Short summary
Fathers in Sweden, as elsewhere, take less parental leave than mothers. Several countries have tried to generate incentives to increase men’s share of leave. Evidence from Sweden suggests that the answer lies in the cultural change that occurs in response to changing incentives. In 1995, in an effort to encourage fathers to take more of the leave, Sweden introduced a first “daddy month,” i.e., a pay-related month of leave reserved for each parent. This “daddy month” reform was followed by the introduction of a second “daddy month” in 2002 and a third in 2016.
How did the introduction of the second daddy month affect the allocation of parental leave time between parents? We model parental leave as a joint decision between parents, balancing financial costs with each parent’s value of spending time with their newborn. We take into account i) that there may be future wage penalties associated with taking more than the “normal” amount of parental leave for one’s gender; ii) the childcare costs associated with different lengths of leave; and iii) the preference for parental leave time, while varying with the individual including their gender and education, also depends on the social norm. The social norm changes as peer behavior changes.
Using Swedish data, we focus on first births and compare behavior in a pre-reform period (1998–2001) to that in a post-reform period (2008–2011), that way allowing enough time for culture to change. Between the two periods, the parental leave uptake of fathers rises, while that taken by mothers falls. Importantly, this is not solely a result of men who would have previously taken no leave or only one month of leave changing their behavior. There is also an increase in the proportion of men who take more than the earmarked two months.
We use our model to identify the drivers of these changes and find that shifts in social norms account for the largest share. Finally, we use the model to investigate the effect of various policies on fathers’ leave uptake. Equalizing men’s and women’s wages helps, but has a small effect. Perhaps surprisingly, subsidizing childcare in the first year (when public childcare is unavailable) has almost no effect. A policy that gives each parent a much larger non-transferable leave entitlement of six months, by contrast, has a sizable effect increasing the fathers’ share of overall parental leave to over 40 percent.
Our results should not be taken to imply that economic incentives are irrelevant. On the contrary, they provide the impetus that then leads to cultural change.
Key Findings
- Fathers’ leave increased between the pre- and post-reform periods. Mothers’ leave fell over the same period.
- As shown in the model and also true in the data, the immediate effect of the reform on the division of parental leave was small.
- Changes in norms explain the lion’s share of the changes in leave-taking behavior between the pre- and post-reform periods.
- Policies that increase the number of earmarked months of leave have the potential to increase fathers’ share of parental leave substantially.
Relevance Today
Many countries are expanding “use-it-or-lose-it” parental leave for fathers. The EU’s work-life balance directive, for example, requires each parent to have paid leave with a non-transferable component. Reforms that provide incentives that affect behavior can also transform social norms with respect to parental roles magnifying considerably the large long-run effects of these reforms.
Author Quote
“The reform did not just change economic incentives, it also changed cultural norms. Once some fathers took more leave, increasing leave became more attractive for other fathers.”
Reference: Albrecht, Edin, Fernández, Lee, Skogman Thoursie, and Vroman (2026), “Parental Leave: Economic Incentives and Cultural Change” (RFBerlin Discussion Paper 038/26)
Research summary
“Daddy months” are an important feature of the Swedish parental leave system. These are months of parental leave that are reserved for each parent and cannot be transferred between them. Furthermore, they are highly paid months, i.e., they count towards the total number of parental leave months that are reimbursed at a high rate. These reserved months are intended to give fathers an incentive to assume a greater share of the available leave time since a father who fails to take his daddy month(s) is effectively giving up months that would be reimbursed at a high rate (around 85% of monthly wages) without requiring him to be at work.
We study the effect of the 2002 daddy-month reform. This reform increased the number of daddy months from one to two while simultaneously increasing the total number of highly paid months from 12 to 13, thereby permitting women to still take 11 months of highly paid parental leave if they wished. Comparing the behavior of new parents in a period before the reform (1998-2001) to that of new parents after the reform (2008-2011), we find that the share of leave taken by fathers increased markedly (from 15.4% to 24.0%), and that the fraction of fathers who took no leave fell substantially (from 24.2% to 12.9%). Women decreased their months of parental leave (on average by 8.2%). Our objective is to understand the forces behind these and other changes in leave-taking behavior. How much can these changes be understood as a direct result of the additional daddy month? How important are other economic factors? And, to what extent do changes in social norms resulting from the reform drive the patterns that we see?
