lower than for non-Aboriginal women. Chart 6 shows that, First Nations women had a low-income rate of 27.0% compared to 15.6% for Métis women and 13.5% for Inuit women. Non-Aboriginal women had a low-income rate of 13.7%. One possible contributor to the differences between low-income rates among Aboriginal and non-Aboriginal women is that Aboriginal women are less likely to be employed than non-Aboriginal women (Arriagada, 2016).

There was no statistically significant difference in low-income rates between women and men for any of the categories in this analysis.

Government transfers reduced the low-income rate among First Nations women by 14.8 percentage. Transfers also reduced the low-income rate among Métis women by 14.7 percentage points and by 30.5 percentage points among Inuit women. Statistical significance testing was only performed on the difference in low-income rates between women and men.

Province

Looking at Table 4, in 2016, low-income rates of women and girls by province varied from a high of 17.6% in Newfoundland and Labrador to a low of 7.9% in Alberta. For the most populated province, Ontario, the low-income rate for women was 15%, slightly higher than the national average. The low-income rates of men were not statistically significantly different than those of women in all provinces except Alberta after transfers, where men had a higher low-income rate than women, and Ontario and Saskatchewan before transfers, where women had a higher low-income rate than men.

Transfer programs reduced the gap in low-income rates between provinces. The effect of transfer programs was highest for the provinces with the highest low-income rates.

Rural and urban areas

Chart 7 shows that low-income rates were higher in rural areas. With a rate of 14.4%, women and girls living in rural areas were more likely than women in urban areas to live in low income.

Government transfers helped reduce the low-income rates in both rural and urban areas. Without government transfers, low-income rates of women in rural areas would have been 30.5% while, for those living in urban areas, low-income rates would have been 28.9%. The low-income rates for women and men were not statistically significantly different for rural areas with transfers and for urban areas without transfers.

Statistical analysis of how specific transfer programs have affected low income over the period of 1995 to 2016^Note 17

This section of the paper will review individual federal government transfer programs and their effect on the low-income rate of the targeted group. The government transfers included in this analysis are Old Age Security (OAS) and Guaranteed Income Supplement (GIS) benefits, Canada Pension Plan and Quebec Pension Plan (CPP/QPP) benefits, child benefits, social assistance benefits, Employment Insurance (EI) benefits, the Working Income Tax Benefit (WITB), and the Goods and Services Tax (GST) / Harmonized Sales Tax (HST) credit. The effects of each program are examined by looking at their targeted group: seniors, couples with children, lone-parent families, etc.

Among these transfer programs, OAS/GIS and CPP/QPP are the largest programs in terms of total benefits allocated. In 2016/17, the total amount of payments for OAS/GIS was $48.2 billion, and $42.5 billion for CPP/QPP. Child benefits overtook EI in 2016/17, at $22.1 billion, while EI was $20.7 billion. This was a $4.0 billion increase in child benefits, due to the new Canada Child Benefit (Department of Finance Canada, 2017a; Treasury Board of Canada Secretariat, 2017). In 2016/17, the total amount of GST/HST credits paid out was $4.3 billion (Canada Revenue Agency, 2017c). In 2016 the total amount of payments for the WITB was $1.1 billion (Department of Finance Canada, 2017b). Social assistance programs are administrated by the provinces. In 2016/17 the amount of the Canada Social Transfer was $13.3 billion. This transfer to provinces provides funding for social assistance, social services, post-secondary education, and child development and childcare. Other federal transfer payments also contribute to social assistance (Department of Finance Canada, 2018).

In 2016, for women, the GST/HST credit had the largest number of recipients, followed by CPP/QPP benefits, then child benefits, then OAS/GIS, social assistance, EI, and the WITB. For men, GST/HST benefits and CPP/QPP benefits also had the greatest number of recipients, followed by OAS/GIS, EI, child benefits, the WITB, and social assistance.^Note 18 There were more female than male recipients for all of the transfers except EI and the WITB. Twice as many women as men received social assistance and four times as many women received child benefits.

