The authors have declared that no competing interests exist.
This study examined early external childcare in relation to development from age 7 to 20. A Swiss sample was used (
According to the notion of developmental pathways, early life experiences can shape lives for a long time to come, affecting not only initial development but also long-term socio-behavioral trajectories [
One often asked and studied question is how external childcare is related to child development. An increasing number of studies has examined this question but, though contributing greatly to our knowledge, most offer only a momentary glimpse of development because the underlying data are limited to snapshots of pathways. Specifically, the vast majority of studies could only examine short-term relations of external childcare with development. Few were able to assess long-term relations, because the necessary data are often not available. Thus, little is known about the long-term relation of external childcare with behavior.
Two other aspects that past studies have often overlooked relate to the multi-faceted nature of both external childcare and socio-behavioral development. For example, many, though not all, studies have focused primarily on daycare centers, though many children attend other types of (sometimes informal) external care such as by external family members. In addition, because many studies on external childcare have focused on externalizing behavior, other behaviors remain understudied. These include sub-types of externalizing behavior, including aggressive and non-aggressive, but also prosocial behavior, internalizing problems, and manifestations of problem behavior in adolescence and beyond, notably delinquency and substance use.
In this paper, we addressed these issues by using longitudinal data from Switzerland that capture behavioral development from school entry to early adulthood. Our data include a range of external childcare arrangements as well as behavioral aspects, and enable us to examine the relation of external childcare before Kindergarten with child and youth development from age 7 to 20. We are not aware of other studies that included a similarly wide range of socio-behavioral outcomes, types of childcare, and, particularly, length of follow-up. In order to account for the possibility that our results are spurious and explained by third factors, we controlled for various confounding factors. Nevertheless, like other studies on the same topic, our study is observational and we therefore cannot claim that the remaining relations are more than correlational.
We focused on the amount and type of external childcare but not on its quality, which is another aspect that has been related to child development but was not included in our data [
Past literature reviews on the link between the quantity of external childcare and child development have concluded that external childcare is related to increased problem behavior. For example, Belsky [
Regarding the type of childcare, most prior research in the field has focused on a particular type of external childcare, namely on care in daycare centers, whereas not many studies have explicitly examined the relation between other types of external childcare and problem behavior (for exceptions, see [
Several potential explanations have been offered for the found relations. Though early discussions questioned whether external childcare reduced safe child-parent attachment [
The existing literature has focused more on externalizing behavior than on internalizing problems. Studies that have examined the link between external childcare and internalizing problems using various methods (including both traditional regression-based methods and more robust quasi-experimental techniques) have yielded mixed results. Some studies found some evidence that more time in a daycare center was related to more internalizing problems [
Another question that is still open is to what extent the relation between external childcare and behavior is causal. Particularly, researchers in the field have argued that “too little attention to internal and external validity may be biasing much of the scientific and public thinking on the topic” [
The studies that have used these methods have produced inconsistent results as well. In their review of these studies, Dearing and Zachrisson [
Another way in which studies have tried to circumvent selection bias is by analyzing sudden policy changes that affect a certain part of a country but not another. An example is a study in Quebec, Canada, where researchers have examined the consequences of a policy change that involved the provision of a universal subsidized child care program. It was found that the increased use of external childcare due to this policy change was associated with more anxiety, more physical aggression and opposition, and lower standardized motor and social development in early childhood, although there were no associations with hyperactivity-inattention or separation anxiety [
It is difficult to arrive at an overall conclusion regarding the causality of the link, which is evidenced by the fact that in one prior review, the authors were not very confident that the relation between childcare quantity and problem behavior is causal [
The way in which early experiences shape trajectories is not just dependent on the experience itself and the way it affects subsequent events but also a function of the larger environment in which children develop [
The question of how long relations (in either direction) continue to exist over the life course remains largely unanswered. Although it can generally be expected that more recent life experiences play a larger role than more distant ones and that the relation between early external childcare and development decreases as children get older, life course perspectives also suggest that early differences can exacerbate later in life. For example, the notion of developmental cascades posits that children’s exposure to early risks may contribute to other risks with age [
In the current study, our goal was to examine the long-term relation of various forms of external childcare with social behavior. We examined a range of externalizing behaviors, prosocial behavior, internalizing problems, delinquency, and substance use.
Researchers have argued that international replication has been underutilized as a scientific tool to examine the link between external childcare and behavior and that it is important to grow the international research base so that ultimately, individual dimensions of economic and sociopolitical contexts can be isolated [
Compared to other European countries (though less so to other countries), Switzerland has been rated relatively poorly on indicators of national family-friendly policies (e.g., paid leave for mothers and fathers), ranking in the bottom three along with Cyprus and Greece [
Little is known about the relation between external childcare and development in Switzerland. To our knowledge, four other studies exist. Three found that center-based care was related to more externalizing but not internalizing problems [
Data were drawn from the Zurich Project on the Social Development from Childhood into Adulthood (z-proso), a combined longitudinal and intervention study [
We used data from the youths, their parents, and their teachers at ages 7 through 20. Data collection followed local data protection regulations. Active parental consent was obtained before the data collection in grade 1 and again before the data collection in grade 5. The youths gave their active informed consent starting grade 7. The parents could opt their children out of the data collections in grades 7 and 9.
The parents received shopping vouchers the equivalent of US$30 for their participation. At ages 13, 15, 17, and 20 the youths received an incentive worth the equivalent of US$30, US$50, US$60, and US$75 at each respective wave in exchange for their participation. The size of the incentives was based on wage levels for Swiss youths. For the teachers, participation in the first three years was compulsory. Starting in grade 5, they received a book voucher worth the equivalent of US$50 when they completed more than six questionnaires.
Computer-assisted parent interviews, ceased after age 11, took place in the respondent’s home. Since the mother tongue of around half of the parents was not German, contact letters, telephone recruitment, informed consent, and interviews were made available in nine additional languages. Teachers completed a questionnaire and returned it by mail. Since the frequency of interactions between teachers and the youths significantly reduced after age 15, we analyzed teacher data until that age. For the children, the first three data collections were done in the form of computer-assisted personal interviews conducted by trained interviewers at school (approx. 45 minutes). Starting in grade 5, the youths completed a written questionnaire of approximately 90 minutes in a school setting. At age 20, they completed a computer-assisted self-interview in a lab setting.
We defined external childcare as care outside of the child’s home. Thus, care by live-in nannies was excluded. Similar to prior studies (e.g., [
We acknowledge the possibility that retrospective self-reported data can introduce bias. Two papers have examined the validity of the current EHC. The first examined its criterion-related validity. It was concluded that risk factors measured on the EHC were correlated with behavior outcomes in the expected direction, that the size-order and relative importance of early risk factors were in line with the previous literature, that longer exposure was associated with an added risk, and that the likelihood of problematic outcomes was related to cumulative contextual risk [
For all quarters prior to completion of the EHC, respondents estimated how many days per week a child received external childcare by external family members, acquaintances or neighbors, daycare mothers, daycare centers, and playgroups. Playgroups and daycare centers are similar in the sense that both are group-based, the difference being that playgroups often do not allow children below the age of 3 and usually offer care for a couple of hours up to half a day and for a limited number of days per week, for example only two or three. Different from a previous analysis [
Outcome variables were measured at ages 7 through 20.
