Economic and Social Reports
Patterns of participation in early learning and child care among families with potential socioeconomic disadvantages in Canada

Abstract

Various studies have shown that children from socioeconomically disadvantaged families are more likely to have poorer outcomes than children from more advantaged families and that such gaps could be reduced by participating in early learning and child care (ELCC). However, the patterns of ELCC participation, such as rates of participation, types of care arrangements and care hours, may differ between families with and without socioeconomic disadvantages, and such differences may limit the role of ELCC in improving children’s well-being and families’ opportunities for education or employment. Using the 2019 Survey on Early Learning and Child Care Arrangements, a nationally representative survey that provides detailed information on child care for children aged 0 to 5 years linked to additional socioeconomic information, this study examines the patterns of ELCC participation among families with potential socioeconomic disadvantages in Canada. Low-income families were about 20% less likely than families not in low income to use non-parental child care. After other sociodemographic characteristics were controlled for, the income-based gap in child care use shrank, but persisted. More than one-third of all parents who were using child care reported having difficulties finding a child care arrangement. Finding child care available in the local community and finding affordable child care were the two most frequently reported difficulties. Because of the difficulties in finding a child care arrangement, low-income parents and lone parents were more than twice as likely to postpone or discontinue their schooling or training compared with parents not in low income and parents in two-parent families.

Keywords: child care; low socioeconomic status; family income; parental education; lone parent

Authors

Lan Wei is with the Centre for Income and Socioeconomic Well-being Statistics at Statistics Canada. Leanne Findlay is with the Health Analysis Division, Analytical Studies and Modelling Branch, at Statistics Canada. Rubab Arim is with the Social Analysis and Modelling Division, Analytical Studies and Modelling Branch, at Statistics Canada.

Introduction

Various studies have shown a positive relationship between participating in early learning and child care (ELCC) and child developmental outcomes and parents’ participation in the labour market (Havnes and Mogstad 2011; Lefebvre and Merrigan 2008; Romano, Kohen and Findlay 2010). More specifically, studies have found that the potential benefits of ELCC on child development are greater for children from families with socioeconomic disadvantages (Ruhm and Waldfogel 2012; Van Huizen and Plantenga 2018). However, the patterns of ELCC participation, such as rates of participation, types of care arrangements and care hours, may differ between families with and without socioeconomic disadvantages, and such differences may limit the role of ELCC in children’s well-being and families’ opportunities for education or employment.

However, the research that has examined patterns of early child care use among families with socioeconomic disadvantages in Canada is limited, with little information based on recent data. In addition, little is known about the extent to which families with socioeconomic disadvantages have difficulties in finding child care, the types of difficulties they face and the consequences of such difficulties. Thus, the purpose of this study is to examine child care use among three types of families with potential socioeconomic disadvantages: low-income families, families with low-educated parents and lone-parent families. The focus is on these socioeconomic conditions because previous research has shown that children from low-income families, children with low-educated parents and children from lone-parent families are more likely to have poorer health, behavioural and developmental outcomes (Currie and Stabile 2003; Cutler and Lleras-Muney 2010; McLanahan, Tach and Schneider 2013), and therefore are more likely to experience positive outcomes associated with participating in ELCC (OECD 2016).

Use of non-parental child care among children living in families with potential socioeconomic disadvantages

Various studies have shown that use of non-parental child care is positively associated with household income in Canada. Bushnik (2006) showed that, from 1994/1995 to 2002/2003, the higher the household income, the more likely the child was in non-parental child care. Sinha (2014) found that, in 2011, 34% of parents with an annual income below $40,000 used non-parental child care for their children aged 4 and younger, compared with 65% of parents with an annual income of at least $100,000. Most recently, using the 2016 Census of Population, Guèvremont (2019) showed that families with higher income were more likely to report paying for child care in order to work. For instance, 21% of low-income families^Note reported paying for child care in order to work, compared with 39% of moderate-income families, 56% of middle-income families and 66% of high-income families.

The positive relationship between child care use and household income has also been observed in Quebec (Guèvremont 2019; Kohen et al. 2008), where licensed child care has been provided to some families at a low price since 1997. Kohen et al. (2008) demonstrated that, even after the 1997 child care program reform, Quebec saw greater increases in the use of licensed child care for both low-income families and high-income families, compared with the rest of Canada, but the income-based gaps in child care use were persistent.

