Project topics and materials
Get Complete Project Material File(s) Now! »
Data and Methodology
In this section we presented the variables from the Dutch Longitudinal Internet Studies for the Social Sciences (LISS), we used to study the effect of the Dutch parental leave scheme on parents’ life satisfaction. Additionally, we provided de-scriptive statistics on life satisfaction, decision to take a parental leave, having young children, and decision to reduce working time outside the parental leave scheme.
Data
Our research was based on data taken from the Dutch Longitudinal Internet Studies for the Social Sciences (LISS) panel administered by CentERdata8 (see for details: www.lissdata.nl). The panel is based on a true probability sample of households drawn from the population register by Statistics Netherlands consisting of more than 4500 households over 8000 individuals and 93 monthly waves from November 2007 to September 2016. In the LISS survey, individuals report on several aspects of their life, including their satisfaction with life, parental leave and background information9. Our common sample was unbalanced10 and included 8,590 observations and 2,943 individuals observed over the period 2008-201311.
To be on parental leave a person must have at least worked with the same employer during one year, this allowed us to exclude individuals who had never worked during the entire observation period. We exclude years 2014, 2015 and 2016 from our analysis because the life satisfaction question was asked six month after the question on parental leave12. During the period between 2008-2013, the per-sonality questionnaire was asked only one month after the question on parental leave.
To analyze the impact of taking parental leave on life satisfaction, we first inves-tigated the effect of having children younger than eight on subjective well-being. We chose the age of eight as a cutoff because only parents with children younger than eight are entitled to use the Dutch parental leave scheme. Then, we studied whether on taking parental leave might influence the relationship between having young chil-dren and life satisfaction. As parental leave is a part-time work arrangement that allows parents to have more time for their children and themselves, it is reasonable to assume that while having young children has no clear impact on life satisfaction (Luhmann et al., 2012), having more free time for them does (Tausig and Fenwick, 2001; Pichler, 2009; Drobnič et al., 2010; Adams et al., 1996; Greenhaus et al., 2003). Further, parental leave was distinguish from working less because of children13. Our goal was to compare the effects of working time reduction outside the parental leave scheme with reduction of working time induced by parental leave on life satisfaction.
Our indicator for life satisfaction was based on the question « How satisfied are you with the life you lead at the moment? » the respondent was asked to use an ordinal scale from zero to ten, from not at all satisfied to completely satisfied. This single-item scale life satisfaction question is a widely used measure of subjective well-being. It has the advantage of asking the respondent to focus on an overall evaluation of their life rather than on current feelings or specific psychosomatic symptoms.Veenhoven (2000) and Frey and Stutzer (2002), have shown that life satisfaction is closely related to a number of other potentially more objective measures of happiness.
The well-being distribution of having a child under eight or not in the Nether-lands is illustrated in Figure 1. Very few individuals reported a level of well-being below five or over nine, which is standard in the literature. At a first glance the average value of life satisfaction among the two groups is about the same. In Figure 2, we further distinguish between parents with young children who took parental leave and those who had not. In the middle and higher score groups of seven and eight individuals not on parental leave dominate those on parental leave in percentage, while a wider share of parents on parental leave scored nine out of ten.
An overview of average well-being of parents with under eight children split by gender is given in Table 1. The last column in Table 1 confirms the findings shown in Figure 1 and 2. Average well-being does not differ noticeably between having children younger than eight or not, and between individuals working less hours to take care of their children or not, while individuals taking parental leave are on average more satisfied with their life than those who did not. On the scale from zero to ten, parents who have not taken parental leave scored on average 7,6 while those who took parental leave scored on average 7,9. Comparing the first two columns of Table 1, the average gap between parents who took parental leave and those who did not was slightly higher for women than for men.
