The ‘child penalty’ significantly reduces women’s lifetime earnings and pension savings, but whether this is the result of deliberate forward-looking decisions remains unclear. This column uses survey responses and an experiment with Swiss public-school teachers to examine the role of information constraints in women teachers’ labour supply decisions following childbirth. Long-term financial factors are not top of mind when women make their labour supply decisions. Providing objective information about the long-term costs of reducing their labour supply, such as by going part-time, increases the labour supply of mothers who were the least informed initially.
The reduction in mothers’ labour force participation and income following the birth of a first child is large and persistent across countries (Cortés and Pan 2023, Kleven et al. 2019, Kleven et al. 2024). This ‘child penalty’ implies profound financial consequences throughout the life cycle: women miss out on a significant portion of their potential lifetime earnings and save less for retirement, making them financially more vulnerable and dependent on the main earner. This has been found to be the key driver of remaining gender inequalities in the labour market in industrialised countries (Bertrand et al. 2018, Kleven et al. 2018). However, research on how mothers make labour supply decisions after childbirth is scarce, and little is known about the factors they consider. In particular, do mothers consciously account for the full long-term implications when deciding how much to work, not only following their first birth but throughout their entire working lives as parents?
In Costa-Ramón et al. (2024), we open this black box and shed light on mothers’ decision-making processes around their labour supply. First, we document through descriptive surveys that long-term financial factors are not top of mind when women are making labour supply decisions. While the overwhelming majority of women do not explicitly take this dimension into account, we also document heterogeneity with respect to the degree to which women are aware of reduced hours being costly in the long run.
Then, we apply these descriptive insights to a large-scale field experiment, in which we test whether providing mothers with information about the long-term costs of part-time work influences their labour supply decisions and financial planning. Our partnership with a regional educational department allows us to link the rich survey-based data to administrative records of the employer, enabling us to study whether and how shifts in intentions expressed in the survey translate to actual changes in labour supply. We observe an increase in employment levels among this group of mothers one year after the intervention, consistent with the intervention providing most novel information for women who underestimate the cost of reduced hours at baseline.
Our main study population consists of Swiss female public-school teachers with children. In Switzerland, where relatively conservative gender norms prevail, part-time work is common for mothers and the child penalty is among the highest in industrialised countries. After childbirth, the vast majority of Swiss mothers return to the labour market part-time. Similarly, female teachers drastically reduce their working hours when having children and do not substantially increase their level of employment later in their careers. This results in lifetime income losses of about 20% and a decrease in pension receipt of around 25% compared to the average male teacher.
Long-term costs associated with reduced employment levels are not top of mind
The results of our descriptive analyses indicate that for the overwhelming majority of women in the sample, long-term financial factors are not top of mind when making labour supply decisions after childbirth. In Figure 1, we show that when asked about the most important factors in their employment decision, 87% of survey respondents do not mention any considerations related to long-term financial aspects, such as financial wellbeing, pension savings, or professional trajectories.
Figure 1 Factors considered by women in their labour supply decision after childbirth
Note: This figure shows the percentage of women who mention a given topic when asked which were the three main factors they considered in their labour supply decision after the birth of their first child.
In addition, using a vignette featuring a part-time teacher, we elicit short- and long-term estimates of the financial consequences of reduced hours. We find that female teachers struggle to correctly assess the lifetime costs of salary and pension losses associated with reduced working hours. In particular, there is substantial heterogeneity with respect to how women think monthly salary translates into monthly pension receipt in old age, with about a quarter of respondents overestimating pension receipt under part-time work. Finally, a majority of women indicate that they are interested in learning the correct numbers.
Field experiment: The role of informational constraints
Based on these insights, we design a field experiment to test whether learning about the long-term financial consequences of prolonged part-time work impacts two key levers to closing the gender earnings gap: financial planning and labour supply. This randomised controlled trial involves 2,350 public school teachers from a German-speaking region in Switzerland. Participants are randomly chosen to receive an informational video discussing the long-term financial consequences of reduced working hours, as well as access to an online tool that allows users to make individualised cost projections based on different (future) employment scenarios. The control group is shown a placebo video with unrelated information.
