Women Are Key for Future Growth
Evidence from Canada
  • 1 https://isni.org/isni/0000000404811396, International Monetary Fund

Contributor Notes

How important are female workers for economic growth? This paper presents empirical evidence that an increase in female labor force participation is positively associated with labor productivity growth. Using panel data for 10 Canadian provinces over 1990–2015, we found that a 1 percentage point increase in the labor force participation among women with high educational attainment would raise Canada’s overall labor productivity growth by 0.2 to 0.3 percentage point a year. This suggests that if the current gap of 7 percentage points between male and female labor force participation with high educational attainment were eliminated, the level of real GDP could be about 4 percent higher today. The government has appropriately stepped up its efforts to improve gender equality, as part of its growth strategy. In particular, the government’s plan to expand access to affordable child care is a positive step. However, we argue that to maximize the policy outcome given a budget constraint, provision of subsidized child care—including publicly funded child care spaces—should be better targeted to working parents.

Abstract

How important are female workers for economic growth? This paper presents empirical evidence that an increase in female labor force participation is positively associated with labor productivity growth. Using panel data for 10 Canadian provinces over 1990–2015, we found that a 1 percentage point increase in the labor force participation among women with high educational attainment would raise Canada’s overall labor productivity growth by 0.2 to 0.3 percentage point a year. This suggests that if the current gap of 7 percentage points between male and female labor force participation with high educational attainment were eliminated, the level of real GDP could be about 4 percent higher today. The government has appropriately stepped up its efforts to improve gender equality, as part of its growth strategy. In particular, the government’s plan to expand access to affordable child care is a positive step. However, we argue that to maximize the policy outcome given a budget constraint, provision of subsidized child care—including publicly funded child care spaces—should be better targeted to working parents.

I. Introduction

“It’s not just women’s issues. It’s all of our issues. And teaching that to our kids is really, really important. Because we know if kids grow up in a more equal world, it is a better word.” —Canadian Prime Minister Justin Trudeau.

Canada has made great strides in boosting female labor participation over the past several decades. In 1976, only half of Canadian women were in the labor market, below the United States, and much lower than the Nordic economies (Figure 1). Four decades later, Canada’s female labor force participation rate has increased to nearly 75 percent, a far more impressive progress than in many other advanced economies, including the United States (where the rate has actually fallen since the middle of the 1990s). As a result, the gap with the Nordic economies has almost been eliminated.

Figure 1.
Figure 1.

Selected Advanced Economies: Female Labor Participation, 1976-2015

(Percent, ages 25 to 54 years old)

Citation: IMF Working Papers 2017, 166; 10.5089/9781484309247.001.A001

Source: OECD Stat.

The greater inclusion of women in the economy has been an important source of growth in Canada. Between 1976 and 2015, total labor force in Canada increased by 8.8 million (or 84 percent), 5.2 million of which were women. Without women, Canada’s economy would have been much smaller. In addition, the integration of women in the labor market seems to have enhanced productivity as more and more women with a post-secondary education entered the labor market. Indeed, the increase in female labor force participation especially between the mid-1990s and the mid-2000s was in tandem with an increase in labor productivity in Canada (Figure 2). Today, Canadian women are among the most educated in the world.

Figure 2.
Figure 2.

Canada: Female Labor Force Participation Rate and Labor Productivity

Citation: IMF Working Papers 2017, 166; 10.5089/9781484309247.001.A001

Sources: Statistics Canada; and Haver Analytics

However, despite significant progress in recent decades, gender gaps remain in Canada. Female labor force participation lags male participation by about 9 percentage points. In addition, Canada’s gender wage gap is well above the OECD average (OECD 2017), and progress in promoting women’s representation in senior management has been slow, with women making up only one in four senior managers. Accordingly, more effort is still needed to improve gender equality and further integrate women in the labor market to lift overall economic growth.

Why is gender equality an especially important economic issue now? The Canadian economy, like other advanced economies, is expected to face growth headwinds, arising from unfavorable demographic pressure. The Bank of Canada has assessed that labor input growth could be almost halved (from 0.7 percent a year in 2015 to 0.4 percent in 2020) due to the slowdown in the growth rate of the working age population (Bank of Canada 2016). But female labor resources have not been fully tapped, and policy reform to bring more females to the workplace could effectively counterbalance the demographic pressure that is predicted to rise in the future.

The primary objective of this paper is to analyze the impact of female labor participation on labor productivity growth in Canada. Several studies have analyzed gender equality and growth (see, for example, IMF 2013; OECD, ILO, IMF, and World Bank 2014; and IMF 2017). However, to our knowledge, this paper is the first to explicitly quantify the impact of female labor participation on labor productivity growth using Canada as a case study. The paper is organized as follows. Section II analyzes women’s contribution to growth. Section III discusses policies to tap the full potential of female labor force. Section IV concludes.

II. Analyzing women’s Contribution to Growth

Unless otherwise noted, our analysis focuses on the core working age group (ages 25–54) for the following reasons.

