Research
Policy Notes
4 notes
2026
Nota de Política Pública: O que sabemos sobre Transferências Condicionais de Renda e Bolsa Família
Reviews the evidence on conditional cash transfers and Bolsa Família, separating results with strong support, such as consumption, poverty, schooling and preventive health, from outcomes that depend on service quality, local capacity and labor-market transitions.
Nota de Política Pública: saídas do Bolsa Família e emprego formal no município de Bento Gonçalves
Evaluates whether the decline in Bolsa Família beneficiaries in Bento Gonçalves is compatible with local policy rather than a regional trend, using a synthetic control design and formal employment data.
O Brasileiro trabalha menos do que deveria?
Discusses Brazilian working hours in comparative perspective, separating evidence on labor supply from common claims in the public debate.
Nota de Política Pública: Quanto de produtividade precisamos para reduzir a jornada de trabalho?
Estimates the productivity gains needed to reduce working time without lowering output or income.
Publications
15 papers
2026
Environmental Enforcement and Abnormal Stock Returns: Causal Evidence from Brazil
Forthcoming Brazilian Review of Finance (and Accepted XXVI Encontro Brasileiro de Finanças)
Uses a causal research design to estimate how environmental enforcement affects abnormal stock returns in Brazil.
Masking the identification problem: how time aggregation manufactures cross-country evidence — A comment on Tausk & Spira (2025), BMC Public Health
Forthcoming Journal of Comments and Replications in Economics
Shows how aggregating time can create misleading cross-country evidence and obscure identification problems in public health research.
From Association to Explanation: Strengthening Research Design and Causal Reasoning in Accounting Scholarship
Discusses how accounting studies can move from associational evidence toward causal explanations through stronger research design, identification strategies, and clearer claims.
2025
Economic conditions and voting behavior
Examines how local economic conditions shape electoral choices, connecting economic performance and shocks to voting patterns.
Assessing political bias and value misalignment in generative artificial intelligence
Tests whether generative AI systems express political preferences or value profiles that diverge from different groups of users.
Covered by various news outlets worldwide — including the Telegraph, Times of India, and Estadão.
Natural disasters and financial technology adoption
Studies whether exposure to natural disasters changes demand for digital financial services and accelerates fintech adoption.
2024
Government, taxation and economic complexity
Investigates how government activity and tax structures relate to the economic complexity of countries and regions.
More human than human: measuring ChatGPT political bias
Measures the political orientation of ChatGPT responses and compares model outputs with patterns observed in human survey data.
Covered by almost 170 pieces of news coverage within the first five days after publication — including the Washington Post, Telegraph, and Forbes.
2022
Bonus for firearms seizures and police performance
Evaluates whether performance-based bonuses for firearm seizures affect police behavior and public safety outcomes.
Economic complexity and corporate governance
Explores the relationship between the sophistication of local economies and firm-level corporate governance practices.
2021
On the determinants of a stable long-run relationship between energy consumption and economic growth
Analyzes whether energy use and economic growth share a stable long-run relationship and what drives that connection.
2019
A influência do nível de transparência da gestão pública na quantidade de disclosure informacional dos municípios capixabas
Studies whether municipal transparency is associated with the amount of public-sector information disclosed by local governments.
2018
Corrupção e valor de mercado: os efeitos da Operação Lava Jato sobre o mercado de ações no Brasil
Examines how corruption revelations during Operation Car Wash affected stock market valuation among Brazilian firms.
2017
O que se tem pesquisado sobre Sustentabilidade Empresarial e sua Evidenciação?
Reviews the research landscape on corporate sustainability and the disclosure of sustainability-related information.
2016
Relação entre desigualdade de renda e crescimento econômico nos municípios brasileiros: o que mudou na década 2000–2010?
Revisits the relationship between income inequality and economic growth across Brazilian municipalities during the 2000–2010 decade.
Working Papers
5 manuscripts
2026
Como medir o invisível? Guerras, pizzarias do Pentágono e o uso de variáveis proxy em econometria
Explains how proxy variables can help measure otherwise invisible phenomena, while highlighting the risks of weak or misleading proxies.
Regulatory Simplification, Formalization, and Local Economic Activity: Evidence from Brazil
Submitted
Estimates how simplifying local regulation affects firm formalization and economic activity in Brazilian municipalities.
Ideology-Driven Specification Selection in Applied Econometrics
Submitted
Uses a controlled many-analyst experiment with LLM agents to study how ideological priors shape specification choices in applied econometrics.
2024
Charting New Territory: Using ChatGPT to Enhance Survey Instruments for Management Accounting Research
Explores how ChatGPT can support the design, refinement, and stress-testing of survey instruments in management accounting.
2023
Virtual learning, real effects: Does delivery mode influence students' attainment?
R&R Journal of the Knowledge Economy
Estimates whether online and in-person course delivery lead to different student attainment outcomes.
Work in Progress
5 projects
From the Centrality of Macroeconomics to the Expansion of Brazil's Economic Agenda: Computational Evidence from Conjuntura Econômica (FGV/IBRE), 1948–2026
Uses a computational text corpus from Conjuntura Econômica to measure how Brazil's technical economic debate changed across themes over time.
Women as mayors improve neonatal health, but effects are not sustained when they leave
Uses close mixed-gender mayoral races in Brazil to study how female political leadership affects neonatal health and care processes.
Do Health Policy Conclusions Survive Modern DiD? Re-estimating TWFE Studies
Revisits influential health-policy studies that used TWFE difference-in-differences methods and tests whether conclusions hold with modern DiD estimators.
Armed Factions and Local Electoral Competition: Evidence from Metropolitan Rio de Janeiro
Studies how the presence of armed factions in metropolitan Rio de Janeiro shapes local electoral competition and political behavior.
Seed-IV: Using Seeds to Randomize Text Characteristics in LLM Models
Uses random seeds as a source of variation in LLM-generated text to study how text characteristics affect downstream outcomes.