The creation and resolution of discrepancies between preferred and actual working hours over the life course

ABSTRACT

This article contributes to the analysis of working hour discrepancies, i.e., under- and overemployment, by exploring how they emerge and resolve with special consideration of the household context. It uses a rich longitudinal data set, the German Socio-economic Panel, for a discrete duration analysis controlling for unobserved heterogeneity. We focus on the most relevant household and job characteristics. Findings suggest that job autonomy plays a crucial role in the creation and resolution of discrepancies. We especially contribute to previous studies by also examining path dependence and find that both the creation and resolution of discrepancies are characterized by positive duration dependence but by negative occurrence dependence.

KEYWORDS:

• Working hour preferences
• working hour discrepancies
• household context
• life course
• working-time arrangements

JEL CLASSIFICATION:

• J21
• J22
• C23
• C41

Acknowledgments

The authors are grateful to Susanne Wanger, Ines Zapf and Johann Ludsteck for helpful suggestions and valuable input. The authors also benefited from comments from participants of the Doctoral Workshop on Applied Econometrics at the University of Strasbourg and the Jour fixe of the Graduate Programme of the Institute for Employment Research (IAB) and the School of Business and Economics of the University of Erlangen-Nuremberg (FAU). Franziska Zimmert gratefully acknowledges the financial support provided by the Graduate Program of the IAB and FAU.

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability

The article uses confidential data from the German Socio-economic Panel maintained by the German Institute for Economic Research (DIW Berlin). It can be obtained free of charge after application at https://www.diw.de/en/soep. Stata do-files can be provided by the corresponding author.

Notes

¹ Marginal and self-employed do not indicate agreed working hours. As these groups are included in the analysis, it is relied on the measure actual working hours.

² We run linear regressions of each potential endogenous variable on all other included covariates. The resulting predictions are used in the second stage.

³ The agreed monthly working hours are generated by multiplying the weekly hours with the factor 4.348.

⁴ Wunder and Heineck (2013) additionally look at the partner’s hour discrepancy and Reynolds (2014) at hour preferences couples have for each other. Partner spillovers, however, are not in the focus of this paper.

⁵ For several robustness checks see the appendix.

⁶ For better readability, these results are not shown in Table 2, but can be obtained from the corresponding author upon request.

⁷ For better readability, these results are not shown in Table 3, but can be obtained from the corresponding author upon request.

⁸ Only for the resolution of underemployment, the adjustment of actual hours tends to gain importance the longer the discrepancy has already lasted.

Additional information

Funding

This research did not receive any specific grants from funding agencies in the public, commercial, or not-for-profit sectors.