Social behavior was measured using the Social Behavior Questionnaire (SBQ). The SBQ has shown satisfactory reliability and validity in both the z-proso [
Responses from the parents and teachers were recorded on a 5-point Likert scale. An easily understandable picture-based yes/no format similar to the Dominique Interactif [
We measured delinquency and substance use in five ways. First, from age 13, the youths reported on their past-year prevalence of 14 types of delinquency (e.g., theft, drug dealing, vandalism, extortion, robbery, assault). We computed a variety score [
Second, four self-reported items measured the past-year consumption of tobacco, beer or wine, liquors, and cannabis. Answers were given on a 5-point scale from 1 (“never”) to 5 (“daily”). We again computed a variety score. This variable was available from age 13.
Third, we included a broader measure to analyze past-year self-reported deviance at an earlier age (from age 11). Besides delinquency (9 types) and substance use, this variety score included peer aggression (teasing, stealing/destroying possessions, physical violence, and rejection; [
Fourth, the teachers answered seven items on the youths’ deviant behavior, including truancy, assault with injury, carrying a knife or weapon, using threats to get something, smoking cigarettes, drinking alcohol, and taking illegal drugs in the past 6 months. We constructed a variety score.
Finally, official delinquency records between ages 10 and 17 were obtained from the Youth Justice Authorities. Over 97% of youths who participated in the study at age 17 provided active consent. We computed a binary variable that indicated whether or not participants had a record and a count measure that assessed the number of trials for which youths had been registered (capped at 2).
We included three types of control variables, all identified a priori, based on prior research, and measured at age 7. The first set concerned characteristics that were known or present at birth, namely the child’s sex (‘1’ signifying males;
The second set was derived from the EHC and included parental separation (after the child’s birth;
The third set included variables pertaining to the child’s family situation at wave 1 (age 7). Though including these could result in conservative estimates of the associations between childcare and behavior, it is similar to practice in prior studies (e.g., [
We used growth curve analyses in MPlus [
In most cases, we used linear models. For binary variables, we estimated logistic models. For self- and teacher-reported delinquency and substance use, we used censored models. MPlus does not provide absolute fit statistics for the latter models and therefore, only the Bayesian Information Criterion (BIC) and the Akaike Information Criterion (AIC) are displayed. Where necessary, non-significant variances and corresponding covariances of the slopes were restricted to zero.
The growth curve models resulted in two types of estimates: relations between external care and the initial level of development at the first available time-point (i.e., the intercept), which was often at age 7, and relations between external care and the rate of change at which children developed after this initial level (i.e., the slope). Time-points for the slope factors were fixed between 0 and 1 to reflect the distance between the measurements. Thus, the slope coefficients represent the overall rate of change from the first to the last time-point included. To test the associations between external childcare and child development at different time-points, additional analyses were performed in which we centered on each time-point available. This is recommended in quadratic and cubic models to provide a more informative analysis [
Due to the nature of our sample and data, we could not perform robustness checks to check the sensitivity of our findings against selection bias. For example, change-on-change fixed effects models were not applicable in our case because we analyzed the relation between one prior predictor (external childcare) and a range of later outcomes. Neither could we use sibling fixed effects because with the exception of very rare cases, our data did not include siblings. In addition, propensity score matching requires the matching variables to be measured prior to the main predictor, which was not possible in our data.
We decided against multiple testing and instead guarded against over-interpretation of isolated findings by interpreting our results cautiously, examining overall and consistent patterns in the results rather than isolated findings. Due to space concerns, we refer to
The final sample for all analyses reported in this paper consisted of all children whose parents participated in the EHC (
Because our data were observational, we could not address causality. For example, issues of reverse causality, with parents enrolling their children into external care due to their behavior, may play a role. We performed exploratory analyses to address this issue. Results, suggesting but not proving that parents did not enroll their children into external care due to their problem behavior, are reported in
In total, 22% of children received external childcare from family members before Kindergarten, 3% from acquaintances or neighbors, 12% from a daycare mother, 32% in a daycare center, and 22% in a playgroup. Overall, 67% attended some type of external childcare before Kindergarten.
Children generally spent more time in external childcare as they got older (
The main issue we examined was the association between external childcare and social behavior. Results for the growth curve models are shown by informant in
Unstandardized coefficients from growth curve models.
Aggression | Non-Aggressive Externalizing | ADHD Symptoms | Anxiety and Depression | Prosocial Behavior | |||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Informant | Self | Self | Parent | Tea-cher | Self | Parent | Tea-cher | Self | Parent | Tea-cher | Self | Parent | Tea-cher | Self | Self | Parent | Tea-cher |
