Is Physical Unattractiveness a Risk Factor for Sexual Violence Perpetration? Evidence from the U.S. (with Maria Laura Di Tommaso, Silvia Mendolia, and Giulia Savio). IZA Discussion Paper, No. 18370 (2026).
Abstract. Little is known about whether an individual’s physical attractiveness influences the likelihood of perpetrating sexual violence. Using U.S. data from the National Longitudinal Study of Adolescent to Adult Health (Add Health), we examine the relationship between attractiveness and sexual violence perpetration against one’s partner. Physical attractiveness is measured using interviewer-assigned scores for respondents aged 12 to 17. Among men, a one-point increase in attractiveness (on a 1–5 scale) reduces the likelihood of perpetrating sexual violence in adulthood by 13 percent. We also find that contextual factors, such as parental education and neighborhood cohesion, mitigate this association. Very unattractive men with highly educated parents have substantially lower predicted probabilities of perpetrating sexual violence compared to peers with less-educated parents. Similarly, very unattractive men living in cohesive neighborhoods report markedly lower predicted frequencies of sexual violence perpetration than those in less cohesive neighborhoods.
Children with Disabilities: Parental Labour-Market Penalties Worldwide (with Nicoletta Balbo, Elisabetta De Cao, Francesco De Luca, and Giorgio Nocerino).
Abstract. Today, an estimated 240 million children live with a disability worldwide. Although a child’s disability can profoundly shape parents’ family and economic lives, empirical evidence on its impact remains limited. Using a combination of cross-sectional census data and longitudinal data from 33 countries, we provide the first cross-national comparative analysis of how child disability affects mothers’ and fathers’ employment trajectories. In countries with census data, we implement pseudo–event-study models, constructing synthetic pre-birth employment trends by matching childless individuals to parents of children with and without disabilities on observable characteristics. In higher-development contexts, we document a sizable \textit{child-disability penalty} in maternal employment—mothers of children with disabilities are substantially less likely to be employed than comparable mothers of non-disabled children—while fathers show only marginal effects. By contrast, in lower-development contexts, we find no significant effects for either parent. These results suggest that labour-market structures and broader socio-economic contexts crucially shape parents'—especially mothers'—employment responses to child disability.
Child disability spillovers on schoolmates' adolescent wellbeing. Evidence from U.S. middle and high schools. (with Nicoletta Balbo).
Abstract. Creating an inclusive school environment, where students with disabilities are integrated as much as possible with students without disabilities, has become a central objective in educational policies worldwide. Despite this trend, however, empirical evidence on the effects of exposure to peers with disabilities remains limited. Using data from the National Longitudinal Study of Adolescent Health (Add Health), we examine child disability spillover effects on internalizing and externalizing behavioral outcomes of students in middle and high school (grades 7-12) in the United States. To address potential endogeneity, we employ a quasi-experimental research design and exploit within-school, between-grade variation in peer composition. Our results show that exposure to peers with physical disabilities is reflected in students without disabilities' increased internalizing problems, with no effects on externalizing problems. No spillover effects are found for learning disabilities. These results are consistent with the hypothesis that students without disabilities socialize the emotional and social challenges experienced by their peers with disabilities. We also find that the effects are substantially mitigated in schools where emotional counseling is provided, suggesting that emotional counseling and the factors influencing its provision may play a key role in promoting a healthier and more inclusive classroom environment for all students.
Does timing matter? Parental separation around educational transitions (with Hannelore Nelissen and Kristof De Witte).
Abstract. Parental separation has become increasingly prevalent in recent decades. While prior research focuses primarily on children’s age at separation, it remains unclear whether separation around key educational transitions matters. This paper examines whether parental separation before an early high-stakes educational gateway shapes children’s subsequent educational and early labour market trajectories. Drawing on Dutch administrative data covering all children born between 1995-2011, we leverage a unique dataset of siblings who all experienced parental divorce by age 18 and exploit within-family variation in whether siblings experienced separation before or after the primary-secondary school transition, applying sibling fixed effects to account for time‑invariant family characteristics. Children who experience separation within the five years preceding the gateway score, on average, 0.11 standard deviations lower on standardized tests than their siblings who experience separation later. They are also about 5 percentage points more likely to receive a vocational track recommendation and 2 percentage points more likely to be placed in the vocational track. Separation before this stage is associated with a 1‑percentage‑point reduction in high school graduation and a 2‑percentage‑point decrease in college enrollment. In the short term, however, these children show higher employment levels and wages, consistent with earlier entry into the labor market. These findings underscore the importance of family disruptions around institutional educational transitions and the need for targeted support during these critical periods.