(with Hernan Winkler, Vincenzo Di Maro, Kelly Montoya and Sergio Olivieri )
World Bank Policy Research Working Paper No. 10794, 2024.
A growing body of literature investigates the labor market implications of scaling up “green” policies. Since most of this literature is focused on developed economies, little is known about the labor market consequences for developing countries. This paper contributes to filling this gap by providing new stylized facts on the prevalence of green occupations and sectors across countries at varying levels of economic development. Green occupations are defined using the Occupational Information Network, and green sectors are those with relatively lower greenhouse gas emissions per worker. The paper offers an initial assessment of how the implementation of green policies—aimed at expanding green sectors and strengthening the relative demand for green skills—may affect workers in developing economies. It finds that the share of green jobs is strongly correlated with the level of gross domestic product per capita across countries. When controlling for unobserved heterogeneity, a 1 percent increase in gross domestic product per capita is associated with 0.4 and 4.1 percentage point increases in the shares of new and emerging, and enhanced skills green jobs, respectively. The paper then focuses on Latin America and finds that only 9 percent of workers have a green job with respect to both occupation and sector. The findings show that within countries, workers with low levels of income and education are more likely to be employed in non-green sectors and occupations, and to lack the skills for a greener economy. This evidence suggests that complementary policies are needed to mitigate the potential role of green policies in widening income inequality between and within countries.
Link to the working paper
(with Deborah Winkler)
Revise and Resubmit on Review of World Economics. World Bank Policy Research Working Paper No. 10332, 2023.
This study provides new evidence on the local labor market impacts of trade, differentiating between the employment, income, migration, and informality channels. It uses a unique dataset matching information on exports and imports from customs with indicators on employment and labor incomes for around 2,000 Mexican municipalities over 2004–14. The analysis uses an instrumental variable approach that combines the initial structure of trade across municipalities with global trends in trade between low- and middle-income countries (excluding Mexico) and the United States by sector. First, the study finds that expanding exports per worker in Mexico’s municipalities increased labor force participation but not employment rates. Exports also raised total labor incomes but not average labor incomes, implying a growing labor supply. The results also find that export and import expansion increased immigration and lowered the rate of informal workers. Second, the analysis examines differences by geography and sectors. It finds that trade affected labor markets in the North through the income and migration channels and in the South through the employment and informality channels. Exports benefitted the total incomes of workers in both the manufacturing and service sectors but reduced informality only in manufacturing. Third, the study suggests a more favorable role of intermediate relative to final imports, driven by manufacturing imports. It also finds evidence for positive spillovers from global value chain participation through the employment and income channels. Finally, it examines how local policy mediates the labor market effects from trade, focusing on connectivity, labor market flexibility, and education spending.
Link to the working paper
(with Carlos Rodríguez-Castelán)
Revise and Resubmit on Review of Development Economics. World Bank Policy Research Working Paper No. 10129, 2022
Understanding the selection of workers into informality is a policy priority to design programs to increase formalization across Sub-Saharan Africa, where nine out of ten workers are informal. This paper estimates a model of self-selection with entry barriers into the formal sector to identify the extent of involuntary informality in Senegal, a representative country in terms of levels of informality in West Africa and with one of the most rigid labor markets in the world. Results show that the desire of being formal is greater for workers with formal education, married and with a lower proportion of children under the age of 5 living in the household. The individual's preference for the formal sector also grows with age at a decreasing rate. Results also show that labor informality is mainly a voluntary phenomenon with 30 percent of informal workers being involuntarily displaced into the informal sector. Results are robust to different model specifications, definitions of labor informality and heterogenous groups of workers.
Link to the working paper
CEDLAS Working Paper No. 308, 2022
This paper studies the effect of conditional cash transfer programs on the concentration of students of lower socioeconomic status in certain types of schools. Using data from an urban municipality in the state of Mexico, the paper assesses whether the Oportunidades program (formerly Progresa) had an impact on the type of secondary schools in which beneficiaries enrolled. To that end, it applies a regression discontinuity design, comparing the pretreatment characteristics of high schools attended by students just above and just below the program's eligibility threshold. The study finds that Oportunidades did not induce beneficiaries to attend secondary schools with more resources, with another orientation or location, but rather to concentrate on retentive schools, that is, those that tend to receive repeaters, that support students so that they do not drop out and where the approval of grades and subjects is higher. The consequences on learning of this segregation of beneficiaries and non-beneficiaries between retentive and non-retentive schools that is induced by the program are discussed and analyzed, since the null average effect of Oportunidades on learning that is found can be broken down into a positive effect (although not significant) for beneficiaries in non-retentive schools and a strongly negative effect of approximately 1 standard deviation among beneficiaries who attended retentive schools.