The positive association between family income and participation in child care has also been observed in many other Organisation for Economic Co-operation and Development (OECD) countries. As indicated by the OECD (2016), for children younger than 3 years of age, participation rates in formal ELCC increased with family income and with maternal education in most OECD countries. However, in many OECD countries, particularly countries where the ELCC services are publicly operated or directly subsidized, the gaps in child care use between family income levels largely shrank or disappeared once maternal employment was taken into account.

In addition to family income, non-parental child care use has also been associated with family structure and parents’ work status (Bushnik 2006). In 2011, the lowest rates of child care use among children aged 4 and younger were for children in two-parent families where only one parent worked for pay (42%), whereas 58% of lone-parent families where the parent worked for pay used non-parental child care (Sinha 2014). The highest rates of child care use were among two-parent families where both parents worked for pay (71%). Overall, lone-parent families and families where both parents worked for pay or studied were more likely to rely on non-parental child care than two-parent families where one parent worked for pay or studied (Bushnik 2006). This reflects the strong association between parental employment and child care decisions (Morrissey 2008).

Types of child care arrangements among children living in families with potential socioeconomic disadvantages

Beyond the use of non-parental child care, several Canadian studies have examined the relationship between the types of child care arrangements and household income (Bushnik 2006; Cleveland et al. 2008; Sinha 2014). For example, parents with a household income below $40,000 were most likely to use a daycare centre over any other type of child care arrangement (Sinha 2014). Home daycare was the most common child care arrangement for parents with a household income between $40,000 and $100,000, while daycare centres and private arrangements were the leading choices for parents with a household income above $100,000. In terms of licensed care, Cleveland et al. (2008) suggested that parents in the lowest income quintile were equally as likely to use types of care that were licensed (which may be more expensive) than those with middle income. This may reflect the greater availability of child care subsidies for low-income families.

Types of child care arrangements also vary by family structure. Among children attending child care, children in lone-parent households have been shown to be much more likely than children in two-parent households to be in a daycare centre (40% vs. 28%; Bushnik 2006). This may reflect the absence of a second parent who could care for the child and the fact that lone-parent households, on average, have lower household income than two-parent households and therefore are more likely to be eligible for daycare centre subsidies.^Note Similar findings were observed by Cleveland et al. (2008) and Cleveland and Forer (2010) using more recent data.

Child care hours among children living in families with potential socioeconomic disadvantages

Finally, the number of hours spent in non-parental child care is associated with family structure and parents’ working status. According to Bushnik (2006), children with a lone parent who worked or studied spent 4 additional hours per week in their main care arrangement compared with children with two parents who worked or studied, and 10 additional hours per week compared with children with two parents where one parent worked or studied. This may be because children in lone-parent households were much more likely than children in two-parent households to be in a daycare centre, and that children whose main care arrangement was a daycare centre spent more time there per week (Bushnik 2006). These findings are consistent with those of Cleveland and Forer (2010), which showed that lone-mother families were more likely than all families^Note to use considerable amounts (more than 30 hours) of child care per week.

In summary, the existing Canadian evidence has shown associations between patterns of child care use and family socioeconomic characteristics. However, more recent information on child care use and types of care is necessary given the changing landscape in child care (e.g., the 2017 Multilateral Early Learning and Child Care Framework). Furthermore, there is little information available to describe the reasons why parents are or are not using care and the difficulties and consequences experienced by parents in finding child care. Therefore, the purpose of the current study is to use data from the 2019 Survey on Early Learning and Child Care Arrangements (SELCCA) to explore child care use patterns among families with potential socioeconomic disadvantages in Canada.

Data and methods

Data source and definitions

The SELCCA was designed to assess non-parental child care use, parent and child characteristics, and characteristics of care of children younger than 6 across Canada. The survey was delivered through an electronic questionnaire or through computer-assisted telephone interviewing in both official languages from mid-January 2019 until mid-February 2019, with randomly sampled people knowledgeable about the child care arrangements (mainly a parent) for a child who lived in the same household in all provinces and territories. Participation was voluntary.

The target population was children aged 0 to 5, but the information was obtained from a parent, guardian or person who was knowledgeable about the child’s child care arrangements (or lack thereof). The respondent was female in 91% of cases. Children living in institutions or on reserves were excluded from the target population. The response rate was 55% in the provinces and 40% in the territories, yielding a sample size of 7,548 children. The final SELCCA sample represented approximately 1.3 million children across Canada.

