Maternity leave versus early childcare—What are the long-term consequences for children? Updated

Effects of maternity leave expansions

In 1990, paid and job-protected leave for mothers was extended by 12 months in Austria. The extension was unusually large and take-up almost universal. As is common for parental leave policies, eligibility for the extended benefits depends on a sharp cut-off date, allowing for a comparison of children born just before to those born just after the cut-off with a regression discontinuity design. Using rich administrative data, a 2020 study analyzes educational, labor market, and health outcomes of children in the long term (ages 17–23) [1]. While health of children is positively affected, the study finds no effects on educational (PISA scores and high track attendance) or labor market outcomes (employment and wages). There are no variations by socio-economic status (SES) or gender; but, interestingly, all outcomes are positively affected when the counterfactual mode of care is informal (mostly grandparents) rather than formal childcare.

Comparable evidence comes from a maternity leave expansion in 1977 that introduced paid maternity leave for the first time in Norway (paid leave increased from zero to four months and unpaid leave rose from three to 12 months) [2]. The study finds a reduction in the dropout rate from high school and a 5% increase in wages at age 30 among the affected children. These effects are stronger for children whose mothers had less than ten years of education. Importantly, maternal income remained the same while maternal time spent with the child increased. As in the study from Austria, the positive effects are identified in a setting when the counterfactual arrangements are largely informal such as friends, relatives, or unlicensed caregivers.

In 1984, the Danish government increased birth-related leave by almost 50%, from 14 to 20 weeks. A study compares outcomes of children born two months before and two months after this reform [3]. Results show no significant differences in reading scores, the probability of high school enrolment, or high school grade point average between these groups. The author discusses some reasons for the non-findings: (i) the counterfactual type of care to parental care in Denmark is high-quality childcare; (ii) less pressure on the childcare system may have meant better conditions for those attending it; (iii) parents make compensating investments—for instance, giving more of their time (by cutting down on leisure time, say) or effort (“quality time”) or even more purchased goods to a child who attends daycare, so that child outcomes are the same regardless of care type.

At the end of 2000, maternity leave in Canada was extended by about half a year in the first year of a child's life. While the extension had no measurable development effects on children in the first two years of their lives, a follow-up study examines effects on children at ages four and five using the Peabody Picture Vocabulary Test - Revised (PPVT-R) and Number Knowledge Test as well as non-cognitive indicators to assess the effects on both cognitive and behavioral outcomes [4]. The study finds statistically insignificant or even small negative effects on cognitive skills and no effects on behavioral measures. These findings stand in contrast to most findings from the literature and the authors argue that, as suggested by developmental psychology, the second six months of a child's life may be critical for some developmental milestones and separation from a mother at this age may be harmful.

This proposed explanation points toward the importance of analyzing maternity leave extensions at different child ages, ideally within the same country. In Germany, three major maternity leave policy changes were enacted in 1979, 1986, and 1992, which increased the length of leave from two to six months, from six to ten months and, later, from 18 to 36 months (increases from ten to 18 months took place in-between). Extensions within the first year but also at older ages can therefore be examined using the same data and identification strategy. A study evaluating these three reforms estimates difference-in-differences models [5]; it does not find significant positive effects on children's long-term outcomes for any of the reforms with respect to choice of school track, the probability of completing high school, wages, or years of employment. In fact, the third reform actually showed detrimental effects on children in terms of educational attainment. The study further brings evidence from the German Time Use Survey, indicating that when mothers work during a child's infancy they do not reduce the time spent caring for children; instead, they reduce their leisure and household activities.

Taken together, the evidence of maternity leave expansions points toward mixed effects when considering children's educational or labor market outcomes in the medium to long term, as hardly any beneficial effects in terms of child development are found after the first six months. The introduction of a short paid or unpaid leave, however, can improve child outcomes in both the short and long term. But longer expansions may even be harmful for child development, especially for children from disadvantaged backgrounds. When the counterfactual mode of care is informal and of lower quality, effects are more likely to be positive. Zero effects may arise because parents tend to take time away from other activities to spend more time with their children.

Effects of childcare expansions

A study from Denmark links daycare registers to educational registers to investigate the effects of enrolment in different types of non-parental childcare on academic achievement in the final school-leaving year [6]. It uses entire birth cohorts of ethnic Danish children, enrolled in either formal center-based daycare or the more informal family daycare at age two and exploits the right to access to preschool in some municipalities. In Denmark, both types of care are regulated with respect to safety, nutritional guidelines, playing facilities, health, and hygiene. However, while staff at centers hold a pedagogical degree (15–16 years of education), family daycare givers typically have either a vocational degree or high school or less education. Centre-based daycare improved grades in Danish language in the final year of compulsory school by 0.2 standard deviations, but had no effect on Mathematics scores. In terms of parent SES, the results show that effects were strongest among children of low SES backgrounds, but even children of high SES backgrounds benefited from attending formal center-based care rather than informal family daycare. Additionally, boys gained consistently more than girls.

