publications

The direct and spillover effects of large-scale affirmative action at an elite Brazilian university  (with E. Riehl and G. Reyes) -- Journal of Labor Economics (Accepted)

Firm and Worker Responses to Extensions in Paid Maternity Leave (with C. Szerman and V. Neto) -- Journal of Human Resources v.59 issue 3 (2024)

Large motherhood penalties in US administrative microdata (with D. Almond  and Y. Cheng) -- Proceedings of the National Academy of Sciences (July 2023)

Gentrification and the Rising Returns to Skill (with Lena Edlund and Maria Sviatschi) -- Economica, 89: 258-292 (2022)

Versions: latest,  NBER WP 21729 and IZA DP 9502

Press coverage:  The Washington Post, The Atlantic, The Chicago Tribune, Bloomberg View, and The Huffington Post 

Centralized College Admissions and Student Composition (with Christiane Szerman) -- Economics of Education Review, Vol 85 (2021)

Versions: latest, first draft and IZA DP 10251

Press Coverage:  O Globo, Valor Econômico, Folha de São Paulo

Instrumental Variables and the Sign of the Average Treatment Effect (with Azeem Shaikh and Edward Vytlacil) -- Journal of Econometrics v.212 issue 2 (2019) 

Version: latest

Unobserved Selection Heterogeneity and the Gender Wage Gap -- Journal of Applied Econometrics v.32 issue 7 (2017)

Previous version: "Selection, Heterogeneity and the Gender Wage Gap", IZA DP7005

How the Other Half Lived: Marriage and Emancipation in the Age of the Pill (with Lena Edlund) -- European Economic Review 80 (2015) 

Previous versions: "Marriage and Emancipation in the Age of the Pill"; "Pill Power: the Prequel", IZA DP 5468

working papers

It's the Phone, Stupid: Mobiles and Murder (with Lena Edlund) -- NEW (SUBMITTED)

Previous versions: NBER WP w25883 

Homicide rates fell dramatically in the US in the 1990s and stayed low through the 2000s and 2020s. We propose that the mainstreaming of cell phones played an important role in this development through its impact on drug dealing. Cell phones reduced the role of the open street market and thus undermined the capacity of street gangs to cartelize the retail end. With lower gang profits, violence also fell, we propose. To obtain plausibly exogenous variation in cell-phone uptake, we use data on location and year of construction of antenna structures to proxy for network build-out. Analyzing annual county-level data from 1970-2009, we find that service expansion lowered homicide rates. The effects were concentrated to the 1990s and urban counties, and stronger among Black or Hispanic young males; as well as homicide categories more closely associated with criminal gangs.

Better Neighborhoods or Better Houses? The Effects of Housing Policies on Poor Households in Brazil (with Laisa Rachter) -- NEW (SUBMITTED)

This paper evaluates the effects of a housing program that built houses for low-income families from the city of Rio de Janeiro (Brazil). We explore the lotteries used to select the program’s beneficiaries to provide evidence of its effects on location, housing quality, housing costs, and household choices. The program induced households to move to less populated, more impoverished, and more distant neighborhoods. However, it increased the houses’ quality in which these households lived and decreased their housing costs. Increases in other expenditures did not compensate for the decline in housing costs. Furthermore, we find the program did not influence labor force participation and income and weakly increased teenagers’ enrollment. Overall, our evidence contributes to understanding the mechanisms through which housing programs affect well-being.

Too Late for a Change? Youth Unresponsiveness to Cash Transfers (with V. Pinho Neto and Christiane Szerman) -- available upon request

Identifying successful interventions for disadvantaged youth has recently proven challenging. This project exploits an exogenous variation in the provision of cash transfers in Brazil to identify whether an extra exposure to cash assistance can improve long-term economic outcomes of disadvantaged youth using a combination of difference-in-differences and regression discontinuity designs. We combine the universe of recipients of the Bolsa Familia program with several sources of detailed administrative data to construct educational, labor market, and fertility outcomes up to 10 years after the exposure. We find negligible effects of additional program expansion to teens in poverty, confirming optimal intervation time should target them when young.

How and When: Conditional Cash Transfers and Birth Outcomes (with Fernando Mattar and Marina Palma) -- available upon request

Endowments at birth are key determinants of several short and long run socioeconomic outcomes. While conditional cash transfers (CCTs) alleviate poverty and improve wellbeing of beneficiaries, little is known about the relative importance of cash transfers and conditionality compliance for pregnant women and their babies. This research fills this gap by constructing a large, rich and longitudinal dataset with detailed information on birth outcomes and the timing and strength of in utero exposure of babies to the Bolsa Família CCT program. Exploring exogenous variation in income transfers that are due to sharp eligibility requirements related to dates of births of family members, we recover the causal effects of additional cash transfers during gestation on outcomes at birth. Overall, the results point to a null effect of additional income on birth outcomes. However, for women meeting prenatal care compliance, there is a strong and sizable reduction in the incidence of prematurity. Our findings speak to the role complementarity between prenatal care and family income in producing health at birth: even small amounts of cash transfers are effective in improving birth outcomes when coupled with adequate prenatal care.

The Effects of Better Houses on Infant Health (with Laisa Rachter) -- available upon request

This paper examines the effects of better houses on infant health in the context of Brazil’s Minha Casa Minha Vida program, which built roughly 900,000 million houses to poor households in Brazil during the period 2010-2017. We use a regression discontinuity design and administrative data to estimate the program’s effects on health at birth and infant health. We find the program reduced the share of households living in inadequate houses by 18 percentage points. We find this improvement in housing conditions led to increases in birth weight and decreases in infant (before 1 year) mortality caused by conditions originating in children’s perinatal period. We find no effect of the program in children with more than one year. Our results point out the importance of better houses in improving health at birth.

WORK IN PROGRESS

Labor Market Effects of a Credit Crunch (with M. Bonomo & B. Martins)

Gender Gaps in Multinational Firms (w. M. Ferreira, M. Foguel & C. Terra)

Cash and Care ( with T. Bonomo)

Conditional Cash Transfers and Political Behavior in Brazil (with V. Pinho Neto) 

PUBLICATIONS IN PORTUGUESE AND BOOK CHAPTER

tba