Families with potential socioeconomic disadvantages: This study focuses on three types of families with potential socioeconomic disadvantages: low-income families, families with low-educated parents and lone-parent families. Family income information was taken from the linked T1 Family File. Following previous studies (Bushnik 2006; Guèvremont  2019), families with income below the after-tax low-income measure cutoff were considered as low-income families. Parents and guardians were asked about their highest education level completed, which was grouped into three categories: a high school diploma or equivalent or less, more than high school but less than a university degree, and a bachelor’s degree or higher.^Note A family was considered as a two-parent family if, in the linked T1 Family File, a family record number was available for both the recipient of the Canada Child Benefit (CCB) and their spouse. If the family record number was available only for the CCB recipient or for their spouse, the family was considered as a lone-parent family.

Types of child care: Child care use was determined by asking parents to report one or more types of child care arrangements they had usually used in the past three months from a list of seven possible responses: daycare centre, preschool or centre de la petite enfance; care by a relative other than a parent; care by a non-relative in the child’s home; family child care home; before or after school program; other child care arrangement; or that they do not use child care. When parents reported multiple types of care, a subsequent question asked which child care arrangement the parent or guardian considered to be the main one. Parents and guardians were also asked whether their main child care arrangement was licensed.^Note

Time in care: For the current study, two elements of time spent in care were explored: the number of hours per week and the use of child care in the evenings and on the weekends. Parents and guardians were asked, “In the past three months, how many hours per week did your child usually spend in child care arrangements?” In addition, for each type of care arrangement, parents and guardians reported whether the child attended in the evening, on the weekends, both evenings and weekends, or neither evenings nor weekends. A dichotomous summary variable was created to identify whether the child attended any arrangement in the evening or on the weekends.

Difficulties finding care: Whether or not parents and guardians had difficulty finding care and the reasons for and consequences of those difficulties were also of interest. Those respondents who were using child care indicated whether or not they had experienced any difficulties finding child care. Those who had experienced difficulties were asked to indicate whether those difficulties related to finding the following: care that was available in their community, affordable child care, care that fit their work or study schedule, the quality of care they desire, licensed care, a qualified care provider, care that could accommodate more than one child, care that meets their child’s special needs, or other difficulties.

Consequences of difficulties in finding child care: Respondents who reported using child care and having difficulties finding a child care arrangement were asked to indicate the consequences based on the following options: postponing their return to work; deciding to work at home; working fewer hours than they would have; using multiple or temporary arrangements; splitting care with a spouse, partner or relative; postponing or discontinuing schooling or training; changing their work schedule; other; and none of the above.

Reasons for not using child care: Parents and guardians who were not using child care were asked to indicate the main reasons, which were grouped into the following categories: unemployed; maternity, paternity or parental leave; one of the parents has decided to stay home with the child; shortage of places or waiting list; prefer to adjust work or study schedules to accommodate care needs; the cost of child care is too high; the child is in kindergarten; and other reasons.^Note

Analytical strategy

Descriptive analyses were performed to describe rates of participation in child care among Canadian families and characteristics of child care use. Three types of low-socioeconomic-status (SES) families were of interest: low-income families, families with parents with a high school diploma or less education, and lone-parent families. Other sociodemographic variables included child’s age group (younger than 1; 1 to 3; 4 to 5), parental working status (working, looking for work or at school; parental leave or at home or volunteering; unable to work; other), whether the child was from an immigrant family, whether the child had an Indigenous identity, whether the child lived in rural area, and the province where the child lived.^Note Means and proportions are reported, as appropriate. Ordinary least squares regressions were used to examine the relationship between the patterns of using child care (use of care, licensed care and hours in care) and low SES, after controlling for other sociodemographic characteristics that might affect the patterns of using child care. Survey sampling weights were applied to render the analyses representative of Canadian children aged 0 to 5 living in the provinces or territories. Bootstrap weights were also applied when testing for significant differences (p < 0.05) to account for the complex survey design.