|
|||||||||||||||||
|
|||||||||||||||||
Family | 0.00 | -0.02 | 0.00 | -0.00 | 0.00 | 0.00 | -0.05 | -0.01 | 0.03 | 0.01 | 0.00 | 0.02 | -0.01 | -0.01 | 0.04 | ||
Family sq. | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | |
Acquaintances | 0.01 | 0.02 | -0.02 | 0.04 | 0.00 | -0.02 | 0.06 | 0.20 | 0.10 | -0.00 | 0.03 | -0.03 | 0.01 | 0.11 | 0.02 | 0.07 | |
Acquaintances sq. | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | |
Daycare mother | -0.00 | -0.01 | 0.02 | 0.02 | 0.00 | 0.02 | 0.03 | 0.04 | 0.02 | 0.06 | 0.05 | 0.01 | -0.02 | 0.01 | 0.03 | -0.06 | |
Daycare mother sq. | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | |
Daycare center | -0.02 | 0.02 | 0.00 | 0.03 | 0.02 | 0.02 | -0.01 | 0.00 | 0.00 | 0.01 | -0.01 | ||||||
Playgroup | -0.00 | -0.02 | -0.01 | -0.12 | -0.01 | 0.04 | 0.06 | 0.05 | -0.08 | 0.03 | 0.03 | 0.01 | 0.01 | -0.05 | |||
Playgroup sq. | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | |||||||
Family | -0.00 | 0.12 | -0.02 | -0.16 | -0.00 | -0.02 | 0.13 | -0.12 | -0.00 | 0.01 | 0.03 | -0.02 | 0.02 | -0.01 | 0.12 | 0.00 | 0.05 |
Family sq. | -- | -- | -- | -- | -- | -- | -0.08 | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- |
Acquaintances | 0.00 | -0.18 | -0.11 | -0.06 | 0.02 | -0.09 | -0.13 | -0.45 | -0.06 | 0.00 | -0.81 | 0.28 | -0.53 | ||||
Acquaintances sq. | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | |
Daycare mother | 0.01 | 0.18 | -0.03 | 0.31 | -0.01 | 0.01 | 0.10 | -0.08 | -0.02 | -0.11 | -0.03 | 0.03 | 0.05 | -0.01 | -0.20 | -0.08 | 0.25 |
Daycare mother sq. | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | 0.05 | -- |
Daycare center | -0.16 | -0.03 | -0.00 | -0.02 | 0.11 | 0.01 | -0.28 | -0.02 | 0.01 | 0.03 | 0.01 | 0.13 | -0.03 | 0.31 | |||
Playgroup | -0.02 | -0.62 | 0.10 | 0.26 | 0.00 | -0.44 | -0.31 | -0.20 | -0.08 | 0.31 | 0.01 | 0.00 | 0.19 | 0.01 | 0.22 | -0.76 | |
Playgroup sq. | -- | -- | -- | -0.80 | -- | 0.51 | -- | -- | -- | -- | -- | -- | -0.17 | -- | -- | -- | 0.45 |
Family | -- | -0.28 | 0.01 | 0.35 | -- | 0.01 | -0.13 | 0.12 | -- | -0.12 | -0.03 | -- | -- | -- | -0.31 | -0.01 | -0.04 |
Family sq. | -- | -- | -- | -- | -- | -- | 0.08 | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- |
Acquaintances | -- | 0.28 | 0.09 | -0.42 | -- | 0.08 | 0.09 | -- | 0.45 | -- | -- | -- | 1.30 | -0.23 | 1.25 | ||
Daycare mother | -- | -0.39 | 0.04 | -0.77 | -- | -0.03 | -0.41 | 0.06 | -- | -0.21 | 0.02 | -- | -- | -- | 0.55 | 0.18 | -0.38 |
Daycare mother sq. | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -0.07 | -- |
Daycare center | -- | 0.44 | 0.04 | -- | -0.01 | -0.13 | -- | 0.39 | 0.02 | -- | -- | -- | -0.41 | 0.03 | -0.74 | ||
Playgroup | -- | 1.46 | -0.06 | 0.07 | -- | 0.26 | 0.83 | 0.08 | -- | -1.09 | -0.02 | -- | -- | -- | -0.92 | -0.21 | 2.31 |
Playgroup sq. | -- | -- | -- | 1.32 | -- | -0.39 | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -1.14 |
Family | -- | 0.18 | -- | -0.14 | -- | -- | 0.08 | -- | -- | 0.15 | -- | -- | -- | -- | 0.17 | -- | -0.05 |
Family sq. | -- | -- | -- | -- | -- | -- | -0.02 | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- |
Acquaintances | -- | -0.12 | -- | 0.45 | -- | -- | -0.05 | -- | -- | -0.03 | -- | -- | -- | -- | -0.62 | -- | -0.73 |
Daycare mother | -- | 0.23 | -- | 0.45 | -- | -- | 0.28 | -- | -- | 0.27 | -- | -- | -- | -- | -0.35 | -- | 0.14 |
Daycare center | -- | -0.29 | -- | -- | -- | -- | -- | -0.18 | -- | -- | -- | -- | 0.27 | -- | 0.43 | ||
Playgroup | -- | -0.86 | -- | -0.29 | -- | -- | -0.60 | -- | -- | 0.80 | -- | -- | -- | -- | 0.73 | -- | -1.65 |
Playgroup sq. | -- | -- | -- | -0.68 | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | 0.83 |
χ2-Value | 35.60 | 59.66 | 52.15 | 164.21 | 37.39 | 41.92 | 150.33 | 52.45 | 38.05 | 123.54 | 37.83 | 55.48 | 185.70 | 25.83 | 47.83 | 57.06 | 174.33 |
χ2 df | 20 | 24 | 23 | 100 | 20 | 21 | 100 | 23 | 21 | 96 | 44 | 20 | 145 | 19 | 24 | 24 | 100 |
CFI | 0.98 | 0.98 | 0.99 | 0.98 | 0.98 | 0.99 | 0.98 | 0.97 | 0.99 | 0.99 | 1.00 | 0.96 | 0.98 | 0.99 | 0.98 | 0.98 | 0.98 |
TLI | 0.93 | 0.90 | 0.96 | 0.96 | 0.93 | 0.97 | 0.96 | 0.91 | 0.97 | 0.99 | 1.01 | 0.89 | 0.98 | 0.96 | 0.93 | 0.94 | 0.95 |
RMSEA Estimate | 0.02 | 0.04 | 0.03 | 0.02 | 0.03 | 0.03 | 0.02 | 0.03 | 0.03 | 0.02 | 0.00 | 0.04 | 0.02 | 0.02 | 0.03 | 0.03 | 0.03 |
SRMR | 0.01 | 0.01 | 0.01 | 0.02 | 0.01 | 0.01 | 0.02 | 0.01 | 0.01 | 0.01 | 0.01 | 0.01 | 0.02 | 0.01 | 0.01 | 0.01 | 0.01 |
BIC | 35836.53 | 45128.54 | 41953.37 | 52165.21 | 36172.85 | 41206.06 | 52697.66 | 42264.50 | 45928.93 | 57625.20 | 49107.11 | 43094.43 | 56182.06 | 29778.80 | 42109.18 | 42655.22 | 51818.22 |
AIC | 34533.30 | 43595.33 | 40542.82 | 50458.23 | 34869.62 | 39657.52 | 50990.69 | 40853.95 | 44508.16 | 56046.00 | 47676.11 | 41791.21 | 54705.07 | 28470.46 | 40575.97 | 41122.01 | 50111.25 |
***p < 0.001
**p < 0.01
*p < 0.05.
Unstandardized coefficients from growth curve models.
Approx. age | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 15 | 17 | 20 |
---|---|---|---|---|---|---|---|---|---|---|
Aggression | n. s. | n. s. | n. s. | |||||||
Non-aggressive externalizing | n. s. | n. s. | n. s. | n. s. | ||||||
ADHD symptoms | n. s. | n. s. | n. s. | |||||||
Anxiety and depression | n. s. | n. s. | n. s. | |||||||
Prosocial behavior | n. s. | |||||||||
Aggression | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | ||||
Non-aggressive externalizing | n. s. | n. s. | n. s. | |||||||
ADHD symptoms | n. s. | n. s. | n. s. | n. s. | ||||||
Anxiety and depression | n. s. | n. s. | n. s. | n. s. | n. s. | |||||
Prosocial behavior | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | ||
Aggression | n. s. | n. s. | n. s. | b = -.09, |
n. s. | |||||
Non-aggressive externalizing | n. s. | n. s. | b = -.04, |
b = -. 06 |
n. s. | |||||
ADHD symptoms | n. s. | n. s. | n. s. | n. s. | n. s. | |||||
Anxiety and depression | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | ||
Prosocial behavior | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. |
Unstandardized coefficients from growth curve models.