The effects of these programs on the low-income rate were also different. OAS/GIS and CPP/QPP benefits had the largest effects in terms of reducing low-income rates for women. These were followed by child benefits, social assistance and EI. The WITB and the GST/HST credit had marginal effects. Total income from EI and social assistance fell after the 1990s, and average benefit rates from these programs also declined. Consequently, social assistance became less effective after this time. Income from child benefits increased over time, and consequently became more effective over time (Heisz and Murphy, 2016).

As shown in Chart 8, in 2016, the low-income rate among women was 13.8% and for men was 12.2%, all transfer programs applied. A comparison of low-income rates for each transfer excludes the transfer in question while holding all other transfers constant. The full length of each bar represents the low-income rate that would have been the case without the transfer. We see that OAS/GIS reduced the low-income rate by 4.1 percentage points for women and 3.1 percentage points for men. For CPP/QPP benefits it was 4.6 percentage points for women and 3.4 for men. Social assistance was 1.3 for women and 1.1 for men. EI was 1.5 for both women and men and WITB was 0.1 for both women and men. The GST/HST credit was 0.6 for women and 0.4 for men.

Old Age Security (OAS) and Guaranteed Income Supplement (GIS) benefits and Canada Pension Plan (CPP) and Quebec Pension Plan (QPP) benefits

Introduced in 1965, the Old Age Security (OAS) program along with the Canada Pension Plan (CPP) and the Quebec Pension Plan (QPP) are the basis of the public retirement income system in Canada. The OAS program provides a base income for seniors aged 65 and older while the Guaranteed Income Supplement (GIS) provides additional benefits to low-income OAS recipients. CPP and QPP are retirement pensions to which individuals contribute during their working years. It is intended to provide a partial replacement of eligible earnings at the time of retirement. CPP also includes survivor benefits and a disability pension for individuals who cannot continue regular work at their job.^Note 19

In 2011, 96.5% of Canadians aged 65 years and older received OAS/GIS and 92.1% received CPP/QPP (Heisz and Murphy, 2016). This percentage was likely similar in 2016. In 2016, approximately 3,024,000 women received OAS/GIS and 2,474,000 men received it. That year approximately 3,804,000 women and 3,344,000  men received CPP/QPP.^Note 18

CPP/QPP and OAS/GIS programs are indexed to the consumer price index (CPI) in order to maintain parity with the cost of living. OAS rates are recalculated four times a year on the basis of the CPI. OAS benefits remain the same when the cost of living decreases (Government of Canada, n.d.a). Other than this indexing, the average benefit of these programs has not changed in the time period covered by this paper (Heisz and Murphy, 2016).

Most Canadians aged 65 years and older are eligible for OAS. In 2017, seniors with an annual income under $121,314 were eligible for an OAS pension. The pension amount depends on a person's income as well as on how long they have lived in Canada. The maximum OAS monthly pension from October to December 2015 was $569.95. Individuals with low incomes are eligible to also receive GIS payments. The maximum monthly GIS in January 2018 for an unattached person who receives OAS was $876.23 (Government of Canada, n.d.b).

Effect of OAS/GIS and CPP/QPP on the low-income rate of unattached seniors^Note 11

OAS/GIS and CPP/QPP had an effect in reducing low-income rates among all seniors. In 2016, these transfers reduced the low-income rate from 75.6% to 34.3% among senior unattached women, and from 69.9% to 32.5% among senior unattached men (Chart 9). Over time, OAS/GIS and CPP/QPP have become less effective in reducing low-income among seniors. This is particularly true for elderly unattached women. In 1995, OAS/GIS and CPP/QPP reduced low-income rates for elderly unattached women by 69.3 percentage points (from 78.6% to 9.3%), whereas in 2016 these programs reduced low-income rates by 41.3 percentage points (from 75.6% to 34.3%). Low-income rates steadily increased for elderly unattached women during the period from 1995 to 2016, from 9.3% in 1995 to 34.3% in 2016. These trends were similar for elderly men, although not as pronounced. The low-income rate for unattached elderly men was statistically significantly lower than that of unattached elderly women in most years. For some years, including 2013 to 2016, the difference with transfers was not statistically significant (see Table A2 in the appendix).

