Reductions in women’s labour supply after childbirth remain a persistent driver of gender earnings inequality. Using data from West Germany from the 1960s to 2013, this column explores the evolution of the ‘child penalty’ – post-birth earnings reductions for mothers – in Germany. The child penalty increased over time because women are earning more, so the opportunity cost of motherhood has grown. Family policies can affect gender inequality through changing post-birth earnings decisions. Future research should examine how policies influence pre-birth choices and interact with gender norms over the long run.
Reductions in women’s labour supply after childbirth remain one of the most persistent drivers of gender earnings inequality (e.g. Kleven et al. 2018, 2019a, 2019b, 2024, Angelov et al. 2016, Cortés and Pan 2023). Understanding the magnitude and causes of these reductions matters not only for reasons of fairness and equality. Their fiscal consequences are also substantial – especially in times of growing budgetary pressures from defence spending and demographic change.
In a recent paper (Glogowsky et al. 2025), we provide long-run evidence for West Germany using administrative pension data going back to the 1960s. We document the evolution of post-birth earnings reductions for mothers, commonly referred to in the literature as ‘child penalties’. These are defined as the percentage drop in mothers’ earnings at various points after childbirth, relative to what their earnings would have been in the absence of children. While the term may carry the connotation of an externally imposed loss, it in fact captures a combination of personal choices and care responsibilities. We follow the convention in the literature and use the term for clarity and brevity.
Mothers today face larger long-term earnings gaps
Our first key finding is that child penalties have increased over time. Figure 1 illustrates the monthly evolution of maternal earnings around childbirth. In the first two months after birth, earnings drop by nearly 100% across all cohorts – reflecting immediate leave from the labour market. For mothers who gave birth in the 1960s, earnings began to recover sharply after two months, when the job protection and benefits ended: the child penalty dropped to around 60%. It then continued to decline to just under 40% ten years after birth.
Figure 1 Monthly evolution of maternal earnings around childbirth, 1960s, 1980s, 2000s
By contrast, mothers giving birth in the 1980s or 2000s faced much larger penalties during the first three years after birth. This reflects the expansion of parental leave policies during that period, as we address in more detail below. Long-run child penalties also increased: even ten years after childbirth, mothers’ earnings remained roughly 60% below the counterfactual. In sum, post-birth earnings reductions – child penalties – are substantial even 10 years after birth. More surprisingly, they have increased over time.
As overall gender inequality falls, child-related inequality rises
A second key finding concerns the changing role of motherhood in shaping overall gender inequality. Figure 2 shows the earnings gap between men and women aged 25 to 45 from 1980 to 2013. During this period, overall gender inequality declined from 71% to 54% – meaning that in 1980, women in this age group earned 71% less than men on average, compared to 54% in 2013. Despite this improvement, gender inequality in Germany remains high by international standards.
Figure 2 Earnings gap between men and women aged 25-45, 1980–2013
The figure also decomposes overall inequality into two components: child-related and child-unrelated inequality. The former captures earnings differences driven by child penalties – that is, the post-birth reductions in mothers’ earnings. The latter reflects inequality that would persist even if motherhood had no effect on earnings. The striking pattern is that child-related inequality has steadily increased and now exceeds the child-unrelated component. By 2013, women aged 25 to 45 earned on average 34% less than men due to child penalties. As a result, the share of overall gender inequality attributable to children rose from 14% in 1980 to 64% in 2013.
Rising female earnings potential has amplified the cost of motherhood
The German experience stands in contrast to findings from countries such as Austria or Denmark (Kleven et al. 2019, 2024), where the share of child-related inequality in overall gender inequality has increased but child-related inequality itself has remained stable. To better understand the different pattern in Germany, we decompose child-related inequality into three components.
The first component is the child penalty itself: the percentage drop in maternal earnings after childbirth. How given values for child penalties translate into earnings inequality is captured by the second and third components. The second component is the share of mothers in the population – the more women have children, the more weight child penalties carry in affecting gender inequality. The third, more subtle, component is mothers’ counterfactual earnings relative to male earnings – that is, the income mothers would have earned relative to men had they not had children. The higher this potential relative income, the greater the absolute earnings loss from a given percentage reduction.