Our main finding is that social norms play the dominant role. While the additional daddy month and other economic factors did affect behavior, their direct impact was relatively small. The more significant driver of behavior change was the evolution of social norms. This can be understood as a “snowball effect:” As fathers in the current year are induced to take more leave, fathers in the next year adjust their views about what is “normal.” This effect accumulates over time, giving rise to a total effect of the reform on parental leave. Without this channel, the effect of the reform would be much smaller. Although our results are specific to Sweden, the idea that relatively small direct effects of policy on leave-taking behavior can generate significantly larger effects via a change in social norms should be applicable in other countries that want to move towards greater gender equality in childcare responsibilities. It also has implications for reforms that create incentives to change other behaviors (e.g., smoking, contraceptive use, or littering).
To understand these mechanisms, we use a model in which parents jointly decide how to divide parental leave. In making this decision, they weigh several factors: current and future income, childcare costs, and the value of spending time with their child. Future earnings may be affected if taking longer leave signals lower career commitment. At the same time, preferences for staying home vary across individuals and households. Crucially, these preferences are influenced by social norms, which evolve as people observe the behavior of others.” We estimate our model by adjusting its parameters so that the results of simulating the model closely match the patterns that we observe in the data both before and after the 2002 reform.
We use our model estimates in two ways. First, we decompose the factors that drive the changes in parental leave. For example, for fathers in couples in which both parents have completed three years of university, our simulations closely match the average number of months of parental leave they were taking prior to the reform – 3.14 months (very close to the 3.25 months in the data). After the reform, they took 4.61 months (4.54 months in the data). Holding other economic factors fixed and keeping social norms at their pre-reform level, the reform itself only accounts for 8.8% of the average increase in these fathers’ leave time. If, in addition to the direct effect of the reform, we also allow income related variables (wage parameters, wage penalties, and childcare costs) to take their post-reform values, we can explain 20.4% of the change in average leave time.
Next, we look at the effect of social norms on their own, i.e., holding all income variables at their pre-reform levels. This explains 32.8% of the simulated increase in these fathers’ parental leave time. The remaining part of the increase in these fathers’ average months of parental leave is accounted for by the interaction between the changes in income-related variables and in norms. Similar conclusions regarding the quantitative significance of economic factors on their own relative to social norms apply to women. The conclusion to draw from this exercise – and other related ones in the paper– is that if we neglect the effect of changes in social norms on parents’ choices, we will very much underestimate the effects of parental leave reforms similar to the one we study. Focusing only on short-run changes would therefore lead to a serious underestimate of how effective parental leave reforms can be, and suggests that their full impact may only become visible over time as behavior gradually adjusts.
We next use our estimated model to investigate the potential effects of other policies that may be used to attempt to narrow the gender gap in parental leave. We first study a policy that reduces the high cost of private childcare before age 1 to the same level as that of the relatively inexpensive public childcare which is available only once the child turns one. Perhaps surprisingly, we find that this policy has no significant effect. We next consider an “endowment policy” under which each parent is given 6 months of non-transferable leave (plus one additional month of paid leave that the parents can use as they wish, leaving a total of 13 months of highly paid leave as before). This endowment policy is similar to current policy in Finland and in Iceland. This policy has a large effect on the gender gap in parental leave, increasing men’s share to over 40%.
Conclusion
In summary, the 2002 reform increased fathers’ use of parental leave and reduced that of mothers. However, most of this change was not driven by direct incentives, but by induced shifts in social norms. This highlights a broader lesson: policies can have effects far beyond their immediate incentives by changing what people see as normal. When social norms matter—as they often do in areas shaped by gender roles—policy changes that directly affect a small proportion of individuals can lead to large behavioral shifts in the population.
Disclaimer
The opinions and views expressed in this publication are those of the author(s) and do not necessarily reflect those of the ROCKWOOL Foundation Berlin (RFBerlin). While research disseminated through this series may address policy-relevant topics, RFBerlin, as an independent research institute, does not take institutional policy positions.
Publications in the RFBerlin Research Insights series may represent preliminary or ongoing work that has not been peer-reviewed. Readers are advised to consider the provisional nature of such research when citing or applying its findings.
These publications aim to make scientific work accessible to a broader audience and to encourage informed, research-based discussion. All materials are provided by the respective authors, who bear responsibility for appropriate attribution and rights clearance. While every effort has been made to ensure accuracy and proper acknowledgment, RFBerlin welcomes notifications of any concerns regarding authorship, citation, or intellectual property rights. Please contact RFBerlin to request corrections if needed.
Use of these materials for the development or training of artificial intelligence systems is strictly prohibited.