Other research supports these findings. Elderly people are vulnerable to low income in many countries. Low income among elders is typically higher among women, much higher among single women, and rises with age. Effects of transfers on elderly people are large in developed countries (Smeeding, 2003). From the Longitudinal Administrative Databank, in 2012, of couples who received CPP aged 70-79, 5% were in low income based on the LICO and 10% would have been in low income without CPP. For single seniors aged 70 to 79, the low-income rate was 31% and would have been 52% without CPP (Employment and Social Development Canada, 2017).^Note 20

Effect of OAS/GIS and CPP/QPP on the low-income rate of seniors in a couple

In 2016, retirement transfers reduced the low-income rate among elderly couples from 51.6% to 8.3% (Chart 10). The low-income rate of this group increased slightly from 1995 to 2016, and government transfers became somewhat less effective over this period. However, neither of these trends were as large as what is seen with unattached elderly women and men. The low-income rate of senior couples with and without transfers was statistically significantly lower than that of unattached senior women in all years, by a large magnitude.

Child benefits

Child benefits in this analysis include all federal and provincial child benefits. Up until mid-2016, the child benefits in effect were the Universal Child Care Benefit (UCCB), the Canada Child Tax Benefit (CCTB), and the National Child Benefit (NCB). In 2016, these three programs were replaced by the Canada Child Benefit program (CCB), a change that is partially represented in the 2016 data in this paper. The CCB was introduced in July 2016, with the other programs ending in June 2016. In 2016, approximately 3,734,000 women and 779,000 men received child benefits.^Note 18 This much larger number for women is partly due to the fact that the Canada Revenue Agency (CRA) states that if the child is living with a female parent, then that parent is usually considered to be the primary caregiver for the purposes of receiving child benefits (Canada Revenue Agency, 2017b).

In 2015, the maximum UCCB was $1,920 per year for each child under 6 and $720 per year for each child aged 6 to 17. The maximum CCTB was $1,503 per year for each child under 18. The maximum NCB was $2,329 per year for the first child, $2,060 per year for the second child, and $1,960 per year for additional children. Therefore all the programs depended on the number of children, and the size of the benefits of the CCTB and NCB depended on family income (Office of the Parliamentary Budget Officer, 2016). These programs were replaced by the Canada Child Benefit in July 2016; the maximum benefits from July 2016 to June 2018 were $6,400 per year for each child under 6 and $5,400 per year for each child aged 6 to 17. The amount of benefits therefore depended on the number of children and on family income (Government of Canada, 2018). Income from child benefits has increased over time, and accordingly these transfer programs have become somewhat more effective at reducing low income (Heisz and Murphy, 2016).

Effect of child benefits on the low-income rate of lone-parent families

Low-income rates are generally very high for lone-parent families in many countries, but there is little analysis on the difference between single mothers and single fathers (Picot, Morissette and Myles, 2003). Differences in characteristics of female and male lone-parents may explain differences in low-income rates. In Canada in 2011, the median age of female lone-parents was 46.4 and the median age of male lone-parents was 49.4. That year 12% of lone mothers and 9.3% of lone fathers had children age five and under (Milan, 2015). According to the Labour Force Survey, in 2016, a much lower percentage of female lone-parents were employed than male lone-parents. This disparity got larger for families with younger children. For families with children under 16, 68.2% of lone mothers were employed and 82.2% of lone fathers were employed. For families with children under 3, 48.8% of lone mothers were employed and 80.2% of lone fathers were employed.^Note 21

Low-income rates with and without transfers were statistically significantly lower for male lone-parent families and couple families than for female lone-parent families in all years. The magnitude of these differences was large (see Table A3 in the appendix).

In 2016, child benefit programs reduced the low-income rate among individuals in female lone-parent families^Note 22 by 14.7 percentage points, from 54.4% without transfers to 39.7% with transfers (Chart 11). In comparison, the low-income rate of individuals in male lone-parent families was 21.8% without transfers and 13% with transfers. The large discrepancy in low-income rates between female and male lone-parent families illustrates the fact that certain family situations (such as being a lone-parent) affect the living standards of women and men and their children differently.