Approx. age | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 15 | 17 | 20 |
---|---|---|---|---|---|---|---|---|---|---|
Aggression | n. s. | n. s. | ||||||||
Non-aggressive externalizing | n. s. | |||||||||
ADHD symptoms | ||||||||||
Anxiety and depression | ||||||||||
Prosocial behavior | n. s. | n. s. | n. s. | n. s. | ||||||
Aggression | n. s. | n. s. | n. s. | n. s. | ||||||
Non-aggressive externalizing | n. s. | n. s. | n. s. | |||||||
ADHD symptoms | n. s. | n. s. | ||||||||
Anxiety and depression | n. s. | n. s. | n. s. | n. s. | n. s. | |||||
Prosocial behavior | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | |||
Aggression | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | |||
Non-aggressive externalizing | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | ||
ADHD symptoms | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | |||
Anxiety and depression | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | ||
Prosocial behavior | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. |
Unstandardized coefficients from growth curve models.
Approx. age | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 15 | 17 | 20 |
---|---|---|---|---|---|---|---|---|---|---|
Aggression | n. s. | n. s. | n. s. | n. s. | ||||||
Non-aggressive externalizing | n. s. | n. s. | ||||||||
ADHD symptoms | n. s. | n. s. | n. s. | |||||||
Anxiety and depression | n. s. | n. s. | n. s. | |||||||
Prosocial behavior | n. s. | n. s. | n. s. | n. s. | ||||||
Aggression | n. s. | n. s. | n. s. | n. s. | b = .04, |
n. s. | ||||
Non-aggressive externalizing | n. s. | n. s. | n. s. | |||||||
ADHD symptoms | n. s. | n. s. | n. s. | n. s. | ||||||
Anxiety and depression | n. s. | n. s. | n. s. | n. s. | n. s. | |||||
Prosocial behavior | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | ||
Aggression | b = -.12, |
b = -.09, |
n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | ||
Non-aggressive externalizing | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | ||
ADHD symptoms | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | n. s. | ||
Anxiety and depression | b = -.12, |
n. s. | n. s. | n. s. | ||||||
Prosocial behavior | b = 0.19, |
n. s. |
The second pattern of findings occurred only twice and at a fairly late age: a quadratic relation between care by a daycare mother and self-reported aggression at ages 15 and 17. These relations were similar to those for teacher-reported social behavior (see
Next, we examined the associations for daycare centers.
Other results were less consistent. For example, daycare center care was associated with more self-reported aggression at ages 7 and 8, but the relation flipped at ages 13 and 15, where center care was related to less aggression. Because this latter finding was somewhat unexpected given our other findings, we examined it further and performed robustness checks, which are reported in
Finally, we examined the associations for playgroups.
Second, playgroups were associated with lower self-reported aggression at ages 13 and 15. At age 17, the relation was quadratic and also suggested that time spent in a playgroup was associated with lower aggression (
Third, the most consistent results were found for teacher-reported aggression, anxiety and depression, and prosocial behavior. All associations were quadratic and similar in shape (see
Summing up, evidence suggested that daycare center attendance was related to more problem behavior according to the parents and, to some extent, the children themselves. Spending more than two to three days per week at a daycare mother and visiting a playgroup on at least three days per week was related to more externalizing and internalizing behavior and less prosocial behavior according to the teachers. Relations decreased as the children got older.
We proceeded with growth curve analyses for the relation of external care with delinquency and substance use. Results are shown in
Unstandardized coefficients from growth curve models.
Delinquency | Deviance | Substance Use | Delinquency and substance use | |
---|---|---|---|---|
Informant | Self | Self | Self | Teacher |
Family | 0.12 | -0.01 | 0.06 | -0.01 |
Acquaintances | 0.57 | 0.08 | ||
Daycare mother | -0.02 | -0.03 | -0.07 | -0.01 |
Daycare center | 0.16 | 0.03 | -0.08 | -0.05 |
Daycare center sq. | -- | -- | -- | |
Playgroup | -0.57 | -0.78 | 0.09 | |
Playgroup sq. | -- | -- | ||
Family | 0.01 | -0.09 | -0.30 | -0.30 |
Acquaintances | -3.90 | -0.46 | -0.25 | -0.48 |
Daycare mother | -0.48 | -0.01 | 0.35 | 0.28 |
Daycare center | -0.81 | 0.02 | 0.04 | |
Daycare center sq. | -- | -- | -- | |
Playgroup | 0.80 | -0.01 | ||
Playgroup sq. | -- | -- | ||
Family | -0.30 | -- | 0.23 | 0.33 |
Acquaintances | 3.09 | -- | 0.12 | 0.90 |
Daycare mother | 0.42 | -- | -0.28 | -0.42 |
Daycare center | 0.80 | -- | -0.22 | |
Daycare center sq. | -- | -- | -- | |
Playgroup | -0.04 | -- | -1.11 | -0.11 |
Playgroup sq. | -- | -- | 0.68 | -- |
χ2-Value | -- | -- | 24.42 | -- |
χ2 df | -- | -- | 22 | -- |
CFI | -- | -- | 1.00 | -- |
TLI | -- | -- | 0.99 | -- |
RMSEA Estimate | -- | -- | 0.01 | -- |
SRMR | -- | -- | 0.01 | -- |
BIC | 44845.04 | 57308.44 | 50772.95 | 39039.89 |
AIC | 43408.94 | 55877.45 | 49096.64 | 37608.90 |
***p < 0.001
**p < 0.01
*p < 0.05.
As an alternative measure of delinquency, we examined official delinquency records in relation to external childcare. There were no significant relations (
Summing up, we found little consistent evidence for a relation between external childcare, delinquency, and substance use in adolescence.
Finally, we examined relations between vulnerability, external childcare, and behavior. Similar to prior studies (e.g., [
Besides the overall growth curve analyses with the earliest available age as the centering points (
We found no evidence for a relation between external care, risk, and behavior for childcare by family members and in playgroups (
Some evidence suggested relations between external childcare by acquaintances, risk, and (mostly teacher-reported) social behavior (
Regarding external childcare in daycare centers, there was some evidence for a relation between daycare attendance, risk, and (primarily teacher-reported) social behavior (
We repeated our analyses on the relation between external childcare and risk for delinquency and substance use. In other words, we examined whether our earlier findings that external childcare was mostly unrelated to delinquency and substance use applied to children from both more and less vulnerable backgrounds (
In a final step, we examined the relation between external care, risk, and the official criminal records (
In order to provide a concise summary of the study’s main findings, an overview is included in
Type of external childcare | Results by child development domain |
---|---|
External family members | |
Acquaintances | |
Daycare mother | |
Daycare center | |
There was some evidence that children’s background mattered. Specifically, there was some evidence that for children from more vulnerable backgrounds, attending a daycare center was associated with less externalizing behavior and more internalizing problems. For children from less vulnerable backgrounds, however, attending a daycare center was associated with more externalizing behavior. | |
Playgroup | |
We examined the relation of external childcare before Kindergarten with child and youth development from ages 7 to 20. Our study was novel because to our knowledge, no prior study has examined a similarly wide range of social behavior, external care arrangements, and particularly length of follow-up. Our findings showed that external childcare was related to social behavior in a number of ways, but that the type of childcare, the type of social behavior, age, and informant mattered.
