RESEARCH

This paper examines the unintended effects of fast-track, non-prosecution criminal procedures on domestic violence against women. Brazil provides an ideal setting for this analysis, as Law 9.099/95 established the Juizados Especiais Criminais (JECrim), which mandated that all misdemeanor crimes, defined by sentencing thresholds in the Penal Code, be resolved through conciliation hearings rather than formal trials. This nationwide reform aimed to improve judicial efficiency by redirecting minor offenses outside the traditional court system. Although gender-neutral by design, the law applied to many domestic violence cases, particularly minor assaults. We exploit variation in pre-reform misdemeanor assault rates across judicial districts to estimate the impact of the policy on female homicide rates at home. Our difference-indifferences estimates indicate that, over the decade following implementation, a district with average baseline exposure experienced an increase of about 0.16 female homicides at home per 100,000 women, corresponding to roughly a 15% rise relative to the pre-reform mean. The results are robust across specifications and highlight how efficiency-oriented legal reforms can produce adverse spillovers when gender dynamics are not explicitly accounted for.

Early socioemotional skills depend not only on material resources, but also on the psychological stability of caregivers. We study whether maternal victimization due to domestic violence generates spillover effects on children's socioemotional development. Using a randomly selected household-based Brazilian survey with rich socioeconomic information that links partner-behavior-based measures of victimization to standardized assessments of children's skills, we conceptualize domestic violence as a multidimensional and potentially cumulative stressor that may distort early non-cognitive skill formation. Children aged three to five whose mothers experienced domestic violence exhibit socioemotional deficits of 0.6 to 1.2 standard deviations relative to their peers. These deficits are largest for sexual violence, increase with cumulative psychological abuse, and persist even when violence occurred prior to conception, consistent with enduring maternal stress channels. To probe residual selection, we implement an instrumental variables strategy exploiting proximity to specialized Women's Police Stations, yielding effects of similar or larger magnitude for mothers whose victimization risk responds to institutional access.

This study examines the impact of the establishment of women's police stations (Delegacias Especializadas de Atendimento à Mulher, DEAMs) within the framework of a nationwide policy to address violence against women on female homicides in Brazil. The policy expanded the institutional capacity of DEAMs to investigate and prevent gender-based violence, enhancing their effectiveness relative to pre-existing mechanisms. Exploiting the staggered rollout of DEAMs across municipalities, we implement a difference-in-differences strategy with multiple periods that accounts for heterogeneous treatment effects. Our results indicate an average decrease of 13% in the female homicide rate, particularly those inflicted by firearms. We further document heterogeneous effects across victim characteristics, with larger reductions observed among white, more educated, and higher-skilled women. The pattern we find suggests that the spatial distribution of women's police stations is central to explaining these disparities, as police stations are more frequently located in wealthier and predominantly white areas. These findings underscore the importance of accessibility in shaping policy effectiveness and suggest that equitable placement of specialized services is crucial to ensuring protection for all women at risk of violence.

Violence against women remains a critical governance challenge in middle-income countries, where legal frameworks have advanced faster than the territorial reach of protective institutions. This study develops a theoretical mechanism linking organizational design to formal visibility — the probability that a violent episode is captured in an administrative record — through three sequential stages: state accessibility, reception quality, and recording routines. We test this mechanism by examining whether the spatial distribution of specialized institutions shapes health-system violence notifications across the 96 districts of São Paulo between 2018 and 2024. Using spatial autocorrelation methods and the Spatial Lag Model, we find that notifications are strongly spatially clustered and shaped by conditions in neighboring districts. Women’s Houses, which raise all three stages of the visibility mechanism by collocating psychosocial, legal, and social assistance under one roof, are positively associated with higher notification rates, with spillover effects extending beyond the districts where they are located. Women’s Police Stations, by contrast, show no significant association with health-system notifications, consistent with their operation primarily through criminal justice rather than health-system pathways. The most vulnerable districts, concentrated in the city’s southern periphery, systematically lack the institutional coverage they most need. These findings demonstrate that formal policy visibility depends not on the quantity of institutions but on their organizational design and territorial distribution, contributing to theory on administrative burden, organizational architecture, and spatial equity in public administration.

This paper estimates the causal impact of homicide investigation reforms on clearance rates and lethal violence in the Brazilian state of Paraíba. Exploiting the spatial variation in road travel time to specialized homicide units as a source of differential exposure to the 2021 institutional reforms — which standardized a clearance indicator, digitalized court records, and expanded the investigative workforce — we find that proximity to these units significantly increases clearance rates and reduces investigation time. Using a difference-in-differences design with municipality and region-year fixed effects, we estimate that an additional hour of travel time to the nearest unit reduces the clearance rate by approximately 3%. These results provide novel causal evidence regarding institutional reforms strengthening criminal investigation — with implications for public safety policy in Brazil and the broader Global South.