Despite major advances in women’s education and labour force participation, large gender gaps in earnings and career progression remain. A growing body of research shows these disparities largely arise after the birth of the first child. Mothers experience persistent declines in earnings and employment trajectories, while fathers’ careers are affected to a much lesser extent. This motherhood penalty has been documented across many countries and institutional settings (Kleven et al. 2019, 2019a, 2019b, 2025, Casarico and Lattanzio 2023, Glogowsky et al., 2025).

The policy debate around how to reduce these penalties has intensified in recent years. Governments have expanded parental leave schemes, childcare provision, and financial support for families. Yet an important dimension of the problem concerns how work itself is organised. Traditional workplace arrangements based on fixed schedules and physical presence can make it difficult to reconcile paid work and childcare responsibilities, especially for mothers of young children.

The rapid diffusion of work from home (WFH) following the Covid-19 pandemic introduced a potentially important change. Working remotely may help parents — and particularly mothers — sustain their work input by reducing commuting time and increasing flexibility in daily schedules. At the same time, concerns have emerged that work from home could reinforce traditional gender roles within households or reduce career opportunities if physical presence remains associated with commitment and productivity.

In recent research (Basso et al. 2026), we study whether the expansion of remote work in Italy mitigated the motherhood penalty for mothers and whether fathers’ access to flexible work arrangements also matters for women’s labour market outcomes.

Measuring working from home and the motherhood penalty

Our analysis combines several administrative datasets covering workers and firms in the Italian private sector between 2012 and 2024. We use social security records to reconstruct earnings and employment histories and combine them with data identifying births and family links between mothers, fathers, and children. These data allow us to follow women’s and men’s labour market trajectories before and after childbirth with a very high level of precision.

We also exploit administrative information on work-from-home contracts collected by the Italian Ministry of Labour. Although the records do not report worker identifiers, they can be matched to social-security employment records within firms using year, gender, age, and citizenship. We then assign workers the observed probability of holding a WFH contract within very fine-grained firm-socio-demographic cells.

Estimating the causal effect of work-from-home arrangements is not straightforward because workers with access to it differ systematically from those without. Those who can work from home are more likely to be employed in high-skilled occupations, in larger firms, and in sectors with better career opportunities.

To address this issue, we exploit the uneven expansion of work-from-home arrangements across sectors and local labour markets after the pandemic. Some jobs were inherently more suitable for remote work before Covid-19, while others required physical presence. Moreover, its diffusion depended partly on the quality of local digital infrastructure.

We therefore construct an indicator of ‘remotability’ that combines two factors: the pre-pandemic share of occupations that could be performed remotely in each sector, and local internet quality measured before the pandemic. This measure captures firms’ ability to adopt work-from-home arrangements for reasons largely unrelated to individual workers’ choices.

Importantly, before 2020 mothers employed in high-remotability and low-remotability contexts followed very similar labour market trends, strengthening the interpretation that post-pandemic differences are driven by the expansion of work from home opportunities.

Remote work substantially reduced mothers’ earnings losses

Our findings show that work from home significantly mitigated the earnings losses associated with childbirth. After 2020, mothers employed in jobs compatible with remote work experienced substantially higher earnings in the year following childbirth relative to mothers in less flexible ones (Figure 1, panel a).

The estimated effects are economically meaningful. Using an instrumental variable approach that isolates the causal effect of actual work from home adoption, we find that access to WFH contracts offsets about 77% of mothers’ earnings losses around childbirth compared with similar mothers who cannot work remotely. These gains persisted in subsequent years. Part of this effect reflects stronger labour market attachment. Mothers in more flexible jobs worked more weeks during the year following childbirth and relied less on parental leave (Figure 1, panel b). We also find important changes in job quality. Mothers with greater access to remote work were more likely to remain in full-time employment and less likely to switch to part-time contracts after childbirth.

Interestingly, remote work did not significantly affect the probability of remaining employed per se. Most mothers in our sample kept their jobs regardless of the availability of WFH arrangements. Instead, remote work mainly affected the intensity and continuity of employment, limiting reductions in hours worked and earnings after childbirth.

The positive effects are particularly strong among younger mothers and among women in the lower part of the earnings distribution. This suggests that flexible work arrangements may be especially important for workers facing tighter time and financial constraints. The effects are also larger for women who commute across municipalities, highlighting the importance of eliminating commuting time in easing work-family conflicts.

Figure 1 Labour market outcomes of mothers

Notes: This figure reports reduced-form coefficients and their 95% confidence intervals from a difference-in-differences model on repeated cross-section of mothers with different exposure to remotability. Panel A considers earnings outcomes: annual earnings in the year after childbirth, and the change in annual earnings between one year after and two years before childbirth. Panel B considers labour supply outcomes: total full-time-equivalent weeks worked in the year after childbirth, the change in full-time-equivalent weeks between one year after and two years before childbirth, and weeks of parental leave. 

Fathers’ flexibility matters too

One particularly interesting finding concerns the role of fathers. Consistent with the broader literature, we find no evidence of a ‘fatherhood penalty’. Fathers’ earnings trajectories remain largely unchanged after childbirth, leaving little room for remote work to improve their own labour-market outcomes. However, fathers’ access to remote work has important indirect effects on mothers’ labour market outcomes. Mothers whose partners work in highly remote-compatible jobs experience significantly smaller earnings losses after childbirth. The magnitude of this indirect effect is surprisingly similar to the direct effect of mothers’ own access to remote work (Figure 2).

