Bayesian treatment effects models with variable selection for panel outcomes with an application to earnings effects of maternity leave

Abstract

Standard-Nutzungsbedingungen:Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden.Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen.Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in AbstractChild birth leads to a break in a woman's employment history and is considered one reason for the relatively poor labor market outcomes observed for women compared to men. However, the time spent at home after child birth varies significantly across mothers and is likely driven by observed and, more importantly, unobserved factors that also affect labor market outcomes directly. In this paper we propose 1 We are grateful to J.J. Heckman, Remi Piatek, Siddharta Chib and seminar participants at the University of Melbourne, Monash University, University of Innsbruck, as well as at the ESOBE 2012 meeting, the 5th Rimini Bayesian Workshop, and the Melbourne Bayesian Econometrics Workshop 2013 for helpful comments and suggestions. We gratefully acknowledge funding from the Austrian Science Fund (FWF): S10309-G16 and University of Melbourne.2 Corresponding author. email: Sylvia.Fruehwirth-Schnatter@wu.ac.at 1 two alternative Bayesian treatment modeling and inferential frameworks for panel outcomes to estimate dynamic earnings effects of a long maternity leave on mothers' earnings in the years following the return to the labor market. The frameworks differ in their modeling of the endogeneity of the treatment and the panel structure of the earnings, with the first framework based on the modeling tradition of the Roy switching regression model, and the second based on the shared factor approach. We show how stochastic variable selection can be implemented within both frameworks and can be used, for example, to test for the heterogeneity of the treatment effects. Our analysis is based on a large sample of mothers from the Austrian Social Security Register (ASSD) and exploits a recent change in the maternity leave policy to help identify the causal earnings effects. We find substantial negative earnings effects from long leave over a 5 years period after mothers' return to the labor market, with the earnings gap between short and long leave mothers steadily narrowing over time.

Paid Parental Leave and Female Labour Supply: A Review

“…Jacobi et al . () consider the 2000 reform in Austria which increased the period of paid leave from 18 to 30 months. It is clear that this reform extended the amount of maternity leave taken by a substantial amount.…”

“…Dahl et al conclude that these payments are expensive with little effect and come at a high cost both in absolute terms and in terms of redistributing income to high-and middle-income families. Jacobi et al (2016) consider the 2000 reform in Austria which increased the period of paid leave from 18 to 30 months. It is clear that this reform extended the amount of maternity leave taken by a substantial amount.…”

“…'Too much', and women stay out of the labour market longer making a return more difficult (or less 'profitable', e.g. see Blau & Kahn, 2013;Jacobi et al, 2016). 'Too little', and women are more likely to drop out of the labour force because the end of paid leave does not coincide with the time that women are ready to return to work (Bergemann & Riphahn, 2015).…”

Regularization and Confounding in Linear Regression for Treatment Effect Estimation

“…Our proposed method can be seen as an alternative to Wang et al (2012), with certain conceptual and computational advantages, namely ease of prior specification and posterior sampling. Other papers elaborating upon this approach include Wang et al (2015), Lefebvre et al (2014) and Talbot et al (2015); see also Jacobi et al (2016). Zigler and Dominici (2014) and An (2010) focus on Bayesian propensity score models (for use with binary treatment variables).…”

Parents’ Strategies in Dealing with Constructions of Gendered Responsibilities at Their Workplaces

“…Consequently, mothers continue to have the main responsibility for unpaid care and family work. Recent findings have shown that fathers, even if they have been on parental leave or have worked part-time, experienced significantly fewer negative consequences with regard to career and salaries than mothers, as their leave periods are, for the most part, shorter and their part-time work mainly involves more working hours than mothers' part-time work (Glauber 2019;Bünning 2016;Jacobi et al 2016;Schiffbänker and Holzinger 2014). Furthermore, research from Germany has revealed a strong correlation between company structures and fathers' working hours: when companies are characterized by official, general, and transparent rules, fathers are more likely to work reduced hours (Bernhardt and Bünning 2017).…”