Files in this item



application/pdfdeAzevedo_Viviane.pdf (855kB)
(no description provided)PDF


Title:Essays on poverty reduction in Latin America
Author(s):de Azevedo, Viviane Martins Rodrigues
Director of Research:Villamil, Anne P.
Doctoral Committee Chair(s):Villamil, Anne P.
Doctoral Committee Member(s):Baer, Werner W.; Arends-Kuenning, Mary P.; Puzzello, Daniela
Department / Program:Economics
Degree Granting Institution:University of Illinois at Urbana-Champaign
Latin America
Conditional Cash Transfers
Abstract:In recent years, reducing poverty has been a core objective of social policy in Latin America. In 2008, the incidence of poverty reached almost a third of the population, of which 13 percent are classified as extremely poor. Most Latin American countries have invested vast resources in poverty reduction programs, particularly Conditional Cash Transfer schemes (CCTs), which have become a critical element in social policy. The effectiveness of CCTs has been a hotly debated topic. Most social programs in Latin America lack data for comprehensive evaluations to be carried out, but CCTs are an important exception. Many programs have been designed to have measurable impacts on various outcomes, generating a variety of empirical evidence that has been critical for designing and implementing new CCTs as well as fine-tuning existing programs. Evaluations cover a range of issues from access to education and health services, consumption patterns, to the impact on poverty, inequality, and labor market participation. Structural models of household behavior as well as empirical modeling of program implementation have been used to study the effectiveness of the conditionality and the size of the cash transfer. Nonetheless, the design of CCTs can be improved by better understanding households’ behavior in different environments as well as by constantly revising program rules such as transfer’s schemes and targeting instruments. CCTs have proved to be an effective form of social assistance, but are not a panacea to eradicate poverty. A careful examination of which factors generate disparities in living standards is crucial to design poverty reduction interventions. A common feature in many Latin America countries is the presence of geographic disparities in living standards that persist over time, in spite of growth at the national level. The role of infrastructure and geography vis-à-vis human capital and other determinants of household consumption growth have the potential to explain how aggregate economic growth translates into changes in household welfare. Moreover, researchers and policymakers alike are aware that the appropriateness of poverty reduction programs includes an in-depth analysis of geographical characteristics to aid the design of effective policy interventions to reduce poverty. For example, a better understanding of the geographical distribution of poverty can help fine-tune existing CCTs. The essays presented contribute to the literature on analytical work on poverty reduction programs, the effects of these programs on human capital accumulation and propagation of poverty, and the geographic determinants of household consumption. The first two papers analyze Mexico’s Conditional Cash Transfer Oportunidades. Mexico’s CCT is one of the oldest in Latin America and currently reaches 5 million households, one fourth of the country’s population. In terms of budget, the program costs about 0.5 percent of the country’s gross domestic product (GPD) and the transfers represent around a fifth of the mean household consumption. As in many other CCT, Oportunidades has two main objectives: reduce poverty in the short run and increase the human capital of children, which weakens the transmission of poverty across generations. To reach the second objective, transfers are conditional on household investments in children’s education, health, and nutrition. The first two of papers in this dissertation proposes analytical tools to fine tune the current design of Oportunidades along two dimensions. First, it analyzes the program current targeting system and proposes an alternative iii methodology to select beneficiaries. This is important for program effectiveness (measured as a reduction in targeting errors) and fairness, since the program’s objective is to reach the poorest households that under invest in the human capital of their children. Second, the paper assesses how a critical aspect of Oportunidades, school enrollment, changes in response to a new scheme of transfers and the corresponding costs in terms of the program’s budget. The first paper proposes a multidimensional targeting instrument that identifies CCT beneficiaries more efficiently by selecting beneficiaries with deprivations on the dimensions that are aligned with CCTs objectives. Targeting mechanisms used by CCTs have been generally successful in identifying the income poor, but have fared less well in identifying households that under-invest in children’s human capital. CCTs have applied targeting mechanisms based on proxy means test, generally variants of principal component analysis, probabilistic models or ordinary least square regression that identify monetary or well-being composite measures. Nonetheless, CCTs have traditionally identified their beneficiaries comparing the welfare metrics against a single threshold or cut-off point (or “unidimensionally”). The paper proposes a novel multidimensional approach to identify program’s beneficiaries that accounts for the multiple objectives of the CCTs and the multiple deprivations of the household. The proposed targeting mechanism is applied to the Mexican urban program Oportunidades and significantly improves the selection of households with children who are most deprived in the dimensions often relevant to CCTs (e.g. poor households with children that are not attending school). A robusteness exercise is carried out using an ex-ante evaluation technique that reinforces the contribution of the proposed targeting mechanism. The second paper uses an occupational choice model to ex-ante evaluate the impact of different transfers’ schemes on school enrollment and child labor. This paper assesses the occupational choice model of Bourguignon, Ferreira and Leite (2003) as a well suited technique to simulate the impact of policy changes in traditional CCT programs. This is important because one of the main setbacks of ex-ante evaluations is that the choice of the model can lead to ‘targeted modeling’ indicating that the results might be driven by the features of the model rather than the policy itself. Furthermore, we relax the model’s identifying assumption and simulate alternative transfer’s schemes with the objective of increasing school attendance. The findings indicate that the model predicts well school attendance and is a valuable technique to analyze ex-ante the impact of policy changes in school enrollment and its subsequent effects on poverty and inequality indicators. One of the counterfactual simulations indicate that eliminating or reducing school subsidies for primary education and increasing transfer for older students is a costeffective way to raise overall school enrollment. Another result indicates that increasing school attendance of 16-year-olds to 80 percent or more requires a substantial increase in the transfers. Another key stylized fact in the poverty literature is the existence of poverty traps. From the beginning of the development theory, poverty traps have called the attention of researchers and policymakers alike. As defined by Azariadis and Stachurski (2005) “A poverty trap is any reinforcing mechanism that causes poverty to persist.” At the macro level, the concept of a poverty trap has been used to explain differences observed in per capita income between countries and in the case studied in this paper within regions of a country. iv The third paper sheds light on the spatial distribution of poverty, and the intra-regional dynamics of income distribution within a country (Ecuador). The paper assesses the importance of geographical, physical, and human capital variables in explaining differences in consumption growth in Ecuador— a country characterized by significant geographical and cultural heterogeneity. In particular, the paper uses novel pseudo-panel data techniques to analyze the link between geographic factors and household welfare. The evidence indicates that geographical variables are important determinants of the household consumption growth after accounting for household characteristics such as education and labor market experience. Furthermore, the results indicate that individuals with similar characteristics experience increments in growth levels differently, depending on where they reside. Altogether, the results indicate that in other to reduce poverty, efficient and equitable government policies should take into account the role of geography and geographical diversity, including the regional distribution and provision of public goods.
Issue Date:2010-08-31
Rights Information:Copyright 2010 Viviane Martins Rodrigues de Azevedo
Date Available in IDEALS:2010-08-31
Date Deposited:2010-08

This item appears in the following Collection(s)

Item Statistics