Child benefits increasingly reduced low-income rates among women and men over time due to the increase over time of total payments, including the switch to the Canada Child Benefit partway through 2016. In 1995, child benefits reduced the low-income rate among female lone-parent families by 8.8 percentage points. In 2016, this difference was 14.7 percentage points.

Effect of child benefits on the low-income rate of couples with children

Low-income rates for couples with children were much lower than those for female lone-parent families and somewhat lower than for male lone-parent families. Among individuals living in a couple family with children, the low-income rates without transfers from the child benefits programs would have been 16.9% while it was 9.6% with the transfers (Chart 12).

The effect of the child benefits transfers for couples with children has increased over time, the same as for lone parents. In 1995, a difference of 2.8 percentage points was observed between low income without benefits and low income with benefits; in 2016, this difference was 7.3 percentage points.

Social assistance

Social assistance includes provincial and territorial programs that use means testing to provide a basic income to those at the lowest end of the income distribution. In the 1990s, social assistance benefits were reduced and eligibility restrained. The number of social assistance recipients declined, and average benefit rates fell (Heisz and Murphy, 2016). Accordingly, the effect of social assistance on low-income rates has fallen over time. In 2015, yearly social assistance for a couple with two children ranged from $8,405 in Manitoba to $25,895 in the Northwest Territories (Tweddle, Battle and Torjman, 2016). In 2016, approximately 1,340,000 women and 672,000 men received social assistance.^Note 18

Effect of social assistance on the low-income rate

Social assistance programs have relatively fewer beneficiaries than other benefit programs, and these beneficiaries are often well below the low-income threshold. As a result, these programs have a small effect on the general population. In 2016, social assistance programs reduced the low-income rate of unattached women from 36.7% to 34.3%. Among unattached men, social assistance reduced the low-income rate from 27.6% to 25.5%. Among couples, social assistance reduced low income from 8.5% to 7.9%. The difference between rates with and without transfers for unattached women and unattached men was statistically significant, as was the difference between unattached women and couples.

Effect on the low-income rate of women and men receiving social assistance

Given the relatively small size of social assistance programs, to see the impact of the benefit it is helpful to examine the effect of the program on its recipients only. In Chart 13, looking at only women who received social assistance in 2016 revealed that social assistance reduced the low-income rate from 48.9% to 39.1%. Among recipient men, social assistance reduced the low-income rate from 59.7% to 48%. The low-income rates with and without transfers of women and men were statistically significantly different (see Table A4 in the appendix).

During this time period, sometimes female recipients of social assistance had higher low-income rates, and sometimes male recipients did, although in most years before 2014 the difference was not statistically significant. In addition, we see that social assistance was more effective until 2001, when it became less effective and remained stable over time at that lower level, due to the reduction in benefits in the 1990s. In 1995, social assistance reduced low income among women receiving benefits by 21.6 percentage points, while in 2001 the reduction was 16 percentage points. By 2016, the reduction in the rate was 9.8 percentage points.

Employment Insurance (EI)

Individuals contribute to the Employment Insurance (EI) program while they are working and collect payments when they become unemployed, take parental leave, or are unable to work as a result of illness. EI is a small program in terms of total benefits given and has a small overall effect on the low-income rate. The amount of total benefits from EI has fallen since the 1990s. There have been several reductions in benefits, and fewer people are eligible (Heisz and Murphy, 2016). In 2016, the maximum insurable earnings were $50,800 (Government of Canada, n.d.c). In 2016, approximately 1,249,000 women and 1,297,000 received employment insurance benefits.^Note 18

Women receiving EI are more likely to receive maternal, parental or compassionate care benefit than men receiving EI benefits (Moyser, 2017).