To quantify the contribution of these three components, we apply a Shapley decomposition. Figure 3 illustrates the results. The share of mothers has declined over time, which by itself would have reduced child-related inequality. However, both the rise in the child penalty and the rise in counterfactual earnings relative to male earnings contributed positively to its increase. As we showed above, reductions in post-birth earnings have become larger over time, directly increasing inequality.
Figure 3 Contributions to change in child-related inequality
More surprisingly, the stronger driver was the third component: the rise in mothers’ potential earnings relative to men. As women have caught up with men in education and employment, the opportunity cost of motherhood has grown – making a given percentage penalty more consequential in absolute terms.
This finding highlights an important tension: rising child-related inequality is not solely a symptom of stagnation or policy failure – it also reflects the fact that women’s economic potential has grown. In this sense, the cost of motherhood has become more visible precisely because opportunities for women have expanded.
Parental leave reforms and their effects
Part of the increase in child penalties over time – especially during the first three years after birth – coincided with expansions in parental leave policy. Until 1979, mothers in West Germany had access to just two months of paid leave with job protection. In the years that followed, both benefit duration and job protection were extended in several steps, with job protection reaching up to 36 months by 1992. To estimate the effect of these reforms on maternal earnings, we use a regression-discontinuity design that allows us to isolate their causal impact.
The reform had significant effects on short-run child penalties. This is illustrated using the 1986 reform as an example in Figure 4, which increased parental leave to ten months, from six. As shown, it led to a marked reduction in child penalties between six and ten months after birth. Our results show that the effect on gender inequality was substantial: roughly 29% of the increase can be attributed to more generous leave entitlements implemented between 1979 and 1992. Notably, the mechanisms at work operate almost entirely through short-run effects – that is, within the first 36 months after childbirth. These reductions were, to a large extent, a direct consequence of the policy design: giving mothers extended time away from work was a core goal. Yet even if intended, such widespread short-run reductions raise measured inequality at the aggregate level.
Figure 4 Child penalty before vs after 1986 increase in parental leave benefits
In 2007, a parental leave benefit reform was implemented. The first major component was a reduction in the maximum benefit duration to 12 months, from 24 months. The second major component was that parental benefits were increased for all mothers who did not have very low incomes before birth. We illustrate the causal effect of this policy reform on child penalties in Figure 5. The reform led some mothers to return to work earlier, which reduced the child penalty between months 12 and 36 after birth. Importantly, we find that this even resulted in a reduction of the child penalties ten years after birth. This implies that the reform successfully counteracted the trend of increasing child-related inequality. Our estimates imply that child-related inequality would have increased by 13.6% more than if the reform had not been implemented.
Figure 5 Child penalty before vs after 2007 reduction in parental leave benefit
Future research: Pre-birth decisions and gender norms
An open question for future research is the extent to which what is typically classified as ‘child-unrelated’ gender inequality actually reflects early decisions made in anticipation of motherhood. Structural models (e.g. Adda et al. 2017) suggest that women sort into occupations before having children based on expected family trajectories – a process captured in the child-unrelated component. This raises a policy question: could more generous and predictable parental leave, by increasing post-birth job security, encourage greater human capital investment beforehand? Future research should aim to disentangle genuinely child-unrelated inequality from that shaped by pre-birth decisions.
Another important area is how policies interact with and potentially reshape gender norms. In our study, for children born after 1992, we observe significantly smaller child penalties in East Germany than in West Germany. Ten years after birth, the child penalty is 56% in West Germany versus 36% in the East. While East Germany retains stronger public childcare coverage, likely a legacy of its socialist past, we also interpret the gap as reflecting different gender norms shaped by the former dual-earner model.
Conclusion
Our findings carry important implications for policy-making, as child-related reductions in mothers’ labour supply become increasingly costly as their earnings potential grows. Our results also highlight how family policies can affect gender inequality through changing post-birth earnings decisions. As outlined, future research should examine how policies influence pre-birth choices and interact with gender norms over the long run.
Source : VOXeu