Results

Table 1 shows non-parental child care use by family characteristics. It shows that 45% of children from low-income families participated in child care, which was significantly lower than 64% of those from families not in low income. Use of child care was also significantly lower among parents with a high school diploma or less education (47%), compared with parents with more than high school but less than a university degree (62%) and those with a bachelor’s degree or higher education (65%). However, there was no difference in child care use between lone-parent and two-parent families (around 61%). For all three types of families with potential socioeconomic disadvantages, child care participation rates were higher in Quebec, where licensed child care may be provided at a lower price compared with the rest of Canada (Appendix Table A.1). However, the gaps in child care use between family income and parental education levels were found in both Quebec and the rest of Canada.



Use of licensed child care as the main child care arrangement also increased with family income and parental education levels. Table 2 shows that 32% of low-income families reported using licensed child care as the main child care arrangement, compared with 46% of families not in low income. Furthermore, 31% of parents with a high school diploma or less education reported using licensed child care as the main child care arrangement, compared with 44% of parents with more than high school but less than a university degree and 47% of parents with a bachelor’s degree or higher education. The income- and education-based gaps in use of licensed child care were observed in both Quebec and the rest of Canada (Appendix Table A.2).



Compared with parents with a high school diploma or less, parents with a bachelor’s degree or higher were more likely to use a daycare centre, preschool or centre de la petite enfance (56% vs. 48%; see Table 3); before or after school program (11% vs. 6%); and care by a non-relative in the child’s home (7% vs. 3%). They were less likely to use other child care arrangements (2% vs. 5%). In contrast, type of child care arrangement generally did not vary by family low-income status or family structure, with a few exceptions. Parents in low income were less likely to use a family child care home compared with parents not in low income (15% vs. 21%, not shown). Compared with two-parent families, lone-parent families were less likely to use a family child care home (17% vs. 21%), but more likely to use other child care arrangements (5% vs. 3%, not shown).



The average number of hours per week spent in child care was similar between low-income families and families not in low income (around 30 hours per week; see Table 4). However, low-income parents were more likely to use child care in the evenings or on weekends compared with parents not in low income (24% vs. 15%, Table 5). Children of parents with a high school diploma or less education spent, on average, 28 hours per week in child care, while children of parents with more than high school education, including those with a bachelor’s degree or higher education, spent, on average, about 31 hours in child care (Table 4). The use of child care in the evenings or on weekends did not vary significantly by parental education level (Table 5). Average child care hours were slightly higher among lone-parent families (32 hours per week) compared with two-parent families (30 hours per week), but the difference between the two groups was not statistically significant. In contrast, lone-parent families were twice as likely to use child care in the evenings or on weekends compared with two-parent families (27% vs. 14%, Table 5).

 

In terms of difficulties finding a child care arrangement, the findings of the survey showed that more than one-third of all parents who were using child care reported having difficulties finding a child care arrangement (Table 6), although this was very similar between low-income families and non-low-income families. Parents with a bachelor’s degree or higher level of education using care were more likely to report having difficulties finding a child care arrangement compared with parents with a high school diploma or less (40% vs. 33%). It is possible that more highly educated parents may be more selective about characteristics such as location, quality and caregiver characteristics.^Note About 39% of lone-parent families reported having difficulties finding child care, compared with 35% of two-parent families, but the difference was not statistically significant. It should be noted that parents who were not using child care at all were not asked the questions about difficulties, and it is possible that the reason they were not using care related to difficulties obtaining care (see below).



Among all parents who were using care and had difficulties finding child care, the top four reported difficulties were finding child care available in their community (53%), finding affordable child care (48%), finding child care that fits their work or study schedule (38%) and finding the quality of care that they desired (37%) (Table 7). The trends in the types of difficulties were consistent and did not vary by family low-income status, parental education level or family structure.



As shown in Table 8, the top three reported consequences of having difficulties finding child care were requirements to change their work schedule (41%), use of multiple or temporary care arrangements (36%) and working fewer hours (34%). These consequences differed by family low-income status. Compared with parents not in low income, low-income parents were more likely to postpone or discontinue schooling or training (33% vs. 8%) and were more likely to postpone their return to work (38% vs. 26%).

The consequences of having difficulties were mostly similar between lone-parent and two-parent families, except that lone-parent families who experienced difficulties were significantly more likely to postpone return to work (37% vs. 25%) and postpone or discontinue schooling or training (20% vs. 8%), and less likely to use multiple or temporary arrangements (28% vs 37%) compared with two-parent families.