Effect of EI on the low-income rate

In 2016, among unattached women, EI reduced the low-income rate from 36.3% to 34.3%. For unattached men, the rate was reduced from 27.6% to 25.5%. Finally, for couples, the EI program reduced the low-income rate from 9.4% to 7.9%. The difference between rates with and without transfers for unattached women and unattached men was statistically significant, as was the difference between unattached women and couples. EI had a greater effect on men than on women. The effect of the program over time has remained relatively stable.

Effect on the low-income rate of women and men receiving EI

Like social assistance, EI is a smaller program. As a result, it had a small effect on the general population. Again, it is helpful to examine its effects on its recipients only (Chart 14). For women who received EI, EI reduced the low-income rate from 17.4% to 8.5% in 2016. Its effect on women was relatively constant over the period from 1995 to 2016. In almost all years, there was no statistically significant difference in the low-income rates of women and men, and in many years there was no difference in the rates for women and men without transfers (see Table A5 in the appendix). The effect of EI on male recipients was larger than its effect on female recipients in most years until 2003. The difference in the low-income rate of recipients was relatively stable over time.

Working Income Tax Benefit (WITB)

The Working Income Tax Benefit (WITB)^Note 23 was introduced in 2007. It is a tax credit for individuals who are working and are in low income. In the 2016 tax year, the maximum income level for eligibility was $28,576 for a family with children and $18,529 for unattached individuals without children. The maximum benefit was $1,028 for single individuals without children and $1,868 for families (Canada Revenue Agency, 2017d). In 2016, there were over 1.4 million recipients for a total of $1.1 billion in benefits (Department of Finance Canada, 2017b). Approximately 635,000 women and 772,000 men received WITB benefits.^Note 18

Effect of WITB on the low-income rate

The WITB effect on low-income rates in 2016 on unattached women, unattached men and couples was marginal or null. The difference between rates with and without transfers for unattached women and unattached men was statistically significant, as was the difference between unattached women and couples.

Effect on the low-income rate of women and men receiving WITB

Chart 15 shows the effect on the low-income rate of the WITB program among women and men who received WITB. In 2016, the credit reduced low-income among women from 36.8% to 35.4%, and among men from 39.9% to 39.1%. The difference in low-income rates for women and men was only statistically significant with and without transfers in 2007, 2008, and 2014 (see Table A6 in the appendix). In these years, male recipients of WITB had higher low-income rates than female recipients, and the effect of the WITB was similar in magnitude for women and men.

Goods and Services Tax (GST) / Harmonized Sales Tax (HST) credit

The GST/HST credit is a quarterly tax-free payment. The goal is to offset the amount of GST/HST that low-income households pay. The credit is calculated according to family net income and the number of eligible children. The maximum payment amount was $276 for the 2015 tax year (paid in July 2016-June 2017) for an individual, plus an additional $276 credit for a spouse, and for each eligible dependent or children. The credit amount started getting clawed back after an annual family income of $35,926 (Canada Revenue Agency, 2017a). The GST/HST credit is a small transfer program in terms of total benefits given, but many people receive this small credit. In 2016, approximately 5,165,000 women and 5,331,000 men received the GST/HST credit.^Note 18

Effect of the GST/HST credit on the low-income rate

The GST/HST credit had a small and constant effect on low-income rates of women over time. In 2016, the program reduced the low-income rate among unattached women from 35% to 34.3%, for unattached men from 26.2% to 25.5%, and for couples from 8.3% to 7.9%. The difference between rates with and without transfers for unattached women and unattached men was statistically significant, as was the difference between unattached women and couples.

Effect on the low-income rate of women and men receiving the GST/HST credit

Chart 16 shows that among women and men who received the GST/HST credit, the low-income rate for women was reduced from 30% to 28.9% in 2016, and the low-income rate for men was reduced from 27.7% to 26.7%. The difference in low-income rates among female and male recipients increased and then decreased again. The effect of the transfer over time was stable. In 2011 and 2013-2015, there was no statistically significant difference in low-income rates with and without transfers between women and men (see Table A7 in the appendix).

