Does Child Care Quantity Cause Externalizing Problems? An International Comparison Using Seven Longitudinal Studies

• In Event: Non-Parental Child Care: An International Perspective on Behavioral and Academic Outcomes

Abstract

The question of whether high quantities of non-parental child care cause externalizing behavior problems in early childhood is still controversial in the field. Correlational studies suggest that children who spend more time in non-parental care, particularly in center-based care, tend to exhibit more externalizing behavior problems (Jacob, 2009; NICHD ECCRN, 2003; see Huston et al., 2015 for review). However, studies that have moved beyond covariate adjustment to account for unobserved selection have failed to replicate such findings (Dearing, Zachrisson, & Nærde, 2015; McCartney et al., 2010). Indeed, a review of recent research using quasi-experimental approaches suggests that the association between hours in non-parental care and externalizing problems is not robust (Dearing & Zachrisson, 2017).
In addition to the lack of empirical consensus, there is a shortage of research using samples from outside the United States. To date, a large proportion of the current evidence still relies on studies in the U.S., in particular, on the NICHD Study of Early Child Care and Youth Development (NICHD SECCYD). Moreover, the scarcity of cross-country evidence makes it hard to understand the extent to which findings from prior studies do or do not generalize internationally. Replication studies using samples both from inside and outside the U.S. would allow us to study variability in childcare effects across diverse sociopolitical contexts (Duncan, Engel, Claessens, & Dowsett, 2014).
In this study, we employ data from seven longitudinal studies across five countries (i.e., Canada, Germany, Netherlands, Norway, and the United States) to test whether within-child changes in hours in non-parental care predict within-child changes in externalizing behavior problems in toddlers and preschoolers. The within-child models allows us to control for potential bias caused by endogenous between-child heterogeneity. The multi-site replication approach allows us to aggregate and compare results across sociopolitical contexts.
In all studies, child care quantity was measured by the number of hours per week the child attended early child care at each time point. We used teacher reports of externalizing behavior problems in datasets that had this information available, and maternal reports otherwise. We also included time-varying covariates, such as single parenthood, new siblings in the family, and parental employment to further mitigate concerns about confounding. Finally, we included maternal education and family income as moderators to test potential interactions with family background.
We found no evidence that increases in time spent in early child care predicted increases in externalizing problems in any of the seven studies (Table 1). Even after using multiple imputation for missing data, the coefficients for hours in care were all statistically indistinguishable from zero. Moreover, results from a fixed-effects meta-analysis for the seven studies, show that, on average, hours spent in child care do not have a detectable association with externalizing problems (Figure 1). In addition, neither maternal education nor family income significantly moderated this null association. The implications of these findings for developmental science and policy will be discussed with a focus on the value of international replications.

Authors

• Catalina Rey-Guerra, Boston College
  Presenting Author

• Henrik D. Zachrisson, University of Oslo
  Non-Presenting Author

• Eric Dearing, Boston College
  Non-Presenting Author

• Daniel Berry, University of Minnesota - Twin Cities
  Non-Presenting Author

• Susanne Kruger, Deutsches Jugendinstitut
  Non-Presenting Author

• Margaret Burchinal, University of Virginia - Charlottesville
  Non-Presenting Author

• Ane Nærde, The Norwegian Center for Child Behavioral Development
  Non-Presenting Author

• Thomas van Huizen, Utrecht University
  Non-Presenting Author

• Sylvana M Côté, Université de Montréal (UdeM)
  Non-Presenting Author