Working Papers

Religious Media, Conversion and its Socioeconomic Consequences: The Rise of Pentecostals in Brazil (with Giulia Buccione)

Presented at: NEUDC 2023, Summer Meeting Econometric Society 2023 (Los Angeles), European Economic Society 2023 (Barcelona), RIGDE 2022, ASREC 2023 (Cambridge, MA), ASREC 2023 (Melbourne), ASREC 2022 (London), NEUDC 2021, LACEA 2021, SBE 2020

We investigate the role of religious media in promoting new religious movements and shaping socio-economic outcomes. Focusing on the rapid growth of Pentecostalism in Brazil, we use quasi-random variation in exposure to a church-affiliated TV channel to estimate its impact. Exploiting the placement of transmitters prior to the channel’s religious affiliation, we find that exposure increased Pentecostal affiliation by 30%. Consistent with the church’s conservative gender norms, municipalities exposed to the channel experienced higher fertility rates, lower female labor force participation, reduced schooling for young women, and increased support for Pentecostal candidates, with no effects on male employment or education. Event-study analysis reveals that the number of Pentecostal churches expanded following the channel’s introduction, underscoring the role of media in driving both religious and social change.

Free Childcare and the Motherhood Penalty: Evidence from São Paulo (with Joao Garcia and Rafael Latham-Proença) 

Presented at: NEUDC 2022, LACEA 2022, GeFam Workshop

Latin America consistently has some of the largest child penalties for female work globally, and while subsidized childcare is often advanced as a remedy, the literature on its effectiveness is scarce in this context. This paper estimates the impact of a rapid expansion of public childcare on mothers’ careers in the city of Sao Paulo. We leverage the rollout and expansion of childcare facilities, coupled with detailed data on the labor market and household characteristics, to identify effects on mothers’ labor force participation and earnings. Using a difference-in-differences approach, we compare the child penalty in districts that experienced a large and rapid expansion of childcare with districts with no significant expansion. Our results show that an additional seat per child leads to an increase of 6.7 p.p. (20%) in the mothers’ formal employment after the first child’s birth. We do not detect any effect of this expansion on mothers-to-be or fathers. Effects are stronger for low-education mothers and in areas with more women as household heads.

Child Disability and Parental Labor Supply (with João Garcia, Rafael Latham-Proença, and Joana Silva) Submitted

Having a child with a severe congenital disability deeply impacts family life, yet evidence on the economic effects remains limited. This paper examines the impact of a child health shock causing lifelong disability on household labor market outcomes and fertility in Brazil. Using an exogenous shock to disability incidence—the Zika virus epidemic, which caused a surge of microcephaly cases leading to lifelong disability in children—we find that mothers of affected children reduce labor force participation, including informal work, while fathers’ informal employment is reduced . The shock increases the motherhood penalty by 66 percent, and informal employment does not buffer the effects on formal employment. Brazil’s disability payment system supplements income but intensifies maternal employment losses. We find no effects on fertility or partner separation. Results suggest that child health shocks and pandemics have persistent economic impacts beyond immediate health conditions and that, in developing countries, informality and welfare programs shape these impacts. 

Work in Progress

Religious Mayors, School-Based Polices, and Teenage Pregnancy (Draft coming soon!)

Ideological commitments often shape public policy, sometimes overriding empirical evidence and generating unintended consequences. In Brazil, the growing political influence of Pentecostal groups has led socially conservative mayors to intervene in public school curricula and policies. Using a regression discontinuity design based on close mayoral elections, this paper estimates the causal impact of these school-based interventions on adolescent outcomes by tracking cohorts of girls through their teenage years. I follow girls aged 8-10 at each election—students who would be in middle school during the mayor's term—measuring their subsequent birth rates, STD incidence, and educational attainment. Cohorts exposed to Pentecostal mayors during middle school experience birth rates 2.86 per 1,000 higher (10\% increase) and elevated STD rates, while slightly older cohorts who passed through school before policy implementation show no effects (placebo test). Effects are concentrated among school-enrolled populations and absent in state schools where mayors lack control, indicating that policy changes in municipal schools drive the results. Municipal schools become 12 percentage points less likely to offer sexual education programs after Pentecostal victories, though this measure captures only one observable dimension of school-based changes.