A study from Norway uses childcare assignment lotteries to estimate the effects of enrolment at ages one to two on cognitive development [7]. Measured at age seven, the authors find that children who received an assignment offer achieved higher test scores in language and mathematics of 0.14 standard deviations and were less likely to score below a certain threshold. Children from lower SES backgrounds, measured by parental education or family income, benefit much more, but results are positive for all subgroups considered in the analysis. As in the study from Denmark, boys experience larger positive effects. An appealing feature of the study is that it can use the assignment lottery in an instrumental variable estimation as the authors have information on the exact age of childcare enrolment. Through this they calculate the effect of starting childcare one month earlier, which increases test scores by 0.04 of a standard deviation on average.

Due to early expansions and rich administrative data much of the evidence on universal childcare regimes originated from Scandinavian countries. Fortunately, the evidence base has expanded substantially in recent years. A German study analyses the effects of an expansion of early childcare (zero to two years) on children's motor and socio-emotional skills [8]. Exploiting differential timing in expansion by school districts, the authors use a marginal treatment effects (MTE) framework to assess how outcomes at mandatory medical assessments before entering primary school were affected. The study finds that boys benefit more than girls, leading them to catch up in their language, motor, and socio-emotional development. Immigrant children also catch up in terms of language skills and children from lower-educated families benefit in their socio-emotional development. The findings highlight that a large set of skills ought to be considered when evaluating the efficacy of a given program or policy, as different groups may be affected differently. The study does not find any subgroups for which negative effects are identified, and early childcare in this setting has an equalizing effect.

Many studies document more positive effects for disadvantaged children; an often-proposed mechanism is the worse counterfactual environment at home, though the precise mechanisms commonly remain a black box and are largely suggestive. A study from Japan analyzing an expansion of childcare slots for two- and three-year-olds sheds light on this concept, as the authors not only observe child outcomes, but also information on parenting activities and indicators of parental well-being [9]. Childcare enrolment reduces inattention and hyperactivity among children of lower-educated mothers. As children from higher-educated mothers are not affected, this leads to a complete elimination of gaps by maternal education, which are prevalent when children are not enrolled. In a second step, the study documents effects on parents’ outcomes. Parenting of higher-educated mothers is not affected by childcare, but for lower-educated ones an index of parenting quality increased by 0.501 standard deviations, which strongly reduces the gap in parenting quality by education. A driver of better parenting quality could be lower stress and higher subjective well-being, where effects are again stronger for lower-educated mothers. These findings are in line with research from Germany analyzing the effects of childcare usage on parenting activities using time-use data [10]. It finds that while mothers’ time with their child is strongly reduced, the effect on time spent engaging in parenting activities is quite small, as mothers compensate for the reduced overall time they spend with their children when they are attending formal childcare with more direct engagement at other hours of the day.

It appears that childcare often has positive effects for children from disadvantaged backgrounds for a range of cognitive and non-cognitive outcomes. At the same time, effects are more moderate for high-SES children, who are likely to be in more formal, higher-quality arrangements when not attending childcare.

Universal childcare does not always produce positive outcomes

While many studies report positive or null findings for effects of childcare on children's development, there is also evidence that, at least for some subgroups, both cognitive and non-cognitive outcomes can be harmed by childcare. A study from Denmark finds that child behavior at age seven (measured along the Strength and Difficulties Questionnaire, resulting from experiencing high-quality center-based daycare at age three is neutral compared to maternal care, while children experiencing informal daycare display worse behavior than those in maternal care, which indicates that informal, lower-quality arrangements might be more likely to negatively affect child development [11]. This result is driven by boys and children of low-educated mothers.

Several studies have examined the large-scale formal childcare expansion in Québec which led many mothers of preschool children to enter the labor market. A study from 2019 considers not only short-term outcomes (before entering school) but also analyzes longer-term effects to check whether effects identified at young ages in earlier studies persist up to age 20 [12]. The authors find that negative short-term effects on non-cognitive skills are highly persistent and that children from cohorts in Québec with larger formal childcare available score substantially worse in terms of anxiety and aggression, and also experience reductions in prosocial behavior during teenage years. This effect was driven mainly by boys, for whom they saw increased criminal behavior and worsened self-reported health, which increased with the “dosage” received—that is, the number of years of exposure to the childcare program. The study, however, does not report heterogeneous long-term effects by parental SES. An important dimension for interpretation is that the childcare expansion in Québec had strong effects on maternal labor supply. This contrasts with the Scandinavian expansions where most mothers were already in the labor market or with Germany where labor market effects were more muted. The hours spent in childcare also increased substantially for children in Québec compared to the rest of Canada. Both aspects may make negative effects more likely.

A study from Bologna, Italy, uses admission thresholds to analyze the effects of daycare attendance at ages zero to two for a relatively affluent population [13]. The authors find that one additional month in daycare reduces children's IQ at ages eight to 14 by 0.047 standard deviations and also negatively impacts the Big Five, a set of basic personality traits which are relatively stable throughout adulthood. In contrast to the findings from Québec, the negative effects are stronger for girls. The study argues that the reduction in one-to-one interactions in childcare can be especially problematic for children from advantaged backgrounds where those interactions in counterfactual care modes are arguably of higher quality.

