Back to search

VAM-Velferd, arbeid og migrasjon

The Gender Pay Gap: Firms, Children, and Unpaid Work

Alternative title: Kjønn og lønn: Hva betyr arbeidsplassen, barn og ubetalt arbeid for forskjeller mellom kjønn?

Awarded: NOK 8.0 mill.

Project Number:

326391

Project Period:

2021 - 2026

Funding received from:

Women have lower income than men. The differences are significantly reduced when comparing individuals within the same company and position. However, we know little about why men and women sort themselves into different companies and why there are also wage differences within companies. Our project contributes to the literature on gender inequalities in the labor market. Our first goal is to provide a descriptive overview of how firms contribute to these inequalities. This allows us to describe how much of the wage differences within firms are due to different tasks or promotions, or if it is the case that women sort themselves into lower-paid firms because these firms compensate with non-monetary benefits. Our second goal is to investigate the underlying economic and social causes of gender differences in the labor market. By using the description of dynamics within firms, we can examine if the costs of having children are different for men and women. We can also investigate the role of firm-specific measures such as flexible working hours and specialization in paid versus unpaid work. To understand more about the underlying mechanisms, we also use new survey data that allows us to investigate whether it is preferences or comparative advantages that drive specialization in paid or unpaid work. This has important implications for policy measures. Policy measures may not necessarily be appropriate if preferences drive gender differences, whereas if the difference is due to constraints or different opportunities, it may be conceivable that measures should be implemented. In the first article, we focus on how to estimate the effect of having children. We design a new estimator that combines elements from a standard event-study with instrumental variable estimators and show how these are related. All approaches estimate a 20 percent reduction in income differences between mothers and their partners as a result of having children. At the same time, we find very different conclusions regarding policy implications - relating to whether it is the mother or the partner who drives the difference. By using the standard approach in existing literature - the event study - we find that the entire response comes from the mother, while our approach shows that only a quarter is due to the mother. The difference arises because the assumption of the event-study model - that it is conditionally random when women have children - does not hold.

Women earn less than men. The earnings difference narrows substantially when comparing individuals within the same firm; however, we know little about why men and women sort into different firms, or earn different wages within the same firm. Our project contributes to the literature on gender inequality in the labor market. The first goal is to offer novel descriptive evidence on the importance of within-firm dynamics for gender differences on the labor market. This allows us to examine how much of the gender differences in wages within firms are due to differences in tasks and promotions, and whether women sort to lower-paying firms because these firms compensate with non-monetary amenities. The second goal is to investigate the underlying economic and social causes of gender differences on the labor market. To achieve this, we build on our within-firm characterization to understand the differential costs of children across men and women, and investigate the roles of firm-level policies, such as flexible work hours, and specialization in paid versus unpaid work. To push the literature on mechanisms even further, we add novel survey data that allow us to understand whether preferences or comparative advantages drive the specialization in paid versus unpaid work arises. This has important policy implications. Policy interventions may not be needed if differences in preferences are driving the gap, while differences in constraints may call for more active policy responses.

Funding scheme:

VAM-Velferd, arbeid og migrasjon