These results suggest that the child penalty is not simply an individual-level issue affecting mothers alone. Rather, it reflects constraints operating at the household level. Flexible work arrangements appear to modify how couples allocate time and caregiving responsibilities within the family, improving mothers’ ability to remain attached to the labour market.

This finding resonates with broader evidence showing that gender inequalities within households play a central role in shaping labour market outcomes. Policies aimed at reducing the motherhood penalty may therefore need to consider not only mothers’ working conditions but also fathers’ ability to participate more actively in childcare.

Figure 2 Earnings of mothers, exposure to both own and father’s remotability

Notes. This figure plots reduced-form coefficients and their 95% confidence intervals from a difference-in-differences model on repeated cross-section of mothers with different exposure to own and father’s remotability. The outcome is the change in earnings between the year after and two years before childbirth.

Work from home and fertility decisions

Our analysis also examines whether remote work affects fertility choices themselves (Davis et al. 2026). We find WFH arrangements are associated to an increase in the probability that women already with children have an additional child. It also raises the likelihood that childless women become mothers for the first time. The estimated effects are modest but economically meaningful given the persistent decline in fertility observed in Italy and many other advanced economies: a back-of-the-envelope calculation suggests that the observed WFH diffusion raises the total fertility rate by 1.2%, a figure that would increase to 3.9% if all ‘remotable’ jobs adopted WFH contracts.

Figure 3 Fertility outcomes and number of mothers, reduced-form estimates 

Notes. Panels A and B report reduced-form coefficients and their 95% confidence intervals from the estimation of a difference-in-differences model combining samples of women before and after the pandemic. Panel A is estimated on a sample of women working in 2015 who, by the end of 2016, had at least one child aged five or younger, and of women working in 2019 who, by 2020, had at least one child aged five or younger. Panel B is estimated on the corresponding sample of first-time mothers. For both groups, a balanced panel is constructed to track fertility outcomes over the five years following the baseline. The outcome is the cumulative number of children.

Implications for policy

The long-term consequences of work from home remain uncertain, and many firms are currently reassessing post-pandemic workplace arrangements. Some companies have introduced return-to-office mandates, especially in the US, while others continue to adopt hybrid models. 

Our findings suggest that these working arrangement decisions may have broader consequences than previously recognised. Work from home is not simply a matter of convenience or productivity. It may also affect gender equality, labour market participation, and demographic trends.

The results also speak to fertility policy. Public discussions about declining birth rates often focus on financial incentives or childcare provision. Our evidence suggests that workplace organisation is an alternative, effective lever for reducing the cost of having children: when jobs make it easier to combine work and care, family formation decisions may change.

Of course, remote work is not a universal solution. Many occupations cannot be performed remotely, and flexible work arrangements may also carry risks for career progression if workplace norms continue to reward visibility and in-person presence. Moreover, remote work alone cannot replace investments in childcare services or reforms to parental leave policies. Nevertheless, our evidence indicates that greater flexibility in work organisation can substantially reduce one of the most persistent drivers of gender inequality in labour markets: a counterfactual exercise suggests that current WFH adoption reduces the lifecycle widening of the gender earnings gap by about 11%, rising to nearly 30% if all remotable jobs offered formal WFH arrangements. In this sense, the post-pandemic expansion of work from home may represent not only a technological transformation, but also an opportunity to rethink how labour markets can better accommodate family life.

Authors’ note: The views expressed in this column are those of the authors only and should not be attributed to the institutions with which they are affiliated.

References

Basso, G, M De Paola, S Lattanzio and M Paradisi (2026), “Workplace Flexibility and the Motherhood Penalty: Evidence from the Diffusion of Remote Work”, CEPR Discussion Paper 21486. 

Casarico, A and S Lattanzio (2023), “Behind the child penalty: Understanding what contributes to the labour market costs of motherhood”, Journal of Population Economics 36(3): 1489–1511.

Davis, S, C G Aksoy, J M Barrero, N Bloom, K Cranney, M Dolls, P Zarate (2026), “Remote work can blunt the fertility decline”, VoxEU.org, 25 March.

Kleven, H J, C Landais and G Leite-Mariante (2025), “The Child Penalty Atlas”, The Review of Economic Studies 92(5): 3174–3207

Kleven, H J, C Landais, J Posch, A Steinhauer and J Zweimüller (2019), “Child Penalties Across Countries: Evidence and Explanations”, American Economic Association: Papers and Proceedings, forthcoming.

Kleven, H J, C Landais, J Posch and A Steinhauer (2019a), “Child Penalties Across Countries: Evidence and Explanations”, VoxEU.org, 14 May.

Kleven, H, C Landais and J E Søgaard (2019b), “Children and gender inequality: Evidence from Denmark”, American Economic Journal: Applied Economics 11(4): 181–209.

Glogowsky, U, E Hansen, D Sachs and H Lüthen (2025), “The rising cost of motherhood in Germany”, VoxEU.org, 23 May.

Biasi, P and M De Paola (2025), “The role of parental leave policies in mitigating child penalties: Insights from Italy”, Economics Letters 253: 112355.