Conclusion

This paper analysed the effect of government transfer programs on low-income rates from a gender-based perspective. It also examined the intersection of gender with other characteristics. More specifically, the paper analysed the effect of Old Age Security (OAS) and Guaranteed Income Supplement (GIS) benefits, Canada Pension Plan and Quebec Pension Plan (CPP/QPP) benefits, child benefits, social assistance benefits, Employment Insurance (EI) benefits, the Working Income Tax Benefit (WITB), and the Goods and Services Tax (GST) / Harmonized Sales Tax (HST) credit. The paper measured low income using the Low-income measure. The principle underlying the LIM is that a family is in low income if their income is less than 50% of the median income of all families, taking into account differences in family size.

Some of the main findings:

• Transfers reduced low-income rates and mitigated the difference in low-income rates between women and men.
• OAS/GIS and CPP/QPP were the largest benefit programs and had the largest effect in reducing low income among the whole population.
• Low-income rates were highest among female lone-parents and unattached elderly women, and government transfers also had the largest effect on these groups.
• The low-income rate among unattached elderly women increased over time, and OAS/GIS and CPP/QPP became less effective in reducing low income over time for this group.
• Child benefits were increasingly more effective in reducing low income over time for women and men, and more so for women.
• Social assistance became less effective over time for women and men.

A simple approach was used to analyze the effect of government transfers on low-income rates. Although this methodology does not account for behavioural effects, it approximates the effect of government transfers, and is particularly useful in comparing those effects on different groups. The impact of these transfers on low income, particularly among certain demographics, was large, demonstrating that government transfers have an important role in reducing low income.

These differences in the effect of government transfers reflect differences in demographics. Certain groups are more likely to be in low income. In particular situations, especially regarding the marital / couple status and living arrangements of women and whether they have children, the low-income rate of women would have been much higher than those of men in the same situation without the transfer. Therefore benefits to seniors have a large effect, not only because the benefits are large, but also because elderly women, particularly unattached elderly women, are at a higher risk of being in low income. Likewise, child benefits have a large effect on female lone-parent families because they are at a higher risk of being in low income.

Some caveats of this research include the fact that this method likely overestimates the effect of transfers because it does not account for behavioural effects, and that the analysis of low income is at a household level, not an individual level. In addition, the available data and scope of the paper did not allow for multivariate analysis, additional disaggregation of demographics for each transfer, or analysis of gender that includes analysis by sexual orientation, and analysis of transgender, cisgender, and non-binary people.

Future analyses could utilize longitudinal survey and administrative data to follow the same individuals over time. This would allow for an analysis of how individuals and families adjust their behaviour according to the transfers they receive, and how transfers in one year can impact low income in subsequent years. Future research that focuses on individual women and men, their relationships and living arrangements, and their incomes and resource sharing could also give additional insight into gender and low income. An analysis with larger data sets, such as the Canadian Census, could also permit disaggregation into more detailed sub-groups.

References

Arriagada, P. 2016. "First Nations, Métis and Inuit Women." Women in Canada: A Gender-based Statistical Report. Statistics Canada. Catalogue no. 89-503-X. Ottawa, Ontario.

Ben-Shalom, Y., R.A. Moffitt and J.K. Scholz. 2011. "An Assessment of the Effectiveness of Anti-Poverty Programs in the United States." National Bureau of Economic Research. Working Paper 17042.

Blank, R.M. and M.J. Hanratty. 1993. "Responding to Need: A comparison of Social Safety Nets in Canada and the United States." Small Differences That Matter: Labor Markets and Income Maintenance in Canada and the United States, edited by D. Card and R.B. Freeman. National Bureau of Economic Research, University of Chicago Press.

