Dércio de Assis - Research

Disrupting Violence, Protecting Lives: Strangulation Laws and Intimate Partner Homicides

with Arpita Ghosh, Sonia Oreffice, and Climent Quintana-Domeque

This paper has been presented at: the Exeter Internal Seminar Series, the CEPR Women in Economics Seminar Series, the Gender Issues and Development Workshop at Universite Paris Nanterre, the 34th Annual Meeting of the American Law and Economics Association, the Workshop on Labour and Family Economics (University of York), Universidad de Granada, the University of Aberdeen, the Nottingham Internal Seminar Series, the 1st Workshop on Violence Against Women: Drivers and Policies (IEB, Universitat de Barcelona), and the 4th Catalan Economic Winter Workshop, the Bruges Family Economics Workshop 2025, the ESPE 2025 Annual Conference, the RES 2025 Annual Conference, the EEA 2025 Annual Conference, and the APHEC 2025 Workshop.

Non-fatal strangulation (NFS) is a dangerous form of intimate partner violence (IPV) and a strong predictor of homicide. We compile data on state NFS statutes in the United States and link them to FBI Supplementary Homicide Reports from 1990 to 2019 to estimate their causal effects on intimate partner homicide (IPH) rates. Using a two-stage difference-in-differences estimator that accommodates staggered adoption and heterogeneous treatment effects, we find that, NFS laws reduce female-victim IPH rates by 14% and male-victim IPH rates by 27%, among those aged 18-49, with no detectable effects for victims aged 50–70 or for homicides committed by strangers. Event-study estimates show flat pre-trends and sustained declines following enactment of NFS laws. Using National Incident-Based Reporting System (NIBRS) data and the same identification strategy, we show that NFS laws increase the share of reported IPV incidents that are classified as aggravated assault among female victims aged 18-49—the most exposed to NFS—and the share of aggravated assaults that result in an arrest for both male and female victims aged 18-49. Our estimates are consistent with NFS laws strengthening the criminal justice response to life-threatening IPV and helping to prevent homicides.

Revised version: University of Exeter DP2501, December 2025.

Old version: IZA Discussion Paper 18006, July 2025.

Childhood Poverty and Life Expectancy: Evidence from Brazilian Human Development Units

Revise & Resubmit, Journal of Demographic Economics

This paper has been presented at: 2nd City/Royal Holloway University of London Workshop on Family and Health Economics, ifo Institute Conference on Understanding Socio-Economic Inequalities with Novel Data and Methods (Munich, Germany), 17th RGS Doctoral Conference in Economics (Essen, Germany), CAEN-UFC, University of Exeter, and Human Capital and Economic Opportunity Group (HCEO, University of Chicago) & Centre for Experimental Research on Fairness, Inequality and Rationality (FAIR, Norwegian School of Economics) summer school on socioeconomic inequality.

This paper analyzes data from urban Brazil using supervised machine-learning techniques to shed greater light on the role that childhood poverty plays in child health and current longevity. By examining a unique dataset collected over a 10-year period from thousands of very small, sub-neighborhood-level geographic areas, I document that child poverty measures have higher predictive power than household income and other major socioeconomic variables in forecasting child mortality and current life expectancy. In addition, using a rich dictionary of hundreds of variables and different data-driven specification selections, the machine-learning models reveal that childhood poverty measures are associated with a 1 percentage point increase in child mortality and a 4 percentage point decrease in current survival to ages 40 and 60. These predictions offer further economic insights on the importance of early life circumstances for human development.

NICEP Discussion Paper available here. Slides here.

Health Decentralization, Prenatal Care and Neonatal Mortality: Evidence from a Constitutional Amendment Act in Brazil

with Victor H. de Oliveira

What is the effect of a massive increase in prenatal care take-up? During the 2000s, Brazil boosted the coverage of prenatal visits (7+) by more than a third. Using population-level administrative data from over 41 million births, we propose estimating the causal effect of prenatal visit coverage (7+) on neonatal mortality rates. To overcome common issues related to omitted variable bias, we employ a shift-share instrumental variable (SSIV) design, leveraging the Brazilian Constitutional Amendment Act No. 29/2000 as a source of exogenous “shifts” in local health spending and state-level measures of health decentralization after the law’s full implementation as exposure “shares”. The main results show that a 10% increase in prenatal visits (7+) coverage leads to a 7% decrease in neonatal mortality rate. No significant effects were found regarding post-neonatal mortality, and fewer prenatal visits (≤3) have a substantial positive effect on neonatal mortality. Even considering the possibility that our SSIV may violate the exclusion restriction does not undermine the robustness of our findings. Finally, we discuss potential mechanisms and policy implications.

New draft coming soon! Old draft is available here. Slides here.

Implementation and Evaluation of a Parenting Program in Rural Northeast Brazil: The Case of PADIN

with Flávio Cunha, Rita Almeida, Leandro Costa, and Jimmy Oliveira

The paper describes the implementation and evaluation of a home visitation with community meetings program targeted to parents living in low-income rural areas in the state of Ceara, Brazil. To evaluate the program, the team used the 2010 Census data to identify target areas. We randomly selected about half of these areas to receive the program whose implementation started in June 2017. The remaining areas were assigned to the control condition. In February 2017, our team collected baseline data in both control and treatment locations and we find that both areas are comparable in dimensions of child development and parent-child interaction. The paper also shows the results of the evaluation of the program impacts after one year of the program.

Publications in Portuguese:

de Assis, Dércio (2020). Descentralização e Resultados na Saúde Infantil no Brasil, Estudos Econômicos, São Paulo, vol.50 n.3, p.447-484.
de Assis, Dércio; Linhares, Fabrício (2018). Dinâmica da Pobreza, Mudanças Macroeconômicas e Disparidades Regionais no Brasil, Pesquisa e Planejamento Econômico, Rio de Janeiro, v. 48, n. 2, p.57-96.
de Assis, Dércio; de Medeiros, Cleyber; Nogueira, Cláudio (2017). Extrema Pobreza Infantil, Crescimento e Distribuição de Renda, Planejamento e Políticas Públicas, Brasília-DF, v.48, n.1, p.175-205. (Prêmio de Excelência em Pesquisa Ipea Roberto Campos).
de Assis, Dércio; Linhares, Fabrício; Penna, Christiano (2016). Efeito da Desigualdade no Crescimento Econômico dos Municípios Brasileiros: Uma Análise com base no Jackknife Model Averaging, Revista de Economia, Curitiba, v. 43, n.2, p.1-18.