Regarding reasons for not using child care at all, several notable differences were observed between low-income families and those not in low income (Table 9). For example, 28% of parents in low income were unemployed, 11% were at home on parental leave, and 11% indicated that the child was in kindergarten. . By comparison, 13% of parents not in low income were unemployed, 27% were at home on parental leave, and 18% indicated that the child was in kindergarten.



Table 10 presents the associations between the patterns of child care use and family SES, controlling for other sociodemographic characteristics, including child age, immigrant status and parental working status. The results suggest that low-income families were 12% less likely to use child care and 8% less likely to use licensed child care as the main child care arrangement, compared with families not in low income. Compared with parents with a high school diploma or less education, parents with more than high school education but less than a university degree were 9% more likely to use child care and 7% more likely to use licensed child care as the main child care arrangement. Parents with a bachelor’s degree or higher education were 13% more likely to use child care and use licensed child care as the main child care arrangement. In comparison with the descriptive statistics shown in Tables 1 and 2, when the other sociodemographic factors were controlled for, the gaps in the use of (licensed) child care between family income and parental education levels shrank but remained statistically significant. Compared with lone-parent families, two-parent families were 8% less likely to use child care and 6% less likely to use licensed child care as the main child care arrangement, and their children spent about four hours fewer per week in child care. Average hours per week in child care did not differ statistically significantly by family low-income status and family structure.



Discussion

This study provides comprehensive descriptive statistics on child care use based on socioeconomic characteristics of families. Several notable differences in terms of the patterns of using child care were observed between low-SES families and families not in low SES.

First, similar to previous research (Bushnik 2006; Guèvremont 2019, OECD 2016), low-income families and families with low-educated parents were less likely to use non-parental child care; they were also less likely to use licensed care as their main arrangement. Even when other sociodemographic characteristics, including child age and parental working status, were controlled for, the differences in child care use across family income and parental education levels shrank but persisted. This may be because low-income parents and parents with low education are more likely to have additional barriers to accessing child care, such as high child care costs and non-standard working schedules. Another potential explanation is that more highly educated parents value different aspects of child care compared with parents with lower educational achievement (Johansen et al. 1996). However, not using child care because parents were unemployed was more likely among low-income families; not using child care because parents were on parental leave was less likely among low income families.. The results may suggest that low-income parents do not necessarily have different desires for using non-parental care but instead are more likely to have barriers to access.

Second, use of care in the evening and on weekends varied by family income and family structure, while hours using child care varied by parental education level. Low-income parents and lone parents were more likely to use child care in the evenings or on the weekends. This may reflect the fact that low-income parents may be more likely to work non-standard hours (which include regular evening and night shifts, weekend hours, rotating shifts, and irregular or on-call hours) (Lero et al. 2019).

Third, more than one-third of parents who used child care reported having difficulties finding a child care arrangement, and the trends in the types of difficulties were generally similar between families with and without potential socioeconomic disadvantages. However, the consequences of the difficulties were associated with family SES. One notable finding is that low-income parents and lone parents were more likely to postpone or discontinue their schooling or training or their return to work because of the difficulties in finding a child care arrangement. This might be due to the nature or type of employment or because low-income parents and lone parents may have more difficulties in adjusting their working hours and schedules (Foley and Schwartz 2002) and are therefore more likely to sacrifice their return or their schooling and training opportunities.

Despite the important contribution that the findings make in terms of understanding child care use among families with potential socioeconomic disadvantages in Canada, several limitations of this study need to be mentioned. First, the gaps in child care use across family income and parental education levels were shown to withstand the controls for sociodemographic characteristics, including child’s age group and parental working status, but additional (unobserved) factors that correlate with family income and parental education and affect child care use might exist (e.g., work schedule, neighbourhood, care availability, caregiver characteristics). Second, it is difficult to collect comprehensive information regarding the quality of child care services from a parent-reported survey and thus this  important aspect of child care has not been investigated in this study. If more nuanced data on the quality of child care (e.g., staff-to-child ratio in daycare centres, caregivers’ education levels) were available, it would be of interest for future research to examine whether and how the quality of child care varies by sociodemographic characteristics.