For externalizing behavior, we found that spending more time in a daycare center was related to more problem behavior according to the parents. The relations persisted until at least age 11 (when the parent interviews were ceased). Some evidence suggested that it was also related to more ADHD symptoms in adolescence (age 15 and 17) as reported by the youths themselves. More than, roughly, three days per week at a daycare mother or in a playgroup across the period before Kindergarten was related to increased teacher-reported aggression, non-aggressive externalizing behavior, and ADHD symptoms until age 11 to 13. We found some relations between external childcare by acquaintances and behavior, but these were inconsistent, which may be due to the small number of children who received this type of external childcare (only 3%).
Our findings on externalizing behavior are in line with some, though not all, of the very few studies that examined long-term relations, which have found that the relation between time spent in a daycare center and externalizing problems may persist until the end of primary school or into adolescence ([
Regarding delinquency, our findings suggested little consistent evidence for a relation with external childcare. Few studies have been done on the relation between external childcare and delinquency. One exception was the study by Baker et al. [
In addition to our findings on externalizing behavior, we also found that spending more time in a daycare center was related to more anxiety and depression at ages 7 to 11. This finding is in line with those from some prior studies [
On the positive side, we found evidence that care by family members was associated with increased prosocial behavior, which to our knowledge has not been examined or found before. Possibly, being embedded in a larger family context can provide children with the love, belonging, and trust [
Finally, we found no strong evidence that children from more vulnerable backgrounds profited more from external childcare compared to other children. There were some noteworthy findings, however. First, care by acquaintances and by daycare mothers was associated with more externalizing behavior for children from vulnerable backgrounds and less externalizing behavior for children from less vulnerable backgrounds. This was somewhat surprising as we are not aware of prior research or theory that might explain this. Possibly, parents from vulnerable backgrounds are more likely to choose acquaintances and daycare mothers who are also in vulnerable circumstances, potentially adding to the child’s stress and promoting problem behavior.
Second, some of our findings showed that spending more time in a daycare center was related to less externalizing behavior for children from vulnerable backgrounds, but to more problem behavior for children from less vulnerable backgrounds. Although the international literature has been mixed on this topic, our findings are in line with some prior studies that, like ours, used a cumulative risk index [
On the other hand, our data uniquely suggest that daycare center care was related to both more internalizing problems and substance use for children from vulnerable backgrounds. For substance use, the relation persisted through adolescence and even into early adulthood. Thus, the structured context and rules in a daycare center may help these young people to control their externalizing behavior but not their internalizing problems and substance use. It is possible that the experienced control and suppression of externalizing behavior leads to internalizing symptoms. Alternatively, children from precarious conditions may experience daycare centers as more unpleasant and hostile than other children, because daycare centers (like non-precarious households) have certain “middle class” behavioral norms, which could lead to a permanent feeling of alienation. Furthermore, in line with the bioecological model [
One interesting finding is that our results differed by informant. Whereas external childcare in a daycare center was related to externalizing behavior according to the parents (and, to some extent, the children themselves), it was external childcare by a daycare mother or in a playgroup that was related to problem behavior according to the teachers. One potential explanation may be that informants observe different contexts. Daycare centers are probably most similar to the school environment, which may be why teachers do not observe effects of daycare centers. Prior research to some extent supports this argument, finding that both daycare centers and schools increased children’s level of stress (i.e., cortisol) and that stress levels at school were just as high for children who had attended a daycare center and those who had not [
Due to the wide range of outcomes, types of childcare, and time-span included, as well as the multi-informant nature of our data, we were able to paint a nuanced picture. Though there were some results suggesting relations between childcare and productive youth development (e.g., the association between external family care and prosocial behavior, and decreased externalizing behavior among vulnerable children in daycare), other results suggested relations between childcare and increased problem behavior (e.g., the association between daycare and externalizing, and increased internalizing problems and substance use among vulnerable children in daycare), highlighting the necessity of examining various outcome domains. All in all, our findings suggest that early life experiences in external care shape children’s pathways, suggesting that a life course view is important in the relation between external childcare and development. Ultimately, other contexts and experiences gain precedence, and most of the relations diminished as children got older, although some remained until late adolescence and early adulthood.
We cannot claim that our results are causal because of limitations of our data. First, as described in the introduction, studies that examine the relation between external childcare and child development cannot rule out selection bias, and our study is no exception. That is, whether or not children attend external childcare is related to characteristics of the children and their families. In turn, child development is related to these characteristics, too, making it unclear whether differences in child development between children can be attributed to external childcare or to other characteristics. Like many previous studies, we have tried our best to reduce this problem by including a range of control variables that may explain why children entered external childcare. However, our data and statistical design do not allow us to rule out other potential explanations. For example, we did not know whether or not parents enrolled their children in external childcare because they displayed problem behavior. As mentioned, we performed robustness analyses to examine how likely it is that our results are influenced by this (see
As another limitation, we did not have data on the quality of external childcare, which varies widely in Switzerland [
In addition, information on external childcare was collected retrospectively, at age 7. Although the reliability of retrospective childcare reports is supported [1990], the EHC used in our study has shown validity [
Also, because of the nature of our longitudinal data, our measures on external care were from the years 1997 to 2002. Since then, the provision of external childcare has seen a rapid expansion. Although the city promotes the quality of care, to our knowledge no data exist on how daycare centers in Switzerland have changed. In the Netherlands, which has seen a similar expansion of external childcare, the increase in the quantity of daycare centers was accompanied by a decline in quality [
Finally, due to space constraints, we could not examine differences by gender [
Although there were limitations to our study, to our knowledge there are no other Swiss studies with comparably extensive data that can be linked to external childcare. Even internationally, few if any similar studies exist. Summing up, our study found that external care tends to be related to increased externalizing behavior (i.e., aggression, non-aggressive externalizing, and ADHD symptoms), internalizing problems (i.e., anxiety and depression), and substance use, but that most though not all relations diminished and disappeared as children grew into adolescents and young adults. An exception was the relation between daycare attendance and substance use, which persisted into young adulthood for those from vulnerable backgrounds.
(DOCX)
(DOCX)
(DOCX)
(DOCX)
(DOCX)
(DOCX)
Unstandardized coefficients from growth curve models.
(DOCX)
Unstandardized coefficients from growth curve models.
(DOCX)
a. Correlations between externalizing behavior and internalizing problems at ages 0 to 1 and enrollment in external childcare at age 2. b. Correlations between externalizing behavior and internalizing problems at ages 0 to 2 and enrollment in external childcare at age 3.
(DOCX)
Unstandardized coefficients from growth curve models.
(DOCX)
Unstandardized coefficients from growth curve models.
(DOCX)
Unstandardized coefficients from growth curve models
(DOCX)
Unstandardized coefficients from growth curve models.