  • It is sensible to conjecture that workers belonging to the core working age have the highest labor productivity potential in their life cycle, as many studies suggest (see for example, Skirbekk 2003 and Feyrer 2007).

  • This age group dominates the labor force, accounting for about two-thirds of the total female labor force.

  • Many people in this group have family and young children. This means that this age group must handle important decisions about work-life balance. This pressure is more acute for women, given the traditional division of labor between men and women in a family.

A. Trends of female labor participation in Canada

In Canada, over the past few decades, the growth of female labor force has constantly exceeded that of male labor force. From late 1970s to the 1980s, the female labor force of the core working-age group increased by 5–6 percent a year on average, compared with 2.2 percent for men (Table 1). The pace of female labor force expansion has since declined but has consistently surpassed male labor force growth. As a result, the gap between men and women in the labor force participation rate has narrowed from nearly 39 percentage points in the late 1970s to about 9 percentage points in the past five years. Furthermore, the number of women with full time jobs has also gradually grown: the portion of women working full time accounted for three-fourths of working women in the late 1970s and now exceeds 80 percent.

Table 1.

Canada: Labor Participation Trends

(Ages 25–54, percent, period average)

article image
Source: Statistics Canada; and Haver Analytics

There has also been a gradual but an important shift in the composition of female labor force. Table 2 shows the composition of hours worked by educational attainment. The rapid increase in women’s hours worked—above 40 percent growth on average per decade in the 1970s and 1980s—was driven mainly by women with relatively lower and moderate educational attainment (namely, “primary or secondary education” and “some or completed post-secondary education”). Over time, though, women with high educational attainment (“university degrees or above”) grew faster than other categories: in the 1960s, only 1.7 percent of total female labor force (measured as worked hours) had university degrees, compared with 4.8 percent for men. The share of highly educated females has since then risen steadily and faster than men, and by 2010–16, it reached about 25 percent of total female workforce, whereas for men it was only about 22 percent.

Table 2.

Canada: Hours Worked by Educational Attainment

(Period average)

article image
Sources: Statistics Canada, CANSIM Tables 383-0024 and 282-0004; and authors’ estimates.

CANSIM Table 383-0024 provides data only through 2010. Data beyond 2011 are based on author’s estimates using CANSIM Table 282-0004.

How important are these trends for boosting economic growth?

The positive labor supply effect is one important factor, which we can calculate using growth accounting. To this end, because the focus of this paper is not on capital, we decompose GDP growth into the change in hours worked (that is, the change in the number of workers multiplied by each worker’s working hours) and the labor productivity growth (change in output per hours worked). Women’s total hours worked increased strongly over the past 35 years: between 1990 and 2015, the average annual growth rate was 1.7 percent, compared with men’s average of 1 percent. The increase in women’s total hours worked contributed to about 30 percent of overall output growth in the 1980s and was shy of 23 percent—still greater than men’s—in the following two decades, as the growth of the female labor force moderated (Figure 3).

Figure 3.
Figure 3.

Growth Decomposition

(Period average, contributions to overall growth rates)

Citation: IMF Working Papers 2017, 166; 10.5089/9781484309247.001.A001

Sources: Author’s estimates based on Statistics Canada CANSIM 383-008, 383-0024, and 383-0029.

Productivity is another factor in the influence of women’s labor participation on growth. The growth of female labor participation has been increasingly driven by women with high educational attainment. As well discussed in the economic literature on growth, better education—and associated improvements in human capital—has been among the important determinants of labor productivity, and thus economic growth (for example, Mankiw and others 1992, Barro and Lee 1993 and 2000). How important has the productivity channel been in Canada? To answer this question, we ran productivity growth regressions, discussed in the next section.

B. Empirics: female labor participation and productivity growth

Empirical strategy and data

The goal of our empirical analysis is to find statistical relationships between female labor participation and labor productivity growth. We also attempt to find how different characteristics of female workers have contributed to labor productivity growth. To this end, we begin with examining the broadest pool of female workers, and then delve deeper into its subgroups, for example by educational attainment. The key variables used in the analysis—labor productivity and labor force participation—are calculated as follows, based on national accounts and labor force survey estimates from Statistics Canada.

  • Labor productivity (LP). Real GDP per total hours worked. Growth of labor productivity is calculated as the first difference of the logarithm of labor productivity.

  • Female labor force participation. We examine two indicators.2

    • Female labor force participation rate (LWPW), defined as the number of female labor force participants (LW) as a percentage of female population (PW). The labor force participation rate for a specific subgroup—for example, females with high educational attainment—is calculated as the number of labor force participants with a specific characteristic as a percentage of the total population with the same characteristic.

    • The share of female workers with a specific characteristic—for example, the share of female workers with high educational attainment, LW,High—in the female population is calculated as LW,HighPW.3

We use data from the 10 Canadian provinces.4 This approach has the advantage of facilitating consistent comparison across provinces over time, thus avoiding the measurement problems associated with many growth studies using large cross-country data sets.