The distribution of individuals having young children and not reducing their working hours, taking parental leave and working less outside the Dutch parental leave scheme is presented in Table 2. The second column of Table 2 describes the distribution of individuals taking parental leave; with the largest group being middle income, highly educated women in cohabitation with a single child under two years old. We observed an equal distribution across the public and private sectors, living environment and number of hours taking leave per week. The duration of parental leave generally exceed two years, however 43 percent of the respondents were effectively on parental leave for six months to up to two years.
Finally, a majority of parents in the Netherlands take parental leave as a part-time work arrangement, allowing them more easily to reconcile work and parenting without completely withdrawing from the labor market.
The first column of Table 2 reports the distribution of individuals having young children and not reducing their working hours. These parents have different characteristics than those taking parental leave; they mainly belong to high income category and work full-time in the private sector. The third column of Table 2 shows the distribution of parents reducing their working time outside the parental leave scheme. A vast majority of them belongs to the lowest income category and are women working part-time. Note that 10 percent of those reducing their working time are already on parental leave14.
All in all, we observed a potential selection effect on time-varying observables, namely income category, level of education, working time arrangements and working in the private or public sector.
Econometric Model
Our dependent variable life satisfaction was measured on an ordinal scale from zero to ten. Panel data allowed us to control for time-invariant unobserved personal characteristics using a linear robust fixed-effects model. In this model the dependent variable was assumed to be cardinal, however when analyzing subjective well-being, the linear fixed-effects estimation performs as well as the fixed-effects ordered logit estimation (Ferrer-i Carbonell and Frijters, 2004)15. Our model was specified as:
LSit = β0 + β1 Y oungN oLeaveit + β2 (Y oungit × Leaveit) +β3 (Y oungit × W ork.Lessit) + Xitβ4 + αi + εit
Where i (i = 1, 2, …, n) refers to individuals t (t = 1, 2, …, T ) stands for year and LSit is the self-reported life satisfaction of individual on a scale from zero to ten. β0 is the constant, Y oungN oLeaveit is a dummy vector of having a child younger than eight and not taking parental leave, Y oungit × Leaveit, denotes the interaction effect between having a young child and taking a parental leave, Y oungit × W ork.Lessit designate the interaction effect between having a young child and working less to take care of young child16 and Xit represents the vector of covariates that may be correlated to both parental leave and Life satisfaction for example work hours categories (Booth and Van Ours, 2013), having a one year old child17, work satisfaction, social contact satisfaction and moment feeling. Additionally, we controled for the usual demographic and socio-economic variables like age, living environment, health, the education level of the respondent, marital status, employment status, the log of personal net monthly income in Euros and year dummies (Booth and Van Ours, 2008). αi represents individual specific time-invariant effects, such as personality and εit is the error term.
Empirical Results
In this section we presented results from our baseline estimates. Further, we per-formed a sensitivity analysis to test the validity of our results. Then, we investigated the existence of reverse causality. Finally, we analyzed of how the life satisfaction of different parental sub-groups is influenced by the Dutch parental scheme.
Baseline Estimates
The results of our linear fixed-effects model are presented in Table 318. All of our models were estimated using cluster-robust standard errors at the individual level. Our results confirmed the U-shaped relationship between life satisfaction and age (Blanchflower and Oswald, 2008). we found that having a child younger than eight has no significant effect for life satisfaction, while taking parental leave increases life satisfaction on average by 0.2 points. The total effect of having young child and taking parental leave was denoted by the sum of the two estimated coefficients β2 and β3. Considering the last model specification in Table 3, having a child younger than eight and taking parental leave increase life satisfaction by 0.17 on average. The Wald-test revealed a significant difference in estimated coefficients between having a young child and not being on parental leave and the interaction between having a young child and taking parental leave. Thus, having a young child but not taking parental leave does not signif-icantly impact life satisfaction, while taking parental leave increases life satisfaction.
Reduction of working time that is not part of the parental leave scheme does not have a significant association with life satisfaction, although the difference in coefficients between working less to take care of a young child and taking parental leave is not statistically significant. Looking at Column 5 of Table 3, we noticed that when we added subjective control variables, namely work satisfaction, social contact satisfaction and moment feelings the significance and the size of the coefficient slightly decreased, but remained statistically significant.