Canada Revenue Agency. 2017a. "GST/HST Credit Including Related Provincial Credits and Benefits. For the Period from July 2016 to June 2017." Government of Canada. Ottawa, Ontario. Retrieved from https://www.canada.ca/content/dam/cra-arc/migration/cra-arc/E/pub/tg/rc4210/rc4210-17e.pdf

Canada Revenue Agency. 2017b. "Canada Child Benefit and Related Provincial and Territorial Programs. For the Period from July 2017 to June 2018." Government of Canada. Ottawa, Ontario. Retrieved from https://www.canada.ca/en/revenue-agency/services/forms-publications/publications/t4114/canada-child-benefit.html#primaresp

Canada Revenue Agency. 2017c. "GST/HST Credit Statistics – 2016-2017 Benefit Year." Government of Canada. Ottawa, Ontario. Retrieved from https://www.canada.ca/en/revenue-agency/programs/about-canada-revenue-agency-cra/income-statistics-gst-hst-statistics/gst-hst-credit-statistics/gst-hst-credit-statistics-2016-2017-benefit-year.html

Canada Revenue Agency. 2017d. "T1─2016, Schedule 6: Working Income Tax Benefit." Ottawa, Ontario.

Chung, L. 2004. "Low-Paid Workers: How Many Live in Low-income Families?" Perspectives on Labour and Income. 16(4): 23–32. Statistics Canada. Catalogue no. 75-001-XPE. Ottawa, Ontario.

Crespo, S. 2017. "Et si les conjoints ne partageaient pas tous leurs revenus? Conséquence sur la mesure des inégalités du bien-être." Cahiers québécois de démographie, 46(1): 73-99.

Department of Finance Canada. 2017a. "Annual Financial Report of the Government of Canada Fiscal Year 2016-2017." Government of Canada. Ottawa, Ontario. Retrieved from https://www.budget.gc.ca/2017/docs/plan/anx-01-en.html#Toc477707546

Department of Finance Canada. 2017b. "Backgrounder: Enhancing the Working Income Tax Benefit." Government of Canada. Ottawa, Ontario. Retrieved from https://www.fin.gc.ca/n17/data/17-103_2-eng.asp

Department of Finance Canada. 2018. "Federal Support to Provinces and Territories." Government of Canada. Ottawa, Ontario. Retrieved from https://www.fin.gc.ca/fedprov/mtp-eng.asp

Drolet, M. 2002. "Can the Workplace Explain Canadian Gender Pay Differentials?" Canadian Public Policy, 28: S41-63.

Drolet, M. and K. Mumford. 2012. "The Gender Pay Gap for the Private-Sector Employees in Canada and Britain." British Journal of Industrial Relations, 50(3): 529-553.

Drolet, M., S. Uppal S. and S. LaRochelle-Côté. 2016. "The Canada-U.S. Gap in Women's Labour Market Participation." Insights on Canadian Society. Statistics Canada.

Employment and Social Development Canada. 2017. "Summative Evaluation of the Canada Pension Plan Retirement Pension and Survivor Benefits." Employment and Social Development Canada. Catalogue no. SP-1138-03-17E. Ottawa, Ontario.

Fox, D. and M. Moyser. 2018. "The Economic Well-Being of Women in Canada." Women in Canada: A Gender-based Statistical Report. Statistics Canada. Catalogue no. 89-503-X. Ottawa, Ontario.

Government of Canada. n.d.a. "Old Age Security Amounts and the Consumer Price Index." Ottawa, Ontario. Retrieved from https://www.canada.ca/en/services/benefits/publicpensions/cpp/old-age-security/oas-price.html

Government of Canada. n.d.b. "Quarterly Report of Canada Pension Plan and Old Age Security Monthly Amounts and Related Figures – October to December 2015." Ottawa, Ontario. Retrieved from https://www.canada.ca/en/employment-social-development/programs/pensions/pension/statistics/2015-quarterly-october-december.html

Government of Canada. n.d.c. "EI Premium Rates and Maximums." Ottawa, Ontario. Retrieved from https://www.canada.ca/en/revenue-agency/services/tax/businesses/topics/payroll/payroll-deductions-contributions/employment-insurance-ei/ei-premium-rates-maximums.html

Government of Canada. 2018. "Canada child benefit – How we calculate your CCB." Ottawa, Ontario. Retrieved from https://www.canada.ca/en/revenue-agency/services/child-family-benefits/canada-child-benefit-overview/canada-child-benefit-we-calculate-your-ccb.html

Heisz, A. and B. Murphy. 2016. "The Role of Taxes and Transfers in Reducing Income Inequality." The Art of the State Volume V: Income Inequality – The Canadian Story, edited by D.A. Green, W.C. Riddell and F. St-Hilaire. The Institute for Research on Public Policy.