Link to the working paper
CEDLAS Working Paper No. 305, 2022
This paper provides a dynamic model of socioeconomic school segregation. It reinterprets and extends the Galor-Zeira model to illustrate the main causes and consequences of this phenomenon. The theoretical model predicts that the segregation of the poor will be greater the larger the quality gap between schools, the lower the interest rate, and the higher the marginal cost of attending high-quality schools for less wealthy individuals. These predictions are consistent with the data observed in the case of segregation between public and private schools. More importantly, the model highlights the self-reinforcing process that exists between socioeconomic school segregation and income distribution. Differences in the quality and costs of schools make the dynamic process converge to an extremely polarized society, self-perpetuating inequality.
Link to the working paper
CEDLAS Working Paper No. 299, 2022
This paper studies the effect of schooling on learning in Latin America according to the level of socioeconomic integration of the students. Using data from the different editions of PISA for 5 countries in the region (Argentina, Brazil, Chile, Costa Rica and Uruguay), the document exploits the exogenous variation in the date of birth of the students around the cut-off date for school entry using a fuzzy regression discontinuity design. The results indicate that, while the average effect of a year of schooling in Latin America is around 30 points on the PISA scale (0.30 standard deviations), it differs substantially according to the level of socioeconomic integration of students. Thus, while the learning gain derived from attending a higher school year for integrated students is around 40 PISA points, segregated students obtain an improvement in their learning in mathematics that is close to 20 points. The greatest learning from schooling that is gained by integrated students is acquired by those who are integrated with both richer and poorer peers. Moreover, this gain comes from those students who are furthest behind in their learning, who benefit most from schooling by being integrated.
Link to the working paper
(with Francisco Haimovich)
Work in progress
School dropout is one of the most serious challenges faced by the education system in Latin America, and particularly in Uruguay. To help identify students at-risk of dropping out, a quantitative profile of individual and household characteristics associated with dropout is estimated, and compared across countries and time. The results show that parental education can be considered the main factor associated to secondary dropout in Uruguay, seriously compromising upward social mobility. In addition, from a regional perspective, Uruguay has the largest dropout gender gap and the highest dropout penalty for large families in LAC. The importance of most of the characteristics analyzed to explain dropout has remained stable over the last 25 years. Estimations based on microsimulations indicates that dropout have a sizeable economic and distributive impact, reducing each year the GDP per capita by almost 6%, and explaining about 40 percent of the poverty rate.
Link to the working paper will be available soon.
(with Mariana Marchionni and Florencia Pinto)
Annals of the Argentine Association of Political Economy (AAEP), XLVIII Annual meeting , 2013
Inequality in PISA test scores of Argentinean students is among the highest of the participating countries. This paper explores the determinants of this inequality and quantifies the marginal contribution of characteristics of students and schools. We estimate Multilevel Models of the determinants of PISA 2009 reading results for Argentina and decompose the overall score inequality based on the methodology proposed by Lerman and Yitzhaki (1985) to decompose the Gini coefficient of total income by income source. The main result is that overall test scores inequality is mainly due to the high socioeconomic segregation among Argentinean schools.
Link to the working paper
(with Florencia Pinto and Mariana Marchionni)
Report for UNICEF Argentina, 2012.
In this paper, the determinants of inequality in the educational performance of Argentine students are explored using the results of the PISA 2009 tests. On the basis of multilevel econometric models, the study estimates to what extent different factors - individual characteristics of young people, of their families, peers, and schools - can explain the observed disparity in standardized test scores in reading, math, and science competencies. Regarding the effects at the school level, results indicate that it is the characteristics of the classmates –peer effects– and not the type of school itself that explain the differences in individual educational performance; in particular, the effects of repeating peers and the educational climate are the most relevant. The geographic location of schools also appears to contribute to the inequality in educational outcomes: students in schools in large cities score better, given everything else. Regarding the family context, the variables that approximate the socioeconomic level are those that are most strongly associated with an increase in the inequality of educational results - occupational qualification of the parents and availability of certain resources at home such as books and computers. Moreover, school repetition lag is associated with significantly lower individual performance on average, although the effect varies between schools.
Link to the working paper.