The sample period is from 1990 to 2015. The annual data are averaged for five 5-year periods. We are interested in long-term relationships and thus want to exclude cyclical effects arising from short-run business fluctuations. The sample data set is balanced (see data description and descriptive statistics in Annex I).

Figure 4 plots statistical relationships between labor productivity growth and various measures of female labor participation for the full sample period. We observe the following:

Figure 4.
Figure 4.

Labor Productivity Growth and Female Labor Participation 1/

(1990–2015 average)

Citation: IMF Working Papers 2017, 166; 10.5089/9781484309247.001.A001

1/ For Newfoundland and Labrador (NL), an outlier outcome for 2002 is excluded. In that year, Terra Nova offshore oil field started operating, boosting oil production by 90 percent and the province’s real GDP by 16 percent. With a small increase in hours worked, labor productivity grew by 16 percent in Newfoundland and Labrador in 2002.
  • A positive relationship between female labor force participation rate and labor productivity growth rate (top left panel).

  • Clear evidence that the participation rate for women with high educational attainment is positively correlated with labor productivity growth (top right panel).

  • The share of female workers with high educational attainment seems to be positively associated with labor productivity growth, although the slope of the regression line is relatively flat (bottom left panel).

  • Positive associations are more evident for female workers with both high educational attainment and full-time employment (bottom right panel).

Regression model

Following the setup used in many growth studies (for example, Barro and Lee 2001), we estimate the following model:

ΔLPit = α + β[female labor participation variables]it + γ[control variables]it + μi + εit,

where i and t denote province and time, respectively; μi denotes an unobserved fixed effect capturing time-invariable heterogeneity across provinces; and εitIID(o,σε2) is a white-noise error term.

The dependent variable, ΔLPit, is labor productivity growth. Explanatory variables are divided into female labor participation variables and a set of control variables. The control variables are those most relevant to productivity growth and following the growth literature include the initial level of labor productivity;5 research and development spending per hours worked; the growth rate of foreign exports and interprovincial trade; the growth rate of capital stock per hours worked; the change in net migration inflows as a percent of population; and the rule of law index (a proxy for the quality of institutions; see Annex I for data description).

We estimate the above equation using the fixed-effects model and the generalized method of moment (GMM) estimator. Among various GMM estimators, we choose the difference GMM estimator from Arellano and Bond (1991), because it addresses problems related to inconsistent estimators due to variable endogeneity and a relatively short sample period combined with a fairly large cross section data (“small T, and large N”).6 We use as instruments, lagged explanatory variables, the number of births per woman, voice and accountability index, the percentage share of female legislators, and that of female senior officials and managers. These instruments are used as proxies for a female-friendly environment, a role model effect, and expected career improvement among women.

Results

We present the main regression results in Table 3.

Table 3.

Canada: Regression Results

article image

GMM specifications. As instruments, lagged explanatory variables, “fertility rates,” “voice and accountability index,” and “share of female legislators, senior officials and managers,” are used for instrumenting female labor force variables. Robust standard errors are in parentheses, with *** indicating significance level at 1 percent, ** at 5 percent, * at 10 percent, and + 15 percent.

  • Models 1 and 2 present the results of fixed-effects models. Consistent with our priors, female labor participation rates are positively and significantly correlated with labor productivity growth, with the greater contribution by females with high educational attainment.

  • Models 3–9 are based on the GMM estimator. The coefficient on “all females” (which includes all types of educational attainment) turns negative and insignificant (model 3), but the coefficient on “females with high educational attainment” is still positive and highly significant. This result is robust even if we control for “males with high educational attainment” (model 5). The negative sign on the males’ coefficient is surprising: it could be interpreted as a negative contribution by men to productivity. But this result may instead reflect a statistical issue: that the participation rate of highly educated men has been broadly flat at a high level for the sample period, and thus lacks meaningful variance both across time and across sections.7

  • The positive contribution of “females with high educational attainment” to productivity growth does not appear to be persistent. Model 6 shows that the coefficient on the quadratic term of the female variable is negative, indicating diminishing marginal returns to scale—i.e., the positive contribution of the females with high educational attainment would eventually diminish after the participation increase wanes.

  • “Females with low educational attainment” is also positively associated with labor productivity growth, but the coefficient is much smaller and insignificant (model 7).

  • The positive association between female participation and productivity growth is particularly strong for females working full time (model 8).

  • An alternative female labor force variable—the share of highly educated female labor in the female population—is also positive and highly significant (model 9).

Simulations of the impact of higher female labor participation on economic growth

The regression results permit a rough simulation of the impact of higher female labor participation on GDP. Although caution is needed when interpreting causality, the predicted impact on GDP can be calculated by estimating the direct impact due to an increase in labor input and the indirect impact through the productivity enhancing channel.

Table 4 presents the simulation result, which underscores the nontrivial contribution of the female workforce for Canada’s long-term growth potential.

Table 4.

Canada: Calculations of Growth Impacts

article image
Source: Authors’ calculation.
  • For men and women with high educational attainment, the gender participation gap averaged 6.8 percent percentage points in 2010-15. This implies that if the gender gap were to be closed, the number of female workers would be raised by 0.3 million, equivalent to 1.8 percent of total labor force (including both men and women with all levels of educational attainment at all ages).