We found a positive moderating role of parental leave on the relationship between having a young child and life satisfaction. This result confirmed our first hypothesis that, in the Netherlands, parental leave reduces the « time-bind » (Hochschild, 1997) generated by having young child i.e. it increases the extent to which workers feel suc-cessful in balancing their work and personal lives. Parental leave in the Netherlands induces a reduction in work-life imbalances leading to higher life satisfaction.
Robustness Test
We perform a sensitivity analysis to confirm the robustness of our results. Using an alternative definition for the dependent variable life satisfaction, re-running the regressions using an ordered logit fixed-effects model and distinguishing working less in addition from parental leave from working less outside the parental leave scheme. Then, we looked at the reverse causality issue and estimated the impact of lag life satisfaction on the actual decision to take parental leave.
Sensitivity Analysis
First we replaced the dependent variable life satisfaction with happiness19. Our estimation results are presented in Table 420. Having re-estimated the model using the happiness as dependent variable, our conclusions did not change in that there was still a positive association between parental leave and happiness as a subjective well-being indicator.
The estimated results using ordered logit fixed-effects are displayed in Table 521. In line with the findings of Ferrer-i Carbonell and Frijters (2004) assuming cardinality or ordinality in the subjective well-being data makes qualitatively little difference, and re-running the regression using an ordered logit fixed-effects model gave similar results to those found using linear fixed-effects model. In contrast to our baseline estimates, however, the estimated effect of taking parental leave on life satisfaction was significantly different from the estimated impact of working less to care of children on life satisfaction.
In a last specification we distinguished working less in addition from parental leave from working less outside the parental leave scheme in Table 622. As mentioned in the descriptive statistics part there was 94 observations for which individuals are taking parental leave and reducing their working hours at the same time. This distinction does not change our main findings. Taking parental leave and additionally reducing working hours, or not still largely and significantly impact life satisfaction, while reducing working hours outside the parental leave scheme does not affect life satisfaction. Nevertheless, the Wald-test revealed a significant difference in estimated coefficients between taking parental leave and additionally reducing working hours and only reducing working hours outside parental leave scheme. Thus, only reducing working hours without the legal framework of parental leave does not significantly impact life satisfaction, while taking parental leave and reducing working hours increases life satisfaction.
To conclude, using happiness as a dependent variable in Table 4, re-runing the analysis using ordered logit fixed-effect in Table 5 and distinguishing working less in addition from parental leave from working less outside the parental leave scheme produced substantively the same results as those presented in our baseline estimate in Table 3.
Reverse causality
To check the validity of our results we needed to take into account the possibility of omitted variable bias and reverse causality in the estimation of the coefficient of taking parental leave on life satisfaction. Using the linear fixed-effects model, we removed omitted fixed variable bias due to individual-specific unobserved het-erogeneity related to both parental leave and life satisfaction, i.e. the phenomenon that fixed individual characteristics, such as personality, may influence the choice to take parental leave and life satisfaction at the same time. The linear fixed-effects model does not consider possible reverse causality, i.e. the fact that an individual whose life satisfaction increases is more likely to take parental leave. Someone who becomes more satisfied with his or her life may be more willing to take free time to enjoy spending time with his or her children. Likewise, a person going through a depressive episode after the birth of a child may be more disposed to spend time at work.
To examine whether or not reverse causality was an issue, we looked at changes in life satisfaction over time and if they might influence the decision to take parental leave. We estimated a linear fixed-effects model in which the dependent variable was the interaction effect between having a child younger than eight or not and taking parental leave or not with independent variables life satisfaction divided by 10023 in earlier periods and the same covariates as before.
