Grants for Graduate Thesis - 2nd edition



In the context of the project "Fostering capacities in Impact Evaluation in Latin America," the Center for Distributive, Labor and Social Studies, CEDLAS with the support of the International Development Research Centre (IDRC-Canada), organize a second call for thesis proposals that use Impact Evaluation techniques for PhD and Master's students based in Latin America.

The selected proposals are:

Amendolaggine, Julian (Argentinean, Maestría en Economía - Universidad Nacional de La Plata) "Efectos de la política de vivienda social. Evidencia para Argentina"

This project aims to assess the impact of social housing policies implemented in Argentina over the last decade within the Planes Federales. In particular, the random rule used for housing allocation in some subnational governments will be exploited to identify the causal effect on subsequent labor market outcomes of beneficiary households, using administrative records of different government agencies. The paper is original because evaluates an intervention for which there is no evidence in the literature about its effects, since it combines granting property rights, improving building materials, giving access to public services network and relocation in a suburb. The first stage of the project will focus on interventions that occurred in the province of Santa Fe, between 2011 and 2013, which had about 3,000 beneficiaries from about 50,000 enrolled in the lotteries.   

Cardoso, Thiago (Brazilian, Master in Economic Theory - University of São Paulo, USP) "Color-Blind Affirmative Action and School Choice"

This research project intends to assess the impact of affirmative actions on school choice. To this end, we explore a Brazilian Law that reserves half of the seats in federal universities to students who were enrolled all high school years in a public institution. Building on affirmative action models and using data from the National Educational Census, we analyze if such policy affects the migration of students from private to public high schools.  A diff-in-diff methodology will be employed using 9th graders from private elementary schools as the treatment group and freshman private high school students as the control group. The Law only affects the migration decision of 9th graders since students who finished one year of high school in a private institution cannot be eligible to the Law even if changing to a public institution in the following year.

Katzkowicz, Noemí (Uruguayan, Maestría en Economía -Universidad Nacional de La Plata) "Cuidados prenatales y salud del recién nacido: evidencia para Uruguay"  

This paper evaluates the impact of the program "Metas Asistenciales" implemented in Uruguay in 2008. The study provides evidence on the effects of prenatal care on the newborn’s health, using data from the Perinatal Information System of PAHO MSP regarding the universe of birth records from the years 2000 to 2013. This program allocated funds to health providers based on indicators about the use and quality of maternal health care services.  Our identification strategy is based on “differences in differences” approach, which exploits the variation in the timing of implementation between the public and the private health sectors. This methodology addresses potential biases due to unobserved heterogeneity, allowing us to estimate the causal effects of the program on birth health outcomes. Our results suggest that this policy had positive effects in terms of efficiency, as shown by the amount of prenatal care and other newborns’ health patterns measured by birth weight, percentage of children with low birth weight, neonatal mortality and pre-term births.  

Mariño Fages, Diego (Argentinean, Maestría en Economía - Universidad de San Andrés) "Efecto de pares en el desempeño académico de alumnos de secundaria"    

The objective of this project is to evaluate the existence of peer effects on the academic achievement of secondary school students. As it is widely stated in the literature, the fundamental difficulty in measuring it is how to disentangle the effects of friendships from the student’s decision to choose some friendships over others (i.e., the problem of self-selection). To overcome this difficulty, we construct a unique panel data set of students’ reported friendships before and after a natural experiment occurs. The experiment consists of reshuffling primary school students into different groups when they enter secondary school, which generates an exogenous variation in students’ peers and friends. Due to this exogenous variation and a particular fact from the networks’ literature, we can adopt an IV strategy. Additionally, the panel feature of the data allows us to control for fixed effects at the individual level. This approach will enable us to identify the causal effect of friendships on achievement.

Muñoz Pedroza, Carlos Martín (Mexican, Maestría en Economía - El Colegio de México) "Deserción en escuelas de doble turno: diseño de regresión discontinua"    

Double-shift schooling (DSS) is an educational policy widely used in developing countries to maximize educative coverage while facing strong budget constraints. There exist concerns that minimizing the cost of public education —through the use of DSS— may increase inequality between students of both shifts. This research proposal, is aimed to examine the potential effects to being assigned to the afternoon shift in the students’ performance, especially, on the decision to permanently leave school and on the probability to be admitted to college.
Analyze the impact of DSS is of considerable relevance, if they do not have a negative impact on students’ performance, they could be replicated by other countries —which, as Mexico, have budgetary difficulties— without fear of diminishing educational quality; but if they increase dropout rates, this can result in an efficiency loss, arising from the lessening of potential human capital and truncating education of students whom —under different conditions— would finished high school.
A regression discontinuity design (RDD) is used to examine students’ responses to shift assignation. This is analyzed through three different mechanisms that could have an impact on student’s performance: 1) the reaction of students while facing a negative incentive, 2) the existence of peer-effects and tracking effects, 3) and the influence of the relative ranking of a student on his motivation and self-confidence. The RDD exploits the shift assignation rule of an important high school in México —which is based in the middle school average grade— and a rich dataset with socioeconomic and administrative information of eight cohorts of students.
The main preliminary conclusion is that, the assignment to the afternoon shift have an impact over dropout. The most striking characteristic of the results is the heterogeneity of the responses between male and female. Assignment to the afternoon shift decreases the probability to dropout for men, while increases the probability for women. The impact is sizable, one of every ten female students dropout just because she was assigned to the afternoon shift, and the opposite occurs for male students.