  • We use model 4 to calculate the impact on productivity. 8 The estimated coefficient is 0.317 (with the standard error of 0.149). Assuming that the participation rate of females with high educational attainment increases by 6.8 percent percentage points to close the gender gap, labor productivity growth would be higher by 2.2 percentage points (with ±one standard error band (68 percent significance) of 1.2–3.2 percentage points).

  • Adding the two effects, the level GDP would be higher by 4 percent (with ±one standard error band (68 percent significance) of 3–5 percent).

III. Policies to Tap the Full Potential of the Female Labor Force

A. Where are the remaining gaps in family support policies?

Despite significant progress in recent decades, gender gaps remain in Canada. The labor force participation rate for all women stands at 82 percent, about 10 percentage points lower than the men’s rate. And internationally, although Canada’s female participation rate is above the OECD average of 72 percent (age 25–54), it is below that of all the Nordic countries, including Sweden at 88 percent (Figure 5).

Figure 5.
Figure 5.

OECD Economies: Female Labor Force Participation Rates

(Percent, ages 25-54)

Citation: IMF Working Papers 2017, 166; 10.5089/9781484309247.001.A001

Source: OECD Stat.

Stronger educational attainment among women has not resulted in a fully corresponding increase in their participation in the labor force. Although 35 percent of women have a university degree (up from 14 percent in 1990; see Table 5), compared with 29 percent of men, the labor force participation rate of these highly educated women remains at around 87 percent, below the 94 percent participation rate of men with the same education level.

Table 5.

Canada: Population and Participation by Education Achievement

(Ages 25–54 years old, percent)

article image
Source: Statistics Canada, CANSIM 282-0004; and authors’ calculations

All these data suggest an opportunity to fully tap underutilized female labor sources to solve the longer-term growth concern. But this means incentives for female integration in economic activity must improve in some areas.

Many studies suggest that economic incentives play an important role in encouraging women to participate in the labor market (for example, Jaumotte 2003, Tsounta 2006, and Thévenon and Solaz 2013). These studies highlight key labor market policies that influence female labor participation: (i) tax incentives for women to work; (ii) child-related leave entitlements; and (iii) child care services. Figure 8 presents the cross-country comparison of these policies.

Our assessments of work incentives in Canada are as follows:

  • Tax policies. For women who are “secondary earners” in households (often wives who earn less than their husbands), their decision to get a job, for example, depends on how much total additional tax the woman and her spouse are required to pay. In the late 1980s, the federal tax reform in Canada eliminated the elements of the tax code that created incentives similar to those provided by a system of joint taxation. Under the previous tax system the connection between a secondary earner’s effective marginal tax rate and her spouse’s marginal tax rate distorted the labor supply of women.9 In the 1990s, with an improved fiscal positon, the federal government actively implemented tax reforms, including tax cuts specifically focusing on families with children, further lowering tax wedges on secondary earners. This has contributed to reducing the tax wedge for the secondary earner from about 35 percent in the early 1990s to 32 percent today (Figure 6).10 These tax cuts proved successful in raising the participation of women in the workplace (Tsounta 2006; Crossley 2006; Bibbee, 2008). However, by international comparisons, the tax wedge in Canada is still relatively high, above the median of the advanced OECD economies, and thus there is potential to further reduce the tax wedge on secondary earners (Figure 8, top panel).

  • Maternity and parental leave. All OECD economies except the United States offer paid maternity and parental leaves of at least three months (Figure 8, second and third panels). In Canada, federal and provincial governments launched initiatives in the late 1990s to support parenting and early childhood development. One of the key reform measures was the lengthening of maternity and parental leaves from a maximum of 37 weeks to a maximum of 52 weeks in 2001. After having a baby, a mother can typically take 17 weeks leave from her job (the leave provisions vary across provinces, between 15 and 18 weeks), and thereafter, both parents have the option of taking leave for another 35 weeks. Thus, if a household decides that the mother would take all the parental leave herself, the total time off work can add to up to 52 weeks, compared with the advanced OECD economies’ median of 41 weeks. Public spending on maternity and parental leave in Canada11 is also higher than the median of these OECD economies.

  • Early childhood education and child care policies. Canada performs poorly in this area compared with other OECD economies. Despite the strengthening of child benefits in the late 1990s and of the national early learning and child care system in the 2000sas part of parenting policy reforms, (Figure 7) Canada still—on average across all provinces—still spends relatively little on early childhood education and child care compared with other advanced economies: only US$82 per child in 2015 (or 0.2 percent of GDP, including in cash and in kind), the lowest among OECD economies (Figure 8, bottom panel). 12

Figure 6.
Figure 6.

Female Participation Rates and Tax Wedges on Secondary Earners

(Percent, period average)

Citation: IMF Working Papers 2017, 166; 10.5089/9781484309247.001.A001

Sources: OECD Stat ‘Taxing Wedges”
Figure 7.
Figure 7.