The choice of school teachers as a study population in our field experiment has several advantages. Teachers’ wages are set according to a deterministic pay scale, and thus, the tool provided to the treatment group can produce accurate individualised projections of long-run financial trajectories. Additionally, cooperation with the studied region’s department of education, a major employer, ensures a sufficient sample size and enables linkage to participants’ administrative records. Finally, we conduct the main intervention at the start of the yearly employment planning amidst a period of teacher shortages, which enables participants to adjust their employment levels flexibly without demand constraints.
We first document that mothers understand the information presented and are able to correctly apply it: treated women are more likely to correctly rank the relative magnitude of long- and short-term financial factors related to changes in employment level compared with women in the control group. This information update also translates into higher demand for financial planning among treated individuals, with treated mothers being more likely to sign up for additional financial information and planning tools.
Importantly, the increased demand for financial tools is driven by mothers who underestimate the long-term financial costs of part-time employment at baseline (in the following referred to as ‘cost-unaware’ women). Figure 2 displays heterogeneous treatment effects on financial awareness outcomes separately by cost awareness at baseline. While the information update (second bar) is of a similar magnitude between the two groups, only cost-unaware women increase their demand for financial tools (third bar).
Figure 2 Treatment effect on financial outcomes
Note: This figure shows the treatment effect on financial outcomes by cost-awareness. First bar: financial index, which aggregates the information update (second bar) and the tools index (third bar). Second bar: information update, measured as correctly assessing the relative magnitude of the financial implications of a labour supply increase using the part-time vignette. Third bar: tools index, measures the willingness to sign up to receive different information materials and resources related to financial planning, including an incentivised sign-up for a financial consultation. All specifications use post-double-selection lasso to determine the set of controls and strata fixed effects.
With regard to employment levels, women in the treatment group report a 3.1-percentage-point higher planned employment level in 10 years as well as a small (insignificant) increase in employment levels for the next academic year. Cost-unaware teachers plan to adjust their labour supply more meaningfully: they report a 3.5-percentage-point increase (6.3% over the mean) in employment level for the next academic year and a 4.5-percentage-point increase 10 years into the future.
We use the linked employer administrative data to assess the impact of the treatment on teachers’ realised labour supply choices one year after the intervention. Figure 3 shows that the change observed in the survey in planned labour supply for cost-unaware mothers (left panel) is mirrored by the changes observed in the administrative data (right panel). Three out of ten cost-unaware women in the treatment group increase their employment level by about half a day per week in the academic year following the intervention and plan to sustain this increase in employment level in the medium- to long-term.
Figure 3 Treatment effect on short-term labour supply
Note: This figure shows the treatment effect on short-term labour supply one year post-intervention by cost-awareness. Left panel: change in next academic year’s planned employment level (wave 1 survey data). Right panel: change in actual employment level, administrative data. P-values for test of equality of coefficients between cost-unaware and aware. All specifications use post-double-selection lasso to determine the set of controls and strata fixed effects.
For those teachers who increase their employment level for a half day per week, this increase is financially meaningful: if sustained, it reduces projected lifetime income loss by 18% and increases occupational pension benefits by 15%. For these women, it also nearly halves the gender pension gap observed between male and female teachers.
These results demonstrate that limited awareness of financial factors plays an important role in women’s employment decisions. Prior studies on maternal labour supply have primarily focused on the institutional determinants of maternal labour market participation (see Olivetti and Petrongolo 2017 for a review). Recent contributions have started to more closely examine the role of mother’s attitudes and perceptions, such as gender attitudes around childbirth (Kuziemko et al. 2018) and beliefs around labour supply and childcare quality (Boneva et al. 2022). To the best of our knowledge, ours is the first paper to isolate the role of informational constraints in mothers’ labour supply decisions.
Our results also provide insights into why efforts to increase mother’s labour force participation may be stalling. While there exists a wide range of family-friendly policies aimed at supporting women’s attachment to the workforce in industrialised countries, families might not fully internalise the long-term financial benefits of an increase in employment, resulting in a muted response to such policies. Emphasising the long-term financial aspect of family policies, such as the provision of childcare expansions and subsidies, could enhance such policies’ effectiveness in promoting women’s participation in the labour force.
Source: cepr.org