The two latter studies examine childcare in settings where it is of low quality compared to evidence from many other European countries. In Québec, childcare was assessed to be of only “minimal quality” in 60% of childcare centers; for children from low-income families especially, educational activities were largely insufficient. In Bologna the adult-to-child ratio was very low in comparison to studies identifying more beneficial effects and the authors emphasize that attention should be paid to this quality dimension as more direct interactions are especially important for younger children. The evidence thus seems to suggest that the quality of the childcare program being offered matters substantially and that inadequate quality is more likely to produce negative effects on child development.

Not a question of “if” but of “how and when”

For a long time, the main question for parents regarding childcare was whether their child should enter childcare or not. In light of historically lower female—especially maternal—labor force participation rates and more conservative gender norms, childcare was often viewed as an option for parents—potentially for the benefit of children—but not essential, and childcare provision was low in many countries. This has shifted dramatically in recent decades. Figure 1 shows the share of respondents agreeing to the statement that a preschool child suffers with a working mother. Albeit on different levels, the share agreeing has declined dramatically since the early 1990s across all countries. To enable maternal employment and to reconcile work and family life for both parents, childcare provision is now often a fundamental requirement. Thus, many parents now face different key questions in this regard: (i) For how many hours should childcare be used? (ii) What is the ideal starting age? And (iii) What constitutes adequate childcare quality?

Evidence on the right intensity, that is, the number of hours per day, of childcare and its effect on child development is still relatively scarce. Naturally, some hours are necessary in order to reap the beneficial effects on learning outcomes and to get a chance to bond with the caregiver, be it the mother or a non-family caregiver. On the other hand, spending many hours in out-of-home care has been shown to have certain negative consequences for children, such as higher stress levels. Indeed, evidence from the German expansion of full-day childcare suggests negative effects of long hours on socio-emotional development. Similarly, few studies have investigated the optimal starting age of childcare. The Norwegian study discussed above finds significant gains associated with starting childcare at a younger age [7]. Yet, longer-term evidence on this question and for settings outside of Scandinavia are still.

What quality is necessary to provide a beneficial environment for children (of all backgrounds)? While quality is a multi-dimensional concept and many potential indicators (e.g. child-to-staff ratios, skill requirements, pedagogical concepts, process quality, etc.) are difficult to compare across countries, studies that produce negative outcomes for children often score low on at least some of those dimensions. Some studies from Denmark and Norway analyze the effects of structural quality of care (e.g. child-to-staff ratios, caregiver qualifications) on child development and identify a positive relationship. Evidence on process quality is even scarcer, but a large-scale randomized controlled trial from Denmark finds that a low-cost intervention targeted at caregiver–child interactions resulted in pronounced positive effects on language and mathematics skills. So, the scant evidence existing to date points toward the importance of both structural and process quality.

Quality indicators vary strongly across countries, for example, child-to-staff ratios range from 3.0 in Iceland to 13.9 in Hungary. To date it has not been well-established what thresholds are required to produce positive child outcomes, and clearly these are also likely to differ across settings and child background. Yet it seems evident that rapid childcare expansions at the expense of quality (as in Québec) are likely to be a bad deal, as this is more likely to hinder child development.

Enrolment gaps by family background

Many studies find stronger positive effects of childcare usage for children from disadvantaged family backgrounds, which is mostly measured by maternal/parental education, but sometimes also by immigrant background or family income. Figure 2 shows enrolment rates for children aged zero to two years by maternal tertiary education for OECD countries (the relationship by household income looks comparable). With a few exceptions, large enrolment gaps by education exist in many countries and, as indicated by the range plots, most of these differences are significant. Interestingly there is no clear link between the average level of enrolment and the gap by maternal education, which implies that an expansion of childcare provision would not necessarily be accompanied by a reduction in enrolment differences by education.

Because children who benefit more from childcare are commonly less likely to be enrolled, in several countries many early childcare systems are universal only in name leading to substantial interest in understanding why such enrolment gaps by family background exist. A 2020 study from Germany considers demand- and supply-side factors to analyze why children from lower-educated parents or parents with a migrant background are less likely to be enrolled in early childcare [14]. Accordingly, differences by demand are small and explain, if anything, only a small share of the gap. An expansion of provision and a fee reduction strongly reduce the enrolment gap by parental education. However, these measures do not close the gap by migration background. Both quality concerns and discrimination could potentially play a role to explain this gap. Other work shows experimentally that behavioral barriers play a stronger role for disadvantaged families as they may be insufficiently informed about childcare and the application system; providing better information can thus help to reduce the gap.

Drivers of enrolment gaps and solutions to close them are strongly context-specific, but policymakers can help to ensure that childcare systems get closer to being truly universal. Information provision, an impartial application system, and transparent selection criteria in the case of over-demand, and a progressive fee structure (or an abolishment of fees) can all be measures that are likely to result in lower enrolment gaps by family background, ensuring that those children who benefit the most have at least an equal chance to participate in the formal early education system.