Canada: Female Labor Participation Rates and Public Spending on Early Childhood Education and Care

Citation: IMF Working Papers 2017, 166; 10.5089/9781484309247.001.A001

Sources: OECD Stat, “Taxing Wedges”
Figure 8.
Figure 8.

Advanced OECD Economies: Selected Family Policy Indicators 1/

(Most recently available year)

Citation: IMF Working Papers 2017, 166; 10.5089/9781484309247.001.A001

Source: OECD Stat, “Family Database,” and “Tax Wedges,” and World Bank World Development Indicators1/ The sample is 20 advanced OECD economies. See Annex 1 for the definition of the indicators.

To sum up, Canada’s public family support policies are mixed in terms of incentives to work. Public spending on maternity and parental leave, as well as the length of paid maternity and parental leave, are generous in Canada compared with other OECD countries. However, despite past efforts, the estimated tax wedge for the secondary earner is higher than that of most OECD peers, and public spending on early education and child care falls well short of that in many advanced economies. In the following section, we review Canada’s public child care assistance programs and assess their affordability, which many studies describe as one of the highest hurdles for women to participate in the labor force (OECD, ILO, IMF, and World Bank 2015).

B. A case for enhancing public spending on child care

Governments’ child care and child support programs in Canada

In the Canadian federation, child care services are within provincial jurisdiction. However, like other programs under provincial jurisdiction such as health care, the federal government may play both a policy and funding role.13 Historically, the federal government has played a relatively limited role in child care.14 Measures have included transferring non-earmarked funds through the Canada Social Transfer which provinces may choose to use for child care. The federal government also allows for a tax deduction for employment-related child care expenses (the Child Care Expense Deduction).

The federal government provides the Canada Child Benefit Program.15 This program—intended to improve the financial situation of low- and medium-income families with children—adds up to a maximum annual benefit of Can$6,400 per child under age six and Can$5,400 per child through age 17. Families with incomes of less than Can$30,000 receive the maximum amount, and the benefit is reduced as family income increases.

At the provincial level, most provincial government have their own child care service provision and funding program. The programs vary in type—from simple “top-ups” to the equivalent of the federal’s Canada Child Benefit Program and various stand-alone child benefit and daycare financial assistance programs—and in generosity and scope. Each province has different objectives in designing early childhood education and child care programs. The financial support instruments—for example, fee subsidies to child care providers versus income subsidies to parents with children—and provincial policymakers use them vary considerably (Figure 9 and Annex II).

Panel Figure 9.
Panel Figure 9.

Canada: Child Care Assistance Programs by Province

Citation: IMF Working Papers 2017, 166; 10.5089/9781484309247.001.A001

Source: Macdonald D., and Friendly, M., 2016, “A Growing Concern, 2016 Child Care Fees in Canada’s Big Cities”, Friendly, M., B. Grady, L. Macdonald, and B. Forer, 2015, “Early Childhood Education and Care in Canada 2014;” and Canada Revenue Agency Child and Family Benefits Calculator.

Quebec’s program stands out. Of Canada’s 10 provinces, Quebec, Manitoba, and Prince Edward Island cap child care fees, and these provincial governments provide more substantial financial assistance than other provinces to child care providers to compensate part of their operating costs, including wages. Of these three, Quebec provides the most generous public support. For example, the basic parent fee for regulated child care centers is set at only Can$8.25 a day if the family’s income is Can$50,545 or less (as of end-2016).16 This is equivalent to about Can$171 a month, much lower than the fees of about Can$450-1,400 a month charged elsewhere in the country. We analyze below how child care fees and other variables affect work incentives. 17

Analytical framework

To evaluate how the tax and benefit system affects work incentives, we estimate the marginal effective benefit in each province. When deciding whether or not to get a job, one of the considerations that a couple with children makes is whether additional family income that the mother earns would pay off the costs of child care. While she stays at home, the couple need not send their children to a child care center. However, once the mother starts working, unless they have somebody (such as other relatives) to take care of their child or children free of charge, the couple must leave the children at a child care center and pay fees. The mother also needs to pay income taxes, and the family might lose some benefits as a result of its higher income. We define the marginal benefit index of work as the following:

Marginal Effective Benefit Index=(Household Net Income)D(Household Net Income)S(Household Gross Income)D(Household Gross Income)S

where D denotes the household consisting of two working parents (we assume here a married couple) and two children, and S denotes the household, consisting of a married couple and two children with only one parent working. Gross income is taxable income, while net income is calculated as:

Net Income=Gross IncomeIncome Taxes PaidChild Care Costs+Child Care Benefits

Child benefits for each province are calculated using Canada Revenue Agency’s Child and Family Benefits Calculator and income tax effects, using a calculator provided by PWC.18 This is done for three income categories of households with two parents and two children (ages two and three). The first income category is a low-income household, with each spouse earning Can$19,000 a year. If both spouses work, they earn Can$38,000 annually, equivalent to half of Canadian median family income (Can$78,870 in 2014). The second income category is a middle-income household, with each spouse earning Can$40,000 annually (if two spouses work, their total income is about median family income). The third income category is a high-income household, with each spouse earning Can$97,000 annually (top 10 percent in the distribution of household income). Data are for 2016.