If a higher level of life satisfaction increased the probability of being on parental leave later on we could have a reverse causality issue. We used three different lags for life satisfaction to allow for effects that take shape quickly or more slowly. The relevant parameter estimates of lagged life satisfaction are presented in Table 7. Estimations a, b and c indicated that a positive shock to happiness of an individual not on parental leave does not increase his or her probability to take parental leave one, two or three years later. Rows d to f show that after controlling for covariates, past life satisfaction does not influences the choice to take parental leave. The same estimation in the sub sample of parents with a young child are shown in Table 8. None of the results are sizable or significant, from this we concluded that reverse causality from life satisfaction to future choice to take parental leave was not an issue (see also Chen and Van Ours (2018)).
Heterogeneity Analysis
In this section, we present an analysis of how the life satisfaction of different parental sub-groups is associated with parental leave in the Netherlands. Firstly, we examined how the Dutch parental leave scheme may shape the relationship. Secondly, we explored the influence of socio-demographic characteristics on the estimate impact of parental leave on life satisfaction.
Variation in Parental Leave Scheme
The estimated coefficient of our linear fixed-effects model by length of parental leave is shown in Table 9, panel a. Taking parental leave is related to an average life satisfaction increase of about 0.3 points for parents who are on parental leave for one month up to one and a half years. A similar significant coefficient size is found for parents during parental leave of between six months to eighteen months. Parents on parental leave for more than one and a half years, however, no longer enjoyed the benefits of this leave. This result may be explain by the process of hedonic adaptation (Brickman and Campbell, 1971) which suggest that individuals return to baseline levels of happiness following a change in life circumstance24.A possible explanation is that after two years parents get used to their part-time work arrangement and their extra free time no longer increases their life satisfaction. Additionally, Adema et al. (2015) find that across European countries negative wage and slower career opportunity progression largely follow long periods of leave from work, e.g. one or two years or more. As a consequence, when the parental leave arrangement is spread over an extended period it may generate negative work outcomes for career progression and wages that offset the work-life balance benefits of taking parental leave.
Estimated effects of parental leave weekly hours on life satisfaction are presented in Table 9, panel b. The size of estimated coefficient does not differ significantly by the number of weekly working hours of parental leave. A reduction of working hours may impact life satisfaction in two contrasting ways. Firstly, taking more hours per week off may help an individual to balance life and work in a better way, leading to an increase in life satisfaction. Secondly, in contrast, reducing an individual weekly working hours can induce lower earnings, reduce their capacity to deal with work demands, restrict their career opportunities, and encourage negative judgments from co-workers (Garnero, 2016). Hence, taking beyond a certain amount of leave per week may have negative work outcomes for an individual: reduc-ing weekly working hours may exceed the work-life balance benefits of parental leave.
Socio-Demographic Characteristics
The estimated coefficient of our linear fixed-effects model by socio-demographic characteristics are shown in Table 10. Although we can see that taking parental leave has a more profound impact on the life satisfaction of (1) men, (2) the lower educated, (3) those belonging to the high income category, (4) and employees work-ing in the public sector, none of the differences are statistically significant. In part, the non-significance of the results can be explained by the limited number of people in the sample taking parental leave. Follow-up research is necessary to examine those differences e.g. gender discrepancy in the effects of parental leave.
Discussion
We investigated the relationship between parental leave and life satisfaction. We showed that although having a child younger than eight has no significant effect on parental life satisfaction, life satisfaction is significantly higher among parents with a young child who are on parental leave. This finding confirms the idea that a family leave scheme reduces the perceived ‘time bind’ of parents and increases their subjective well-being. At the same time, a reduction of working time to take care of children outside the Dutch parental leave scheme does not affect life satisfaction, indicating that the legal framework of parental leave offering job protected leave, the legal obligation for an employer to accept rescheduling a parents’ work arrangement, the possibility to use the LCSS, access to fiscal benefit, and in some case financial support is crucial to enhance parents life satisfaction when taking temporary leave of absence in the form of parental leave to take care of their children. To test the robustness of our results we performed several sensitivity analyses. The findings hold using different estimation strategies and alternative definitions of the dependent variable. Additionally, we did not find evidence of existing reverse causality. The heterogeneity analysis revealed that short parental leave schemes are significantly more conducive to life satisfaction than long parental leave schemes, and no significant differences between subgroups were found.
