Research

Journal Articles

1. How Can Progressive Vouchers Help the Poor Benefit from School Choice? Evidence from the Chilean Voucher System

Awarded Annual Best Paper Prize by the Columbia Committee on the Economics of Education

This work considers a major educational reform in Chile that increased vouchers by 50 percent for students in the lowest 40 percent of the income distribution. This increased the revenues that schools received for these students and lowered the relative prices of private voucher schools for eligible parents. I use a national data set to implement a regression discontinuity design (RDD) exploiting that eligibility is a discontinuous function of a socioeconomic ranking. Results reject that eligible students chose schools with higher test scores or socioeconomic status and that these students are doing better than noneligible students in math and language test scores.

2. Long-term effects of grants and loans for vocational education

Journal of Public Economics

In this paper I study the long-term effect of giving low-income, low-performing students access to state grants and loans for vocational schools. I combine 12 years of administrative records on higher education enrollment and graduation in Chile and exploit the fact that access to state grants for 2-year vocational programs and state loans for 2-year and 4-year vocational programs are both a discontinuous function of a student's GPA. Both grants and loans increase a student's probability of ever enrolling in higher education, as well as the number of years that students spend enrolled. Although imprecise, results show a positive effect on graduation. Findings indicate that financial aid can increase degree completion, even when targeted to low-performing students attending vocational education.

3. Walking in your footsteps: Sibling spillovers in higher education choices

with Juan Matta

Economics of Education Review

Awarded Dean's Grant for Student Research

This paper studies spillovers from older to younger siblings in higher education choices. Exploiting discontinuous admission rules generated by Chile's centralized admission system to higher education, we identify strong spillovers in the choice of college. Having an older sibling enroll in a given college increases by 42% the probability of enrolling in that institution. In contrast, we find no effect on major choices. We show evidence suggesting that sibling spillovers in college choices are mainly driven by siblings deriving benefits from attending college together.

4. College peer effects, gender, and STEM

with Juan Matta and Ana María Montoya

Economics of Education Review

IDB Funding: Boosting growth and economic development by reducing gender gaps in the countries of the southern cone

Women remain significantly underrepresented in Science, Technology, Engineering, and Mathematics (STEM) fields. We examine whether the ability of male and female college peers contributes to this gap. Using administrative data from Chile covering 33 universities and over 100 college majors, we exploit variation in peer composition across cohorts within programs (college major-university combinations) to identify causal effects. We track long-term outcomes including graduation, labor market performance, fertility, and marriage. Our results reveal substantial peer ability effects specific to women in STEM. Higher-ability female peers increase women’s probability of graduating from STEM programs, while higher-ability male peers decrease it. In the longer run, higher-ability male peers also reduce women’s employment, earnings, and work experience, while higher-ability female peers reduce fertility. We find no comparable effects for women in non-STEM fields or for men. These findings reveal that the ability of male versus female peers plays a critical role in shaping women’s outcomes specifically in STEM environments.

Working Papers

1. Joining the Old Boys' Club: Women's Returns to Majoring in Technology and Engineering (Conditionally Accepted AEJ: Applied)

with Juan Matta and Ana María Montoya

IDB Funding: Boosting growth and economic development by reducing gender gaps in the countries of the southern cone

We investigate the economic consequences for women of majoring in Technology and Engineering (TE). We leverage discontinuities built in Chile’s centralized college admission system to identify the causal effects of majoring in different fields on graduation, labor market outcomes, marriage, and fertility. We find that majoring in TE does not increase earnings for women, even when the counterfactual is the low-paying fields of Humanities, Arts and Social Science (HASS). In contrast, we find large positive returns to TE vs. HASS for men. These results are likely explained by gendered occupational sorting among TE graduates and by high childbearing costs in TE.

3. Violence at School (Submitted)

with Fernanda Ramirez-Espinoza and Román Andrés Zarate

This paper estimates the impact of violence perpetrated by peers and school staff on student victims. Leveraging unique administrative data from Chile that links reports of school violence to individual educational records, we address longstanding data limitations that have constrained empirical research on this issue. Using a matched difference-in-differences design, we find that exposure to school violence has persistent negative effects: absenteeism increases by 46–64%, grade retention rates double, and both grades and test scores decline significantly, with impacts lasting up to four years. In the longer term, victims are substantially less likely to graduate from high school or enroll in university, with violence perpetrated by adults having more severe consequences than peer violence. Complementary survey evidence reveals that reported incidents are associated with increased perceptions of violence and discrimination, as well as decreases in school belonging and teacher expectations. While these psychological and perceptual effects tend to fade after one year, the adverse educational consequences persist, underscoring how brief traumatic experiences can lead to long-lasting educational disadvantages.

3. Speaking Up: The Labor Market Costs of Reporting Workplace Harassment

with Ana María Montoya

We provide the first causal evidence on the labor market trajectories of workers who formally report workplace harassment, linking the universe of reports filed with Chile's labor authority to administrative employment records. Complainants are positively selected, tenured, upwardly mobile, and disproportionately in managerial positions, and use a channel designed to preserve the employment relationship. It fails: within a year, roughly 70 percent have left the reported firm, at least a quarter through employer-initiated dismissal. Four years later, employment remains 12 to 14 percentage points below that of matched controls and earnings 40 to 49 percent of pre-report earnings below, with no recovery. Workers displaced by mass layoffs, matched on identical pre-event histories, suffer larger initial shocks but recover steadily; complainants never do. Part of this excess cost likely reflects the health toll of harassment, which we document through sickness leave; the evidence suggests a larger part is a response to the report itself, by the reported firm and by the market. Reported firms show no detectable cost on any margin we observe. Women file two-thirds of reports and, conditional on filing, lose more. Reporting mechanisms thus concentrate the costs of enforcement on the workers who use them.

4. If You Build it They Will Come: Evidence of the Impact of a Large Construction of Childcare Centers on Attendance and Maternal Labor Supply

Master Thesis

This paper provides evidence on the impact of a large increase in public childcare centers in Chile between 2006 and 2009. It estimates the causal impact of the program on childcare attendance, maternal labor supply and health related outcomes. Identification relies on a difference in differences strategy, comparing different counties (comunas) that had a different exposure to the program. Results find a significant effect of the program on childcare attendance, one spot constructed per child increases the likelihood of attendance in 0.409 percentage points for infants and toddlers and in 0.184 percentage points for preschool aged children. However, results reject a full take up of newly constructed places. There is no evidence that this is a consequence of people substituting privately provided with publicly provided care. Instead, results indicate that there are a number of spots in public childcare centers that are not being used. In addition, results find a positive effect of the program over nutrition for infants and toddlers with mothers of low educational levels. Results regarding maternal employment find no significant impact of the program over maternal labor supply. Therefore, this paper concludes that the program was successful in increasing childcare attendance, but it is not possible to infer that it had the desired effects over maternal labor supply.

Google Sites

Report abuse