Results

Figure 10 presents the estimates of marginal effective benefits in each province. A positive number indicates a net gain for the household (see Annex III for detailed calculations).

Figure 10.
Figure 10.

Marginal Effective Benefits, Today’s Policy

(Ratio)

Citation: IMF Working Papers 2017, 166; 10.5089/9781484309247.001.A001

Sources: Canada Revenue Agency, Child and Family Benefits Calculator; and authors’ calculations.
  • For low-income families, Quebec’s marginal effective benefit index is the highest, at 0.7. This implies that if a mother decides to work, total family gross income increases by $19,000, and after paying income taxes and child care costs and receiving child care benefits, the family can still retain a net income increase of almost $14,000. In contrast, in many other provinces, if a mother decides to work, her household’s marginal benefit would be negative (except in Manitoba, and Prince Edward Island). For example, in Ontario, the marginal benefit index is −0.6, indicating that the household’s economy clearly worsens if the mother enters the labor market. Ontario has lower income taxes than other provinces, but this doesn’t help much as its positive effect is more than offset by high child care fees. Indeed, the primary source of this difference arises from child care costs: in Quebec, due to generous public subsidies, the cost of child care is only about Can$3,900 annually, but in Ontario, the cost is much higher at around Can$29,000.19

  • For middle-income families, the marginal effective benefit index is positive for all provinces. However, in some provinces, it is around 0.3, implying that more than two-thirds of the additional income earned by the secondary earner goes toward child care costs, income taxes, and reduced subsidies. For Ontario and British Columbia, the net gain is only marginally above zero. Quebec has the highest marginal benefit index, followed by Manitoba.

  • For high-income families, the marginal effective benefit index is around 0.5. In the high-income case, the increase in income is more than sufficient to offset child care costs in all provinces.

Discussion

The universal Canada Child Benefit introduced in 2016 is generous and targeted to low- and middle-income families. However, it is universal in that it is not specifically targeted to employment, study, or the unemployed actively seeking employment. If one of the parents with children decides to stay at home to take care of their children, the couple is given more child benefits (because the family income is lowered), compared with a case when both parents work. In other words, the child benefit does not add an incentive to work or get job training.

If we assume that provision of child benefits is conditional on employment, study, or job training, and families with a “stay-at-home parent” are not entitled to receive the benefit, what would the marginal benefit index then look like? 20 Figure 12 shows that for low-income households the total economic benefit of a spouse entering the labor market would be positive for all provinces, although only marginally so for British Columbia and Ontario. For Quebec, the total benefit would actually be higher than the increase in gross income. The marginal benefit would also improve for middle- and high- income families in all provinces.

Figure 11.
Figure 11.

Female Labor Force Participation Rates in Quebec and Rest of Canada, Ages 25-54

(Percent)

Citation: IMF Working Papers 2017, 166; 10.5089/9781484309247.001.A001

Source: Stat Canada; and Haver Analytics
Figure 12.
Figure 12.

Marginal Effective Benefits, No Benefits for Families With a “Stay at home” Parent

Citation: IMF Working Papers 2017, 166; 10.5089/9781484309247.001.A001

Sources: Canada Revenue Agency, Child and Family Benefits Calculator; and authors’ calculations.

These illustrative calculations show that Quebec offers greater economic incentives for women to work than do other provinces. Quebec’s approach is unique in Canada in the sense that the provincial government focuses on a supply-side funding approach. Quebec authorities set the ceiling of the price of child care services to households and to compensate their operating costs, provide funding assistance to child care service providers. Several studies (e.g., OECD 2006 and UNICEF 2008) indicate that the supply-side funding approach could result in more uniform quality and superior coverage of childhood populations than parent subsidy approaches. In fact, since the implementation of the major child care program reform in Quebec in 1997–2000, female labor force participation has increased faster than in the rest of Canada (Figure 11), a positive economic outcome highlighted in many studies (see Baker, Gruber and Milligan 2008; Lefebvre and Merrigan 2008; and Fortin, Godbout, and St. Cerny 2012).

As an illustration, we present another scenario: a new federal government funding program to help provinces fund for child care service providers. The national average child care fee is estimated at about Can$865 per child a month. Suppose that to cut the fee by about 40 percent to Can$540, the government pays providers the difference of Can$325 (the total federal cost is estimated at Can$8 billion, equivalent to 0.4 percent of GDP). We assume that this program would be tied to employment. 21 Figure 13 suggests positive effects for all types of families, especially for low-income families.

Figure 13.
Figure 13.

Marginal Effective Benefits, Subsidized Child Care Fees

(Ratio)

Citation: IMF Working Papers 2017, 166; 10.5089/9781484309247.001.A001

Sources: Canada Revenue Agency, Child and Family Benefits Calculator; and authors’ calculations.