Our results may have some implications for public policy. The use of parental leave should be promoted and encouraged by the State and by companies. The recent European directive on work-life balance for parents careers will, in this regard, help to promote paid parental leave supported by national laws. More generally, family friendly policies could be encouraged by public policy in order to avoid work life imbalances caused by parenthood.
Our study has some limitations that should be addressed in future research. One, we have an external validity issue as the analysis was restricted to the Netherlands the results we have found may not be generalized to other countries. In this regard, cultural specificities and legislation on birth-related leave and childcare systems may moderate the effect of parental leave on life satisfaction. Two, we dealt with reverse causality by estimating the impact of parental leave on past life satisfaction but this did not allow us to resolve this issue completely in the absence of a good instrumental variable or natural experiment. An experimental analysis needs to be undertaken to identify clearly the causal impact of parental leave on life satisfaction.
Three, although we included a number of time-varying covariates and applied fixed-effects model to account for time-invariant unobservables, we cannot entirely settle the concern of the possible time-varying unobservables. Four, although we explored for whom the relationship is prevalent, a more detailed heterogeneity analysis is needed. We should look at how different life styles may shape the relationship between parental leave and life satisfaction. For instance, family and career oriented individuals may experience parental leave in different ways. Such a study would require a larger number of people taking parental leave in the data set. Five, we assumed that parental leave reduces work-life imbalances, and so induces higher life satisfaction, however, we lacked the information needed to undertake a mediation analysis. All these limitations need to be addressed in future research.
Table of contents :
General Introduction
Motivations
Theoretical Framework
Research Objectives and Research Questions
Introduction générale
Motivations
Cadre théorique
Objectifs de recherche et questions de recherche
1 Parental Leave and Life Satisfaction
1.1 Introduction
1.2 Data and Methodology
1.2.1 Data
1.2.2 Econometric Model
1.3 Empirical Results
1.3.1 Baseline Estimates
1.3.2 Robustness Test
1.3.3 Heterogeneity Analysis
1.4 Discussion
1.5 Appendix
1.5.1 Appendix A: Details on our Data
1.5.2 Appendix B: Parameter Estimates Baseline model
1.5.3 Appendix C: Robustness Test
2 Informal Caregivers and Life Satisfaction
2.1 Introduction
2.1.1 Caregiving and Health Outcomes
2.1.2 Caregiving and Life Satisfaction
2.2 Data and Summary Statistics
2.2.1 Data
2.2.2 Descriptive Statistics
2.3 Empirical strategy
2.4 Empirical Results
2.4.1 Baseline Estimates
2.4.2 Heterogeneity Analysis
2.5 Robustness Tests
2.5.1 Sensitivity Analysis: Selection Bias and Propensity Score Matching
2.5.2 Alternative Definitions and Specifications
2.6 Conclusion
2.7 Appendix
2.7.1 Appendix A: Merge Procedure
2.7.2 Appendix B: Definitions of Variables
2.7.3 Appendix C: Parameter Estimates Baseline Model
2.7.4 Appendix D: Robustness Test
3 Working hours mismatch and subjective well-being
3.1 Introduction
3.2 Literature Review
3.3 Research Method
3.3.1 Data Description
3.3.2 Empirical Strategy
3.4 Results
3.4.1 Working-time Mismatch and Subjective Well-being
3.5 Robustness Test
3.5.1 Heterogeneity Analysis
3.5.2 Sensitivity Analysis
3.6 Discussion
3.7 Appendix
3.7.1 Appendix A: Overview of Studies onWorking Hours Mismatch and Well-being
3.7.2 Appendix B: Merge Procedure
3.7.3 Appendix C: Definition of Variables
3.7.4 Appendix D: Detailed Results
General Conclusions
Bibliography
