Publications:
Rizzotto, J. S., Sims, K.M., and Gibbs, H.K. (2025). Hot tempers: Differential effects of heat and drought on domestic violence. Review of Economics of the Household [link].
Sena, F. S. M. ; França, M. T. A. and Frio, G. S.; Rizzotto, J. S. (2023) Educação financeira é disciplina a ser aprendida na escola? Uma análise por meio do PISA 2018 [link].
Rizzotto, J. S. and França, M. T. A. (2022) Indiscipline: the school climate of Brazilian Schools and the impact on student performance. International Journal of Educational Development, v. 94, p. 102657 [link].
Rizzotto, J. S. and França, M. T. A. (2020) Does Bullying Affect the School Performance of Brazilian Students? An Analysis Using Pisa 2015. Child Indicators Research, p. 1-27 [link].
Rizzotto, J. S.; França, M. T. A. and Frio, G. S . (2019) Os arranjos familiares importam no momento de decidir em qual rede de ensino matricular os filhos?. Revista brasileira de estudos de População (REBEP), v. 35, p. 1-27 [link].
Selected working papers:
"How to police intimate partner violence against women? New lessons from women’s police stations in Brazil" with Shoshana Grossbard and Marco Túlio A. França [link]
Abstract: This study investigates how geographic access to specialized Women’s Police Station (WPS) relates to whether episodes of violence against women are classified as intimate partner violence (IPV) in Brazil’s health system records. Using national administrative notifications (2010–2019) geocoded to health facilities and linked to WPS locations, we analyze 227,172 women aged 18–59 who identified a male perpetrator. Logistic regressions assess IPV correlates, with results presented as descriptive associations. Three sets of findings stand out. First, and most importantly, distance from a WPS and IPV are related and the association between distance to the nearest WPS and IPV classification varies by region: in the South and Midwest, the odds of IPV classification decline with distance, while in parts of the North and Northeast they rise with distance. Our uncovering that national averages conceal substantial geographic variation in access and reporting is a novel finding that has policy-relevant implications. Second, situational markers strongly predict IPV classification: episodes at the victim's residence, recurrent cases, and incidents involving an intoxicated perpetrator are much more likely to be classified as IPV, whereas weekday and daylight reports show lower odds. Third, violence concentrates among younger, Black, and less-educated women, while state capital cases are less likely classified as IPV, reflecting different urban service pathways. These findings indicate that specialized policing infrastructure correlates with health reporting patterns in complex ways. Aligning health services with policing infrastructure, particularly addressing alcohol-related cases and regional coverage gaps, may contribute to reducing violence against women.
"In God we trust: holy grounds, hidden wounds" with André L. S. Chagas.
Abstract: This paper studies how the expansion of Pentecostal and Neo-Pentecostal churches affects domestic violence outcomes in Brazil. We combine a novel municipality-year panel of religious establishments, built from the universe of registered churches, with administrative records on violence, including hospital reports, hospitalizations, deaths, and hotline calls. To address endogenous church location decisions, we use a Bartik shift-share instrument. An additional Pentecostal church increases female domestic violence notifications by roughly 38.5 cases per 100,000 women and raises calls to the Ligue 180 hotline placed by the victims themselves by about 33.6 per 100,000, while the effects on hospitalizations and on female homicides are small and statistically insignificant. The increase is concentrated in incidents occurring at home, cases involving male aggressors, and recurrent violence. We develop a framework that uses the severity hierarchy across administrative systems to distinguish broad reporting from changes in latent violence. The pattern is difficult to reconcile with either a protective mechanism or a broad, non-selective disclosure channel. It is more consistent with constrained-exit backlash or selective disclosure of serious intimate partner violence below the inpatient-admission threshold. We cannot fully separate incidence from disclosure, but the evidence supports neither a protective nor a purely reporting-driven interpretation.
"The Blue divide: gender wage gaps across blue ecological sectors in Brazil" with Andrea B. Carvalho, Eduarda M. Figueiredo, Wallace P. S. F. de Souza [link]
The blue economy, understood as the sustainable use of ocean resources for economic growth, is increasingly central to coastal development policies. However, its gendered labor outcomes remain poorly understood. This paper examines how wages in oceanrelated formal activities are distributed between women and men across sectors and along the earnings distribution in Brazil’s coastal labor markets. Using linked employer-employee data from the Annual Social Information Report (RAIS) (2006–2023) covering 279 seafront municipalities, we classify formal jobs into five ocean-related sectors and benchmark them against non-ocean activities in the exact locations. We then apply Oaxaca–Blinder and Recentered Influence Function (RIF) decompositions to separate the part of the gender wage gap explained by observable characteristics from the unexplained component. Women earn approximately 4.3% less than men in formal ocean-related jobs, despite having more favorable characteristics, as higher levels of education. The explained component is harmful, indicating that women’s profiles would predict higher wages; the observed gap is therefore entirely sustained by a significant positive unexplained component, consistent with discrimination and unobserved gender-correlated factors. The sector-specific analyses show that unexplained gaps are huge in marine energy and marine transport. Using the RIF decompositions, we find that wage gaps are small at the bottom but widen sharply at the upper quantiles, indicating the presence of glass-ceiling mechanisms rather than sticky floors. These findings show that blue economy development is not distributionally neutral and highlight the need to integrate gender equity explicitly into ocean governance and sustainability agendas.