Working Papers:

We investigate the effect of increasing female representation in local governments on the provision of different public social services, including childcare, elderly care, and other forms of assistance. We exploit a 2014 reform in Italy that mandated 40% gender quotas in the executive committees of municipalities with more than 3000 inhabitants. To exclude the effect of confounding policies that apply at the same population cutoff, we employ a difference-in-discontinuities empirical strategy. We find that, while the policy was effective in increasing female representation, it did not have an impact on any category of social services expenditures.

In this paper, I analyze the correlation between off-school time use of children and achievement outcomes using unique data on Turkish children of age 8-12. With a rich set of variables, I first show significant gender differences in off-school time allocation with girls spending more time on chores and study and less on play. Then, I estimate an education production function with time use variables as inputs. An extra hour spent on house chores or play is associated with a significant reduction in both math and verbal test scores. I also find a gender math gap: being a male increases math test scores, which is due to both a different use of time and a different return to playtime across gender.

We evaluate the effect of households’ financial distress on the incidence of intimate partner violence (IPV). Financial distress is defined based on the timing of bank closures and the fixed schedule of payments, which allows us to identify months in which a household is likely forced to stretch their finances for longer than expected, because of changes in their regular payment schedule. Exploiting the exogeneity of such occurrences and using monthly records from the U.S. National Crime Victimisation Survey, we show that financial distress significantly increases the probability of women experiencing IPV. Estimates indicate that one more day of financial distress increases the likelihood of IPV by 16-17%. Using daily data on expenditures from the U.S. Consumption Expenditure Survey, we also provide evidence of potential channels behind our findings, i.e., changes in consumption behaviour.

Work in Progress: