Principal investigator: Viviane Pires Ribeiro
Paper Title: Universal child care, maternal labor supply, and family well-being
Authors: Michael Baker, Jonathan Gruber and Kevin Milligan
Location of the Intervention: Quebec/Canada
Sample Size: Not specified
Main theme: Health Insurance
Main Variable of Interest: Childcare
Type of InterventionAnalysis of the impact of policy change on family well-being.
Methodology: Difference-in-differences model
A comprehensive assessment of publicly funded childcare requires answers to questions that have not been clearly and comprehensively addressed in the literature. In light of this gap, Baker, Gruber, and Milligan (2008) analyze the introduction of highly subsidized and universally accessible childcare in Quebec, addressing its impact on childcare utilization, maternal labor supply, and family well-being. The study's results suggest caution for other provinces in Canada considering adopting the Quebec childcare model. It is possible that the findings are short-term rather than long-term effects, but more evidence is needed before the program is adopted elsewhere.
Evaluation Context
It is noticeable that, in the last generation, there has been a substantial increase in the proportion of mothers working in the paid workforce in North America. In Canada, the employment rate of mothers with at least one child under 6 years old rose from 31% in 1976 to 67% in 2004. However, this trend was not offset by a decline in the proportion of working fathers, necessitating greater use of paid and unpaid childcare. In that same country, the percentages of children under 6 years of age being cared for by someone other than their parents increased from 40% in 1994–95 to 51% in 2002–3.
Currently, larger subsidies for childcare are being directed toward low-income families in Canada; middle- and high-income families receive modest tax subsidies for daycare or preschool. But interest is growing in more universal subsidies for early childhood care/education, as offered in many European nations. In 1997, for example, the government of the province of Quebec, Canada, introduced a new set of family policies, including major changes to government subsidies for childcare. The centerpiece was a childcare program to provide regulated spaces for all children aged 0 to 4 in Quebec, with a parental contribution of $5,00 per day. Children were eligible regardless of whether their parents were working or not. The program was implemented gradually, beginning with 4-year-olds in September 1997. Subsequently, 3-year-olds became eligible in September 1998, 2-year-olds in 1999, and 0- and 1-year-olds in September 2000.
Universal access to early childhood care has several advantages. Publicly funded systems can provide more equitable access to quality childcare and lead to a greater supply of labor among secondary workers. At the same time, public systems require extensive public funding, which comes at the cost of higher taxes and therefore reduced economic efficiency. Furthermore, it is possible that publicly provided childcare will “drive out” private childcare provision without a net increase in childcare use or labor supply to the market. Finally, and most importantly, there is an ambiguous effect on child outcomes depending on the implications of time spent in childcare versus time spent with parents.
Intervention Details
A comprehensive evaluation of publicly funded childcare requires answers to three questions. First, does public funding affect the quality or quantity of care provided, or does it merely lead to the substitution of one form of care for another? Second, if childcare use increases, what is the size of the associated increase in parental labor force participation, and what does this suggest about the net cost of the policy? Third, what effect does any change in childcare (and associated increases in labor force participation) have on child and family outcomes? Thus, Baker, Gruber, and Milligan (2008) argue that previous studies on childcare policies offer, at best, incomplete answers to one or two of these questions, and that there has been no evaluation of a large-scale public intervention that addresses all three questions.
Therefore, the study conducted by the authors provides such an assessment using a major policy innovation in the Canadian province of Quebec in the late 1990s. Quebec's Family Policy began in 1997 with the extension of full-time kindergarten to all 5-year-olds and the provision of childcare at a price of US$5,00 per day for all 4-year-olds. This US$5,00 per day policy was extended to all 3-year-olds in 1998, all 2-year-olds in 1999, and finally, all children under 2 years old in 2000. This drastic policy change in one of Canada's largest provinces provides a research design to assess the effect of publicly funded childcare.
The analysis is based on the National Longitudinal Survey of Children and Youth (NLSCY). The NLSCY is a continuous panel dataset that tracks the progress of a large, nationally representative sample of Canadian children. In this context, the authors measured the impact of the policy change on mothers' labor supply, childcare utilization, and child and parent outcomes. Because simultaneous program reforms make inferences difficult for single mothers, the authors focused on married women and their children.