This policy is fiscally affordable. Approximately 150,000 women are stay-at-home mothers with high educational attainment living with a spouse or a partner. If they all enter the labor market and start paying taxes, that would raise GDP by 2 percentage points, which would in turn raise federal income tax revenues by about Can$8 billion, fully compensating the cost of the program.22 In other words, the program would be fully financed in a federal government perspective. 23 Importantly, to achieve the policy goal of enhancing work incentive, access to subsidized child care fees should be conditional on employment. 24

Figure 14 summarizes the marginal effective benefit under the three scenarios. Under the second scenario (“no benefits for families with a stay-at-home parent”), economic incentives for work would clearly improve, especially for low-income families and, to a lesser extent, middle-income families. This policy would also lead to fiscal savings but could be politically costly because “stay-at-home mother” families lose benefits. The third scenario (“subsidized child care fees”) would give more uniform incentives across provinces, would be naturally targeted at working couples (ensuring only working couples have access to subsidized fees), and would be politically less sensitive, although the incentives are smaller than in the second scenario.

Figure 14.
Figure 14.

Summary: Marginal Effective Benefits under Alternative Policies

(Ratio)

Citation: IMF Working Papers 2017, 166; 10.5089/9781484309247.001.A001

Sources: Canada Revenue Agency, Child and Family Benefits Calculator; and authors’ calculations.

C. Tackling longstanding cultural norms to promote gender equality and increase economic growth

Reform of child care assistance programs alone would not be sufficient to increase female labor force participation. The toughest question may be how to change social norms pertaining to the role of women. Longstanding cultural stigma often limits policy effectiveness. The low rate of fathers’ usage of parental leave is a case in point.

While parental leave benefits can be shared between the parents, most fathers do not make much use of the benefit after the birth or adoption of children, except in Quebec (which has earmarked father-only leave in addition to parental leave). The available survey data on employment insurance suggest that in 2014, among all insured Canadian mothers, 89 percent received maternity or parental leave payments, whereas only 27 percent of fathers claimed parental leave benefits (Figure 15). In contrast, in Quebec, the share of fathers who took parental leave was 78.3 percent, nearly three times the national average (the share of mothers was also the highest, at 94.5 percent).

Figure 15.
Figure 15.

Percentage of Mothers and Fathers who Take Parental Leave 1/

(Percent)

Citation: IMF Working Papers 2017, 166; 10.5089/9781484309247.001.A001

Source: Statistics Canada, Employment Insurance Coverage Survey, 20141/ As a share of mothers and fathers who have employment insurance. For mothers, Include maternity leave.

Several factors may explain why fathers are less likely to take leave outside of Quebec. Analyzing parental leave programs in European countries, Plantenga and Remery (2005) found five main determinants of take-up rates by fathers: payment level (financial impact); organizational and social culture (expected roles for men and women); program flexibility (when and how leave can be taken); work environment (employer attitude and perceived career advancement); and educational level of parents. Even in Canada, there is anecdotal evidence suggesting that stigma remains on fathers taking parental leave due to peer pressure and an unsupportive workplace (Paterson 2016). 25

Thus, social factors are important. To achieve gender equality, for example in terms of wages that would further encourage women to enter the labor market, a more equal distribution of parental leave could be considered. One option is to introduce parental leave exclusively for fathers, a “daddy quota.” Quebec and Nordic countries have implemented this policy, which seems effective in changing social norms toward a more gender-equal division of parenthood and work.

  • Quebec has its own parental benefits program, called the Quebec Parental Insurance Plan, which the Quebec government introduced in January 2006, replacing a federal program (the Canada Employment Insurance Program). Quebec’s program is more generous than the federal program, in terms of the eligibility criteria and the rate of income replacement. It also includes a five-week “daddy quota” of paternity leave. The introduction of the daddy quota has changed fathers’ behavior quite significantly in Quebec: the incidence of fathers taking paternity leave rose from 32 percent in 2005 to 56 percent in 2006 (Marshall 2008).

  • In Sweden, a “daddy quota” was first introduced in 1995 (Box 1). The Swedish Social Insurance Inspectorate (2012) found that the introduction of the measure had a significant impact on parents’ leave patterns, with mothers’ use of parental leave falling 26 days and fathers’ use increasing by 10 days during the first two years of a child’s life.

IV. Conclusions

The main finding of this paper is that female labor force participation among well educated women has a positive effect on labor productivity. This finding is not a surprise. As well understood in the growth literature, an economy with many highly skilled workers is likely to be much more productive. Our analysis shows that if the current gap of 7 percentage points between male and female labor force participation with high educational attainment were eliminated, the level of real GDP could be about 4 percent higher today. This brings about 0.3 million women with high educational attainment in the labor market26. Today approximately 0.3 million highly educated “stay at home” mothers have at least one child under 5 years, of which nearly 0.2 million are living in a partner or spouse relationship.27. Thus a key element of labor market policies to encourage women to participate in the labor market should be to help improve the work-life environment for these women with young children.