(DOCX)
Unstandardized coefficients from growth curve models.
(DOCX)
Unstandardized coefficients from regression models.
(DOCX)
Unstandardized coefficients from growth curve models.
(DOCX)
Unstandardized coefficients from growth curve models.
(DOCX)
Unstandardized coefficients from growth curve models.
(DOCX)
Unstandardized coefficients from growth curve models.
(DOCX)
Unstandardized coefficients from growth curve models.
(DOCX)
Unstandardized coefficients from growth curve models.
(DOCX)
Unstandardized coefficients from growth curve models.
(DOCX)
Unstandardized coefficients from regression models.
(DOCX)
(DOCX)
(DOCX)
(DOCX)
PONE-D-21-10331
External childcare and socio-behavioral development in Switzerland: Long-term relations from childhood into young adulthood
PLOS ONE
Dear Dr. Averdijk,
Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.
As you will see, the reviewers provide thoughtful, but critical feedback on your work. Both reviewers note significant concerns about the inferences to be drawn from this study design. The reviewers are clear on the issues in how the results are interpreted in light of the non-randomization of children into or not into day care settings. Although I do not think that there solutions that will fully address this issue, I encourage you to think deeply about how to clarify the interpretations, such that causal language is eliminated. The reviewers also note multiple concerns with selection and potential non-representativeness of the sample. Some analytic details are in the supplementary material; it may be advisable to include those analyses in the body of the manuscript. While I agree that there may be challenges in using these data to make causal claims, I see these data as being part of the context of understanding this broad area of research, despite the limitations in non-randomized designs. My hope is that the inferences can be tempered and make statements about this critical issue in the field.
Although I highlight only a small number of the reviewer's comments, please address all comments in full. The reviewers have an appropriately high bar for the scientific evaluation of the work given the developmental and public policy implications of this work.
Please submit your revised manuscript by Sep 27 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at
Please include the following items when submitting your revised manuscript:
A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'. A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'. An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.
If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.
If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see:
We look forward to receiving your revised manuscript.
Kind regards,
Thomas M. Olino
Academic Editor
PLOS ONE
Journal requirements:
When submitting your revision, we need you to address these additional requirements.
1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at
2. Thank you for including your ethics statement: Ethical approval was obtained by the Ethics Committee of the Faculty of Arts and Social Sciences of the University of Zurich
a) Please provide additional details regarding participant consent. In the ethics statement in the Methods and online submission information, please ensure that you have specified (1) whether consent was informed and (2) what type you obtained (for instance, written or verbal, and if verbal, how it was documented and witnessed). If your study included minors, state whether you obtained consent from parents or guardians. If the need for consent was waived by the ethics committee, please include this information.
If you are reporting a retrospective study of medical records or archived samples, please ensure that you have discussed whether all data were fully anonymized before you accessed them and/or whether the IRB or ethics committee waived the requirement for informed consent. If patients provided informed written consent to have data from their medical records used in research, please include this information.
Once you have amended this/these statement(s) in the Methods section of the manuscript, please add the same text to the “Ethics Statement” field of the submission form (via “Edit Submission”).
For additional information about PLOS ONE ethical requirements for human subjects research, please refer to
3. We note that the grant information you provided in the ‘Funding Information’ and ‘Financial Disclosure’ sections do not match.
When you resubmit, please ensure that you provide the correct grant numbers for the awards you received for your study in the ‘Funding Information’ section.
4. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see
"Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see
Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data:
We will update your Data Availability statement to reflect the information you provide in your cover letter.
Additional Editor Comments (if provided):
As you will see, the reviewers provide thoughtful, but critical feedback on your work. Both reviewers note significant concerns about the inferences to be drawn from this study design. The reviewers are clear on the issues in how the results are interpreted in light of the non-randomization of children into or not into day care settings. Although I do not think that there solutions that will fully address this issue, I encourage you to think deeply about how to clarify the interpretations, such that causal language is eliminated. The reviewers also note multiple concerns with selection and potential non-representativeness of the sample. Some analytic details are in the supplementary material; it may be advisable to include those analyses in the body of the manuscript. While I agree that there may be challenges in using these data to make causal claims, I see these data as being part of the context of understanding this broad area of research, despite the limitations in non-randomized designs. My hope is that the inferences can be tempered and make statements about this critical issue in the field.
Although I highlight only a small number of the reviewer's comments, please address all comments in full. The reviewers have an appropriately high bar for the scientific evaluation of the work given the developmental and public policy implications of this work.
[Note: HTML markup is below. Please do not edit.]
Reviewers' comments:
Reviewer's Responses to Questions
1. Is the manuscript technically sound, and do the data support the conclusions?
The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.
Reviewer #1: Partly
Reviewer #2: No
**********
2. Has the statistical analysis been performed appropriately and rigorously?
Reviewer #1: Yes
Reviewer #2: Yes
**********
3. Have the authors made all data underlying the findings in their manuscript fully available?
The
Reviewer #1: Yes
Reviewer #2: No
**********
4. Is the manuscript presented in an intelligible fashion and written in standard English?
PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.
Reviewer #1: Yes
Reviewer #2: Yes
**********
5. Review Comments to the Author
Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)
Reviewer #1: The manuscript "External childcare and socio-behavioral development in Switzerland: Long-term relations from childhood into young adulthood" seeks to answer whether early external childcare -understood as care outside the child’s home- relates to different behavioral outcomes (e.g., externalizing behavior, internalizing problems, prosocial behavior) and risk-taking activities (e.g., delinquency, and substance use) across the lifespan (i.e., from school entry to early adulthood). Although there are strengths to the study, including the use of a large multi-ethnic sample from Switzerland that includes data of life course events and outcomes, as well as the use of multi-informant reported outcomes and the study of different types of childcare arrangements, there are significant limitations that compromise the validity of the inferences made in the study.
Major comments:
Introduction and Results
The introduction section could benefit from a more comprehensive review of the most recent literature on the topic. Given that this article seeks to address a research question that (1) has been very controversial in the field of human development (Dearing and Zachrisson, 2017), and (2) has been most recently approached through quasi-experimental approaches to overcome long-discussed methodological issues of endogeneity/selection bias, I suggest the authors should not only focus on presenting correlational studies but should be more exhaustive in presenting recent research that has used quasi-experimental approaches and strategies to address issues of confounding. Some examples of such studies include:
- Jaffee, S. R., Van Hulle, C., & Rodgers, J. L. (2011). Effects of nonmaternal care in the first 3 years on children’s academic skills and behavioral functioning in childhood and early adolescence: A sibling comparison study. Child Development, 82, 1076–1091. doi:10. 1111/j.1467-8624.2011.01611.x
- Crosby, D. A., Dowsett, C. J., Gennetian, L. A., & Huston, A. C. (2010). A tale of two methods: Comparing regression and instrumental variables estimates of the effects of preschool child care type on the subsequent externalizing behavior of children in low-income families. Developmental Psychology, 46, 1030–1048. doi:10.1037/a0020384
- Baker, M., Gruber, J., & Milligan, K. (2019). The Long-Run Impacts of a Universal Child Care Program. American Economic Journal. Economic Policy, 11(3), 1-26.