Kim, H. 2000. "Do Welfare States Reduce Poverty? A Critical Shortcoming in the Standard Analysis of the Anti-Poverty Effect of Welfare States." Luxembourg Income Study. Working Paper no. 33.

Lahey, K.A. 2010. "Women, Substantive Equality, and Fiscal Policy: Gender-Based Analysis of Taxes, Benefits, and Budgets." Canadian Journal of Women and the Law, Vol. 22.

Milan, A. 2015. "Families and Living Arrangements." Women in Canada: A Gender-based Statistical Report. Statistics Canada. Catalogue no. 89-503-X. Ottawa, Ontario.

Moyser, M. 2017. "Women and Paid Work." Women in Canada: A Gender-based Statistical Report. Statistics Canada. Catalogue no. 89-503-X.

Mumford, K. and P.N. Smith. 2009. "What Determines the Part-Time and Gender Earnings Gaps in Britain: Evidence from the Workplace." Oxford Economic Papers, 61(1): 56-75.

Office of the Parliamentary Budget Officer. 2016. "Fiscal Analysis of Federal Children's Benefits." Ottawa, Ontario. Retrieved from http://www.pbo-dpb.gc.ca/web/default/files/Documents/Reports/2016/CCB/CCB_EN.pdf

Picot, G. and Y. Lu. 2017. "Chronic Low Income among Immigrants in Canada and its Communities." Statistics Canada. Catalogue no. 11F0019M – no. 397. Ottawa, Ontario.

Picot, G., R. Morissette and J. Myles. 2003. "Low-Income Intensity During the 1990s: The Role of Economic Growth, Employment Earnings and Social Transfers." Statistics Canada. Catalogue no. 11F0019MIE – no. 172. Ottawa, Ontario.

Picot, G. and J. Myles. 2005. "Income Inequality and Low Income in Canada: An International Perspective." Statistics Canada. Catalogue no. 11F0019MIE – no. 240. Ottawa, Ontario.

Rodgers, W.M. 2006. Handbook on the Economics of Discrimination. Northampton, MA: Edward Elgar Publishing.

Statistics Canada. 2016. "Low Income Lines: What They Are and How They are Created." Statistics Canada. Catalogue no. 75F0002M – no. 002. Ottawa, Ontario.

Smeeding, T. 2003. Government Programs and Social Outcomes: The United States in Comparative Perspective. University of California-Berkeley.

Treasury Board of Canada Secretariat. 2017. "Canada Pension Plan (CPP) and Old Age Security (OAS) Annual Statistics Tables." Open Government: Government of Canada. Ottawa, Ontario. Retrieved from https://open.canada.ca/data/en/dataset/f064a144-b4e0-4e9b-9970-e0fc2f84d1a8

Tweddle, A., K. Battle and S. Torjman. 2016. Welfare in Canada, 2015. The Caledon Institute of Social Policy. Retrieved from https://maytree.com/wp-content/uploads/CaledonPDF..pdf

United Nations. 2017. Guide on Poverty Measurements. United Nations Publications.

Appendix

Table A1
Low-income rates of individuals by sex, all ages, with and without transfers, 1995 to 2016

Table A2
Low-income rate of elderly unattached women, men, and couples, with and without OAS/GIS and CPP/QPP, 1995 to 2016

Table A3
Low-income rate of individuals in female lone-parent families, male lone-parent families, and couple families, with and without child benefits, 1995 to 2016

Table A4
Low-income rate of individuals who received social assistance, by sex, with and without social assistance, 1995 to 2016

Table A5
Low-income rate of individuals who received EI, by sex, with and without EI, 1995 to 2016

Table A6
Low-income rate of individuals who received WITB, by sex, with and without WITB, 2007 to 2016

Table A7
Low-income rate of individuals who received the GST/HST credit, by sex, with and without the GST/HST credit, 1995 to 2016

Notes