Canada fares well in comparison with other OECD countries on most family policy indicators. Public spending on early education and care, however, is low. If high child care costs mean it is not economically worthwhile for women to work, it will be difficult to realize women’s full potential in the work force. The latest policy initiatives to provide more affordable child care spaces are a positive step forward to encourage more women to participate in the labor market. However, we argue that to maximize the policy outcome given a budget constraint, provision of child care subsidies, including publicly funded child care spaces, should be better targeted to working parents, including those who study or take job training to improve their skills to obtain better jobs in the future.

The Swedish Parental Leave and Child Care System

Sweden’s female labor force participation rate and gender equality ranking—along with other Nordic countries—is among the highest in the world thanks to deliberate policy measures to improve gender equality.

Sweden has implemented two major policy measures to boost female labor force participation.

Paternity leave

Sweden implemented parental leave for fathers in 1974 and introduced one month of paternity leave—known as “daddy month”—under “use it or lose it” conditions in 1995. The length of paternity leave has gradually increased and was extended to 3 months in January 2016.

Paternity leave has increased fathers’ use of parental leave and stimulated more women to work, without decreasing male labor force participation. It has also dampened gender wage gaps (Johansson 2010).

Affordable child care

The supply of child care increased quickly in the 1970s and 1980s. This made it easier for mothers to work. In 2001–03 the federal government offered a large subsidy on child care costs to all municipalities in Sweden and set limits on maximum amounts for how much families would pay out of pocket for child care. No municipality rejected the offer.

The monthly limit is now Skr 1,36228 (just over US$150) a month for the first child; SKr 908 (about US$100) for the second child; SKr 454 (about US$50 for the third; and no cost for any additional children. This measure dramatically lowered household expenditures on child care. The main goal of the program is to encourage labor force participation. Eligibility for subsidized child care is linked to employment or studying, or if the child has special needs due to its family situation or physical, psychological, or other authorized reasons. Parents who are searching for and applying for jobs or on parental leave are given limited access to the program.

The program increased female labor participation, but only slightly because households’ child care fees were already fairly low thanks to high subsidies compared with other developed economies (Lundin and others 2007)—it had, however, a stronger effect on the male labor supply. Studies from other countries with much lower subsidies to start with show much stronger labor supply effects. Another reason for the small effect on women’s labor force participation might be that it was already high before the reform.

Furthermore, bold decisions are needed to change corporate cultures and shift social norms. A glass ceiling apparently remains. For example, Canada’s gender wage gap is well above the OECD average (the third highest among 20 advanced OECD economies),29 and progress in promoting women’s representation in senior management has also been slow (the third lowest), with women making up only one in four senior managers. This glass ceiling could discourage younger generations of women to pursue professional careers. Addressing the problem would require concerted efforts from all stakeholders, including both the private and public sector. One option is to introduce parental leave for fathers, which seems to have been effective in Quebec and Nordic countries.

Since the global financial crisis, countries around the world have been trying to increase GDP growth. Canada is no exception. Low oil prices and weak growth in Canada’s major trading partners have hampered its exports, and its aging population means the labor force is shrinking. Canada must speed up its labor productivity growth. Women are part of the solution: they can help offset the shrinking labor force, reverse the decline in labor productivity growth, boost medium-term growth potential, and raise the living standards of all Canadians.

Women Are Key for Future Growth: Evidence from Canada
Author: Bengt Petersson, Rodrigo Mariscal, and Kotaro Ishi
  • View in gallery

    Selected Advanced Economies: Female Labor Participation, 1976-2015

    (Percent, ages 25 to 54 years old)

  • View in gallery

    Canada: Female Labor Force Participation Rate and Labor Productivity

  • View in gallery

    Growth Decomposition

    (Period average, contributions to overall growth rates)

  • View in gallery

    Labor Productivity Growth and Female Labor Participation 1/

    (1990–2015 average)

  • View in gallery

    OECD Economies: Female Labor Force Participation Rates

    (Percent, ages 25-54)

  • View in gallery

    Female Participation Rates and Tax Wedges on Secondary Earners

    (Percent, period average)

  • View in gallery

    Canada: Female Labor Participation Rates and Public Spending on Early Childhood Education and Care

  • View in gallery

    Advanced OECD Economies: Selected Family Policy Indicators 1/

    (Most recently available year)

  • View in gallery

    Canada: Child Care Assistance Programs by Province

  • View in gallery

    Marginal Effective Benefits, Today’s Policy

    (Ratio)

  • View in gallery

    Female Labor Force Participation Rates in Quebec and Rest of Canada, Ages 25-54

    (Percent)

  • View in gallery

    Marginal Effective Benefits, No Benefits for Families With a “Stay at home” Parent

  • View in gallery

    Marginal Effective Benefits, Subsidized Child Care Fees

    (Ratio)

  • View in gallery

    Summary: Marginal Effective Benefits under Alternative Policies

    (Ratio)

  • View in gallery

    Percentage of Mothers and Fathers who Take Parental Leave 1/

    (Percent)