Conclusion

In conclusion, this study examines child care use, types of child care arrangements, child care hours, reasons for not using child care and difficulties in finding child care, as well as the consequences of any difficulties, among low-income families, families with low-educated parents and lone-parent families in Canada. Child care use was positively associated with family income and parental education levels. Low-income parents and lone parents were more likely to use child care in the evenings or on the weekends. Children with more highly educated parents spent more time in non-parental care. For all types of families, finding child care available in the local community and finding affordable child care were the two most frequently reported difficulties in finding child care. Because of the difficulties in finding a child care arrangement, low-income parents and lone parents were more likely to postpone or discontinue their schooling or training or their return to work. Future research may consider exploring variation in child care quality by sociodemographic characteristics.

Appendix – Tables

 

References

Bushnik, T. 2006. Child Care in Canada. Children and Youth Research Paper Series. Statistics Canada Catalogue no. 89-599-M. Ottawa: Statistics Canada.  

Cleveland, G., and B. Forer. 2010. Child Care Use and Child Development in Immigrant, Lone Mother, Rural, and Official Language Minority Families in Canada. Ottawa: Human Resources and Skills Development Canada.

Cleveland, G., B. Forer, D. Hyatt, C. Japel, and M. Krashinsky. 2008. “New evidence about child care in Canada: Use patterns, affordability and quality.” IRPP Choices 14 (12).

Currie, J., and M. Stabile. 2003. “Socioeconomic status and child health: Why is the relationship stronger for older children?” American Economic Review 93 (5): 1813–1823.

Cutler, D.M., and A. Lleras-Muney. 2010. “Understanding differences in health behaviors by education.” Journal of Health Economics 29 (1): 1–28.

Foley, K., and S. Schwartz. 2002. Leaving Welfare for a Job: How Did SSP Affect the Kinds of Jobs Welfare Recipients Were Willing to Accept? SRDC Working Paper Series 02-03. Ottawa: Social Research and Demonstration Corporation.

Guèvremont, A. (2019). Early Learning and Child Care—Child Care Usage and Costs Among Canadian Subpopulations. Ottawa: Statistics Canada. Unpublished report for Employment and Social Development Canada.

Havnes, T., and M. Mogstad. 2011. “No child left behind: Subsidized child care and children’s long-run outcomes.” American Economic Journal: Economic Policy 3 (2): 97–129.

Hill, S., and L. Findlay. Comparing Parents Reported Use of Licensed Child Care to Confirmed Licensed Child Care Locations Across Canada. Ottawa: Statistics Canada. Forthcoming.

Johansen, A., A. Leibowitz, and L.J. Waite. 1996. “The importance of child-care characteristics to choice of care.” Journal of Marriage and the Family 58 (3): 759–772.

Kohen, D., V.S. Dahinten, S. Khan, and C. Hertzman. 2008. “Child care in Quebec.” Canadian Journal of Public Health 99 (6): 451–455.

Lefebvre, P., and P. Merrigan. 2008. “Child-care policy and the labor supply of mothers with young children: A natural experiment from Canada.” Journal of Labor Economics 26 (3): 519–548.

Lero, D.S., S. Prentice, M. Friendly, B. Richardson, and L. Fraser. 2019. Non-standard Work and Child Care in Canada: A Challenge for Parents, Policy Makers, and Child Care Provision. Guelph: Childcare Resource and Research Unit and the University of Guelph.

McLanahan, S., L. Tach, and D. Schneider. 2013. “The causal effects of Father Absence.” Annual Review of Sociology 39 (1): 399–427.

Morrissey, T.W. 2008. “Familial factors associated with the use of multiple child-care arrangements.” Journal of Marriage and Family 70 (2): 549–563.

OECD (Organisation for Economic Co-operation and Development). 2016. Who Uses Childcare? Background Brief on Inequalities in the Use of Formal Early Childhood Education and Care (ECEC) Among Very Young Children. Paris: Organisation for Economic Co-operation and Development.

Romano, E., D. Kohen, and L. Findlay. 2010. “Associations among child care, family, and behavior outcomes in a nation-wide sample of preschool-aged children.” International Journal of Behavioral Development 34 (5): 1–14.

Ruhm, C., and J. Waldfogel. 2012. “Long-term effects of early childhood care and education.” Nordic Economic Policy Review (1): 23–51.

Sinha, M. 2014. Child Care in Canada. Spotlight on Canadians: Results from the General Social Survey. Statistics Canada Catalogue no. 89-652-X. Ottawa: Statistics Canada.

Van Huizen, T., and J. Plantenga. 2018. “Do children benefit from universal early childhood education and care? A meta-analysis of evidence from natural experiments.” Economics of Education Review (66): 206–222.