Methodology Details
The NLSCY dataset provides information on a rich array of childcare options, as well as tracking parent and teacher assessments of children's development, test scores, and class rankings. The sample size averages approximately 2000 children at each age per year. For this reason, Baker, Gruber, and Milligan (2008) used the weights provided by the survey in all presented results. The primary sample consists of children aged 0–4 years, although for some robustness checks, children aged 8–11 years were also considered. Children aged 5 years were excluded to isolate the effect of the childcare program from the effect of kindergarten. The main sample restriction was to include only children from two-parent families.
The authors estimated difference-in-differences models by comparing outcomes in Quebec and the rest of Canada at the time of the reform. The “pre-reform” period was termed NLSCY waves 1 and 2, covering the period from 1994–95 to 1996–97. The “post-reform” period is NLSCY waves 4 and 5, from 2000–2001 and 2002–3.
The procedure produces intention-to-treat effects because the authors estimated the effects of the reduced form for the children, and not just for the families who chose childcare. This has the advantage of potentially capturing the full impact of the program on both subsidized and unsubsidized care arrangements, as well as any peer externalities. Furthermore, due to some uncertainty regarding responses to childcare questions in the survey, the approach allows some flexibility in deciding by what likelihood of treatment the intention-to-treat effects should be scaled to arrive at the treatment impact on the treated.
Results
The results found by Baker, Gruber, and Milligan (2008) indicate that the introduction of universal childcare subsidies in Quebec led to a significant increase in childcare use. The proportion of children aged 0 to 4 years in childcare increased by 14 percentage points in Quebec compared to the rest of the country. This increase in childcare was associated with a considerable increase in the employment of women in two-parent households. Employment grew by 7,7 percentage points in Quebec. The difference between the increase in employment and the increase in childcare use mainly reflects the reduction in the use of informal childcare arrangements, or the “exclusion” of informal childcare by this new subsidized childcare. In part, as a result of this large exclusion, the taxes generated by the new supply of maternal labor fall short of covering the costs of the increased childcare subsidies.
Consistent and robust evidence of negative effects of the policy change on a variety of child outcomes, parenting, and parental outcomes was also found. Child outcomes are worse for a variety of measures reported by parents, such as anxiety, aggression, motor and social skills, child health status, and illness. Measures of parenting and family function were also negatively affected, and there is evidence of deterioration in parental health and a reduction in the quality of the parental relationship. To our knowledge, the effects on parenting and family outcomes have not been previously investigated. Therefore, the results suggest that, in this case, more access to childcare was bad for both children and parents in the dimensions captured in the data. There are, however, interpretations of these findings that are more benign. While some of these explanations seem inconsistent with the data, it is not possible to rule out the possibility that the findings represent a short-term adjustment to childcare, rather than a long-term negative impact.
Lessons in Public Policy
The study by Baker, Gruber, and Milligan (2008) provides the first comprehensive analysis of a universal subsidized childcare program, tracking its impact from childcare use to employment and ultimately to child and parent outcomes. Strong evidence was found of a shift toward new childcare use, although approximately one-third of newly reported use appears to come from previously employed women with informal arrangements. The impact of job availability is strongly significant. Compelling evidence was also reported that child outcomes have worsened since the program was introduced. Suggestive evidence was found that the families studied became more stressed with the introduction of the program. This manifests in increased aggression and anxiety for children; more hostile and less consistent parenting for adults; and poorer adult mental health.
The authors emphasize that such results are subject to a range of interpretations, highlighting the importance of future work in this area. More importantly, it is unclear whether the negative outcomes for the child are merely short-term problems or foreshadow long-term effects. Furthermore, the study raises the puzzle of why families would benefit from a policy that leads to worse child outcomes, poorer parenting, and poorer parental outcomes. It is possible that the unmeasured benefits of higher family income offset these costs. Alternatively, it is possible that families learn they are not better off under this new regime, and that the ultimate use of subsidized childcare may decline. Again, tracking the long-term evolution of these policy effects will be central to a comprehensive analysis of the program's well-being.
Despite these caveats, the results are particularly relevant to ongoing political debates in the United States and Canada. The estimates do not address the effectiveness of targeted childcare subsidies for groups such as single mothers, but these are generally not the focus of the current debate. Instead, it is the universal expansion of early childhood care/education that attracts the most heated arguments. More directly, the study's evidence advises caution in Canada for other provinces considering adopting Quebec's childcare model. It is possible that the findings are short-term rather than long-term effects, but more evidence is needed before the program is adopted elsewhere.
References
BAKER, Michael; GRUBER, Jonathan; MILLIGAN, Kevin. Universal child care, maternal labor supply, and family well-being. Journal of Political Economy, v. 116, no. 4, p. 709-745, 2008.