Moreover, the authors should also include recent studies that have called into question the internal validity of correlational studies on the relation between time spent in childcare and behavior problems. This recent research has found that conflicting findings on the subject are mostly due to model specifications and the degree of rigor in eliminating selection bias in the models. Indeed, when more rigorous statistical designs have been used to rule out selection bias, the relation between time in child care and behavior problems is, in some studies, no longer statistically significant, suggesting issues of confounding in correlational evidence. Please see the following references for a review of studies:
- Dearing, E., & Zachrisson, H. D. (2017). Concern Over Internal, External, and Incidence Validity in Studies of Child-Care Quantity and Externalizing Behavior Problems. Child Development Perspectives, 11(2), 133–138.
This is also the case for studies that have looked at the association between early child care and behavior problems during adolescence. When studies have used approaches that are less rigorous in ruling out selection bias (using only covariate-adjustment), they have found positive associations between time spent in child care and externalizing behavior problems (e.g., Vandell, Burchinal, and Pierce, 2016), whereas studies using some strategy to address selection, for instance reducing imbalance in observed characteristics by using propensity score matching (e.g., Orri et al. 2019) the relation between child care and behavior problems is even negative. Nonetheless, these important problems related to internal validity are not mentioned in the literature review.
Considering that reviews of the literature clearly demonstrate these issues of selection, I suggest authors revise some of their claims, such as “Our findings on externalizing behavior are largely in line with the international studies discussed earlier, the majority of which have found that more time spent in a daycare center is related to increased externalizing problems at earlier ages (lines 509-511)”. It is not true that the majority of the international evidence indicates such a relation, and most importantly, even if this was the case, the important issue is not whether the majority of papers show a positive association, as if these studies were poorly conducted any inference would be invalid. Instead, the focus should be on internally-valid studies that have used approaches that go beyond traditional covariate adjustment to mitigate issues of selection bias. As already mentioned, such studies tend to find no relation at all and even reductions in behavior problems as children spend more time in child care.
Furthermore, I suggest the authors to revise whether some of the references cited in the article might be equivocally used. For instance, McCartney’s et al. (2010) mentioned that “the evidence linking child care hours with externalizing behavior was equivocal in that results varied across model specifications” (McCartney et al., 2010), and for such reason they tested five causal propositions that “if satisfied, would be consistent with the view that child care hours and externalizing behavior are causally linked.”, which they then show are not satisfied. Since one of the initial claims made in the introduction is that “Prior research on the amount of external childcare has generally concluded that more time in external childcare is related to increased problem behavior.”, citing McCartney’s (2010) article, it seems inaccurate.
Regarding the External childcare in Switzerland subsection, I suggest the authors give a more comprehensive review of the ECEC policies/contexts in Switzerland, and findings are also contextualized accordingly. Switzerland had (still has) the most expensive ECEC in the world (about 50% of disposable income, on average, goes into paying for ECEC), and the country has a very family/mother-unfriendly policies (see Chzhen, Gromada, & Rees, 2019, Are the world’s richest countries family friendly? Policy in the OECD and EU). Although the authors put attention to the counterfactual modes of care, they run the risk of overemphasizing results from a very uncommon context if it is not addressed in more detail.
Results and Discussion
I was disappointed that the authors (from the introduction to the discussion) paid little attention to questions of causality and selection, which are central to the research questions being addressed. Although the authors mention how this study cannot draw causal relations because it uses observational data (lines 325,326), I suggest they clarify that it is not only because of the type of data, but also because the statistical design used does not allow the authors to credibly discard the potential confounders. Even if authors had been capable to rule out reverse causality successfully and prove that parents did not enroll their children into external care due to their problem behavior, there are many other threats to validity that this study does not consider for relying exclusively on covariate-adjustment. Given the implications of a study like this could potentially have, it would not be responsible to present evidence suggesting that child care might be detrimental for young children while relying on a weak empirical strategy that is prone to bias.
Given this, my biggest concern with the paper is the lack of rigor in the language used, as these correlational results could have great implications in practice (i.e., policies on early child care provision, female participation in the labor market). Although the authors acknowledge none of the findings are causal, through the results and discussion they seem to suggest that there might be a causal link. For instance: “Children from less vulnerable backgrounds, however, did not profit from more time in a daycare center or even fared worse” (lines 468-469), or “Though there were some results suggesting beneficial relations (e.g., the association between external family care and prosocial behavior, and decreased externalizing behavior among vulnerable children in daycare), these were offset by less positive ones (e.g., the association between daycare and externalizing, and increased internalizing problems and substance use among vulnerable children in daycare). I encourage authors to address results with much caveat and some careful explanation.
To conclude, while the manuscript could benefit from expanding the literature review and tempering the causal language, the contribution of another correlational study to the field would be minor, considering the long-lasting discussions and evidence showing bias in such studies.
Minor comments:
Methods and Statistical Analyses
Given that ‘time in external childcare’ is the predictor of interest, I suggest moving the information about the data collection using the Event History Calendar back to the article to provide more detailed description of (i) who reported the data, and (ii) the reliability of the measure. Otherwise it becomes less clear who actually reported the childcare information (whether it was the child or the parents) and how the EHC design provides accurate enough data of time spent in childcare that is collected retrospectively. Also, I suggest authors not only give evidence supporting the use of an EHC but also discuss how retrospective self-reported data can bias the results.
It is not clear to me whether the final sample consisted of the 94% of 1,675 first graders that had participated in at least one of the data collections (line 194) or the 1,225 children whose parents had completed the EHC (line 239). I also suggest including the sample size in each of the regression models to be able to compare and check the sample sizes of each model (or clarify that they are the same given that you have done multiple imputation).
I suggest authors discuss further why some of the results showed different directions in the association at different ages: “For example, daycare center care was associated with more self-reported aggression at ages 7 and 8, but the relation flipped at ages 13 and 15, where center care was related to less aggression.”
There are some references that are cited in the text but are not included in the reference list. For example: Vandell et al., 2010 (line 123), Wustmann Seiler et al., 2017 (line 174).
Reviewer #2: The article is well built and well written from a theoretical point of view, and provide a lot of estimations and robustness checks. The article is very (too much?) ambitious, treating several outcomes at different ages, with quite limited, non-representative sample. After reading the summary and the article, I have not a take-away message since results go in several directions according to the informant and child’s age. The major weakness is that the causality between external childcare when the child was young and child’s future development or trend in child’s future development (that the model tests) are not established. How to be sure that the relation observed is causal and not a spurious relationship due to unobserved confounding variable? Even the valuable efforts of authors to control for many of them, this is not enough and I do not belief that growth curve models are particularly adequate for the topic. I would recommend the authors to use policy evaluation methods to treat better causality and to limit the ambitions of the article by limiting the number of outcomes and time periods.
