The effects of family policy on labor supply and fertility decisions: Evidence from a dynamic structural micro-economic model with task-specific human capital, estimated using microdata for France and Germany – FamPol

1) Scope and limitations of family policies

Before implementing our empirical study on the impact of family policies, we have investigated the scope and limitations of family policies for explaining individual behavior.
Melindi Ghidi and Seegmuller (2017) explain theoretically how it becomes possible in a dynamic model to explain the existence of situations with different fertility and growth rates, in economies with the same fundamentals, preferences, technologies and initial conditions. They illustrate their findings using data for French departments.

d’Aspremont et Dos Santos Ferreira (2018) use a comprehensive model of strategic household behavior nesting full cooperative and noncooperative regimes as limit cases. They have shown that the internal structure of the household is quite important for evaluating the consequences of economic policy in general and especially family policies.

Aware of the limitations which arise when trying to explain individual choices, we do our best and try to combine using evaluation methods and structural models in order to identify the component of individual choices, which responds to changes in family policies.

2) The French family policy reform of 2014

An important recent reform in the family policy in France took place in 2014 and consisted in conditioning the amount of basic allowances of early childhood benefits on household income. Nelly El Mallakh (2018) investigated the impact of this reform on fertility choices and labor supply, for both women and men. Nelly El Mallakh used a Regression-Discontinuity Design to examine the impact of this sharp discontinuity in the provision of child benefits on fertility and the hours of work of women and men. The analysis relies on data from the Statistics on Resources and Living Conditions in France that has the unique feature of providing administrative information on both income and social benefits. The results suggest that not being eligible to any family allowances for children decreases the birth probability at the household level. The results also highlight that receiving half the amount of the family allowances for children or not receiving any allowances leads to an increase in the number of hours of work per week for both women and men, compared to individuals who are eligible to the total amount of child benefits.

3) The impact of family policies on fertility in Germany and France
Abiry, Reuss et Stichnoth (2015) and Bonin Reuss et Stichnoth (2016) built a structural life-cycle model of fertility and female labour supply and use it to evaluate the effects of a number of key family policy measures based on data for Germany. Parental leave benefits, child benefits and subsidized childcare are found to have substantial fertility effects. Without these measures, completed fertility is estimated to be lower by 6%, 7%, and 10%, respectively. Income tax splitting, which is fiscally expensive, reduces female labour supply but has a negligible effect on fertility.
This approach is applied to France by Koebel, Lanot et Vergnat (2018) and extended for allowing interrelationships between fertility choices and labour supply.

4) The influence of the number of children on women's wages and hours of work

Using French administrative data, Rodrigues and Vergnat (2016) estimate the impact of the birth of a first, second and third child on hourly wages, as well as for hours worked, for both women and men. They compute the impact on these outcome variables, two, four and six years after the birth of the child, and focus on the distinction
between highly educated women and women with a high school degree or less. Rodrigues and Vergnat also take the maternity leave (or paternity leave in case of men) duration into account. Estimation is done with difference-in-differences and they compute bootstrapped confidence intervals. Results show both lower and highly educated women decrease significantly their working hours after the birth of their child. Men are, for the most part, not much impacted by the birth of their children. Maternity leave duration influences the magnitude of the impact of the birth, especially on the hourly wages of educated women.

Previous studies have shown that childrearing has a different impact on a mother’s professional career, depending, among other reasons, on how much time passed from birth to returning to work. Rodrigues and Vergnat (2018) use a competing risks model to determine which variables may explain time out of work, as well as the transition back to work for young mothers in France. In their study, mothers can decide to go back to the same employer, change employer and/or change labour supply. Their results show that it is mostly the age of the mothers at birth, their pre-birth wages, tenure, firm size as well as the state of the economy as a whole that play a large role in the way young mothers go back to work, if at all. This research highlights the key factors on which causal research should be based in order to advise firms and also the state on how to influence mothers’ labour supply behaviors.

These results have been presented at several conferences and some of them are published as follow:

d’Aspremont, C. and R. Dos Santos Ferreira, 2018, “Enlarging the collective model of household behavior: A revealed preference analysis,” Economic Theory, forthcoming.
El-Mallakh, N., 2018, “Rethinking Family Policies: Fertility Choices and the Labor Market in France,” working paper BETA n°2018-30, submitted
Koebel B. and G. Lanot, 2018, “Labor supply and housing demand,” mimeo.
Koebel B., G. Lanot and V. Vergnat, 2018, “Unobserved preferences, labor supply, schooling and the number of children,” mimeo.
Melindi Ghidi, P., 2018, “Inequality, Educational Choice and Public School Quality in Income Mixing Communities,” Journal of Public Economic Theory, forthcoming.
Melindi Ghidi, P. and T. Seegmuller, 2017, “The Love for Children Hypothesis and the Multiplicity of Fertility Rates,” AMSE WP 2017-11, in revision.
Rodrigues B. and V. Vergnat, 2018, “The time and the transitions back to work in France after maternity,” BETA-Document de Travail n° 2018 – 14, revised and resubmitted.
Rodrigues B. and V. Vergnat, 2016, “The impact on wages and worked hours of childbirth in France,” BETA-Document de Travail n° 2016 – 48, submitted.
Stichnoth, Holger and Mustafa Yeter, 2016, “Cultural Influences on the Fertility Behavior of First- and Second-Generation Immigrants,” Journal of Demographic Economics, 82, 281-314.

Although France and Germany are similar in many socio-economic dimensions, their total fertility rates are at the opposite extremes of the spectrum found in the OECD. German family policy has sought to increase the low rate for years with little success. We believe that much can be learned from a systematic comparison that is guided by economic theory, features detailed modeling of family policy and labor market environments alike, and adequately controls for heterogeneity in preferences.
In the proposed project, we develop an estimable life cycle model with endogenous fertility, career, and labor supply decisions. The model is fully forward-looking, so women choose their careers based on their desired fertility level and the costs of career breaks. These costs are shaped by policy through the prices of childcare, maternity leave benefits, and birth-related job protection policies. They are also influenced by the choice of career itself: Foregone returns to experience and human capital depreciation vary with the task baskets associated with different types of jobs. Our model includes the decision to obtain a university degree; in order to get a meaningful distinction between different career paths we develop a task based approach. Heterogeneous preferences for education, work and fertility ensure that we do not falsely attribute differences in outcomes to variation in the institutional setting. We perform extensive model checks, including validation on holdout samples.
Estimating comparable versions of the model for Germany and France allows us to decompose differences in outcomes into differences in policy, the labor market environment, and preferences. We are also able to investigate possible interaction effects between these three sets of explanatory factors. We also use the model to study specific family policy measures in greater detail. Our main application will be the 2007 reform of parental leave benefits in Germany. Unlike the existing quasi-experimental studies, the structural model will allow us to quantify the separate effects of the different components of the reform, to isolate the reform effects from concomitant changes such as the expansion of childcare, and to predict the long-run fertility effects of the reform, distinguishing changes in completed fertility from pure timing effects. We will also use the model to simulate changes to the current system of parental leave regulations and to approximate the efficient frontier of fertility and female labor force participation, while leaving total fiscal cost constant at the present level.