Ospino, Carlos (Colombian, Doctorado en Economía - Universidad de los Andes) "The role of size-dependent policies on firm outcomes: The Colombian apprenticeship contract"

The current proposal exploits exogenous variation in regulation in Colombia to identify the effects of the compulsory use of apprenticeship contracts.  Hiring apprentices has been mandatory in Colombia since 1960; Law 789 of 2002 modified the criteria with which the mandatory apprenticeship quota is calculated, and the type of apprentices that can be hired.
The paper attempts to provide evidence on an issue which has been at the heart of the discussion of businesses for years in Colombia, but that is also of interest beyond the Colombian case; how do apprenticeship regulations affect firm performance? Under the assumption that the regulation change is exogenous to firm’s decisions the effect of being subject to the regulation on outcomes such as the share of directly hired workers, output, investment, firm and sector productivity can be answered. In particular, the workforce already set at the time of the reform will be assumed to be exogenous to the regulatory change, and used to indicate treatment status. The data used for the project comes from Encuesta Anual Manufacturera (EAM) for the years 2000-2005. The data are proprietary, administered by the National Statistical agency (DANE). They must be accessed on-site at DANE. To evaluate the impact of the Colombian learning contract on firm outcomes, the most suited methodological approach is a difference-in-difference (DD) estimation. In addition to the previous methodology, the discontinuity introduced by the 2002 reform generates variance in the share of apprentices which may be exploited under a control function estimation. The effect of switching from control to treatment by having an additional directly hired worker can be identified once I control for (a continuous function of) the number of workers that are used to determine the apprenticeship quota. The main contribution is to estimate the impact of compulsory learning contracts on firm productivity and other firm performance measures like output and profits, since the training provided through these contracts is different to voluntary training of regular workers by the firm, something the literature has addressed already.

Schiavon, Laura  (Brazilian, PhD candidate - Pontifícia Universidade Católica do Rio de Janeiro (PUC-Rio)) "The impact of financial incentives on legal capacity development"

The project aims at evaluating the impact of incentives and organizational changes on  judicial performance in Brazil. Specifically, it focuses on the role of wages in increasing judges and justice bureaucrats’ productivity. We explore the differential time and magnitude changes in salaries between Brazilian States affecting a wide range of intra-state courts. The wages variations between similar bureaucrats and courts which pertains to different States will be exploited in order to test whether salaries extrinsically motivates individual productivity. The study aims to make an important contribution to the state/legal capacity determinants and public service literature, especially for Latin American countries.

Tavares, Rafael (Brazilian, Ma in Economics - Universidade de São Paulo (USP)) "Mais Médicos Program: Does Political Party Matter?"

This research uses a regression discontinuity (RD) design in close electoral races to evaluate the Mais Médicos (‘More Physicians’) program, a project from the Brazilian Ministry of Health initiated in 2013. The program targets the provision of basic health services and the increase of the physicians per capita rate by transferring professionals to the Brazilian municipalities, which are constitutionally responsible for the basic health provision. By using a more credible source of exogenous variation in political alignment, the RD framework explores the fact that party affiliation of a municipality's mayor changes discontinuously at the threshold between the victory or loss of an affiliated candidate. Specifically, we estimate the effect of the political alignment of local governments on the following program outcomes: i) number of physicians received by municipalities (intensive margin), and ii) the municipalities' probability of participation (extensive margin).

Information about the first call >

What's new

Tenth Meeting

Impact Evaluation Network
Washington, D.C., USA

March 22nd to 24th, 2017


Training on Impact Evaluation of Public Policies - Intermediate Level

Instructor(s): María Laura Alzúa, Guillermo Cruces, Santiago Garganta and Marcelo Bérgolo
La Plata, Argentina
October 17-21, 2016

Basic Training on Impact Evaluation of Public Policies

Instructor(s): María Laura Alzúa, Leonardo Gasparini, Martín Cicowiez
La Plata, Argentina
April 18-22, 2016

Impact Evaluation at Masters in Economics at Universidad Nacional de La Plata

Instructor(s): Maria Laura Alzua
La Plata, Argentina
Since 2009

Connect with us