METHOD
I wonder whether the use of growth curve model is adequate for the topic since the event we are interested in (external childcare when child is very young) occur many years before the large period of observation of outcomes (from age 7 to 20). It might affect rather the initial level rather the hereafter trend. I do not see any rationale why the daycare would affect the rate of change at which child developed after the initial level.
If authors keep this method, a short explanation for non-initiated reader is anyway really needed. The authors comment the different parameters without reminding what they mean (inter-individual variability or intra-individual patterns of change over time), so the results are not easily readable for non expert.
The sample size is only mentioned once, and should be precise in each table. Do the sample size available is the same for all regressions? If not, what has been the choice to keep the common individuals interviewed each wave or to keep maximum individuals at each wave?
The age range from 7 to 20 is very large
RESULTS
I would like to see a description of selection into external childcare. Who are the parents who use external childcare and which type of childcare?
Some results are weird, particularly when they differ by informant. They seem sensitive to measurement, so I have doubts about their validity.
A discussion is missing about the non representative sample. As it is a sample based on a town, it may be possible to control for the daycare attended more precisely.
The fact that more internalizing problems and substance problems are only observed for children attending daycare coming from vulnerable background put doubts on the role of unobserved characteristics. Who are these children? Do they attend same daycare than others? Do they have the same life-course?
More importantly, several unobserved characteristics are possible and strong mediator of results: for instance whether the mother was working part-time full-time (probably by financial necessity for children coming from disadvantaged background) part-time or not working at all, the type of kindergarten attended (of worse quality), the instability of family structure (dissolution and the arrival of a step-parent) from Kindergarden or hereafter, etc.
Minor remarks:
Why is the rationale of precising that the sample is multi-ethnic in the summary since no much emphasis on this point is put in the article?
The limitations section should come back on data, and selection in use of external child care.
**********
6. PLOS authors have the option to publish the peer review history of their article (
If you choose “no”, your identity will remain anonymous but your review may still be made public.
Reviewer #1: No
Reviewer #2: No
[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]
While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool,
Please refer to the uploaded Cover Letter and Response to Reviewers document
Submitted filename:
PONE-D-21-10331R1External childcare and socio-behavioral development in Switzerland: Long-term relations from childhood into young adulthoodPLOS ONE
Dear Dr. Averdijk,
Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.
My apologies for my error in the previous letter.
I was able to obtain the review from one of the original reviewers, but the other was unable to provide comments. In my reading of the manuscript, I see much greater clarity in the ways that you highlight what your study can and cannot conclude based on the design employed in the study. At the same time, the Reviewer is asking for transparency in the methods used by previous studies, particularly those studies that used methods permitting drawing causal inferences. If there were observational studies that appropriately drew causal inferences, some additional clarifications about why those methods could not be used in your work would be helpful.
Please submit your revised manuscript by Jan 08 2022 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at
Please include the following items when submitting your revised manuscript: A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'. A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'. An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.
If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.
If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see:
We look forward to receiving your revised manuscript.
Kind regards,
Thomas M. Olino
Academic Editor
PLOS ONE
Journal Requirements:
Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.
[Note: HTML markup is below. Please do not edit.]
Reviewers' comments:
Reviewer's Responses to Questions
1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.
Reviewer #1: (No Response)
**********
2. Is the manuscript technically sound, and do the data support the conclusions?
The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.
Reviewer #1: No
**********
3. Has the statistical analysis been performed appropriately and rigorously?
Reviewer #1: Yes
**********
4. Have the authors made all data underlying the findings in their manuscript fully available?
The
Reviewer #1: Yes
**********
5. Is the manuscript presented in an intelligible fashion and written in standard English?
PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.
Reviewer #1: Yes
**********
6. Review Comments to the Author
Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)
Reviewer #1: 1. Although the authors have included the suggested literature, I’m worried that the conclusion inferred from the causal evidence and from the most recent reviews (e.g., Dearing & Zachrisson, 2017; Huston, 2015; McCartney et al., 2010) is still wanting. It is not that “different methods may yield to different results” but that more rigorous methods with better identification strategies to address selection bias lead to more internally valid results. I am afraid this idea is still not well addressed in the article and that the reader might potentially draw the wrong conclusions from the article because of this.
This is the case in statements such as: “Past literature reviews on the link between the quantity of external childcare and child development have concluded that external childcare is related to increased problem behavior (see Dearing & Zachrisson, 2017, for an overview).”, which does not clarify that this is the case when less rigorous methodologies such as covariate-adjustment/OLS are implemented but that the opposite relation is found when selection bias is addressed through quasi-experimental approaches. Another example is: “Other studies have used other methods, which have their own unique advantages and disadvantages, such as sibling comparisons, propensity score matching, and instrumental variables and shown that different methods may yield different results (e.g., McCartney et al., 2010).”
2. I strongly suggest that the authors relate the specific methods used in the articles they mentioned to make particularly strong claims. Again, this kinds of statements that have strong policy implications must be stated with enough context. For instance, "the previously mentioned NICHD study found that even though low-quality external care is more often associated with problem behavior than high-quality care, external care was related to more externalizing behavior at all levels of quality (NICHD Early Child Care Research Network, 2003)." or "A Swiss study showed that an increased availability of external childcare provisions would affect parental labor participation only marginally (Stern et al., 2018)." What analytical strategy was this study using? What was the identification strategy used? What was the sample studied? Was it nationally-representative?
3. Although the data used in this study is rich in that it follows children until adolescence and early adulthood and it includes multiple outcome variables, I want to reiterate that this study fails to address selection effects, making it difficult to draw meaningful conclusions. While the authors employ covariates and some initial robustness checks, controlling for observed characteristics, the growth curve model employed is insufficient to establish internally valid estimates given that potential unobserved characteristics that relate to both time spent in external child care and the outcomes might still not be accounted for. It is exactly for this reason that recent research has moved away from correlational designs to address a topic that is very sensitive in terms of policy implications for early childhood care and education supply.
**********
7. PLOS authors have the option to publish the peer review history of their article (
If you choose “no”, your identity will remain anonymous but your review may still be made public.
Reviewer #1: No
[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]
While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool,
All of our responses to the reviewer's and editor's comments can be found in the Cover letter and Response to Reviewers document.
Submitted filename:
External childcare and socio-behavioral development in Switzerland: Long-term relations from childhood into young adulthood
PONE-D-21-10331R2
Dear Dr. Averdijk,
We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.
Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.
An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at
If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact
Kind regards,
Thomas M. Olino
Academic Editor
PLOS ONE
Additional Editor Comments (optional):
Reviewers' comments:
PONE-D-21-10331R2
External childcare and socio-behavioral development in Switzerland: Long-term relations from childhood into young adulthood
Dear Dr. Averdijk:
I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.
If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact
If we can help with anything else, please email us at
Thank you for submitting your work to PLOS ONE and supporting open access.
Kind regards,
PLOS ONE Editorial Office Staff
on behalf of
Dr. Thomas M. Olino
Academic Editor
PLOS ONE