none

Indiana University-Purdue University Indianapolis (IUPUI) / Observational data are frequently used for causal inference of treatment effects on prespecified outcomes. Several widely used causal inference methods have adopted the method of inverse propensity score weighting (IPW) to alleviate the in uence of confounding. However, the IPW-type methods, including the doubly robust methods, are prone to large variation in the estimation of causal e ects due to possible extreme weights. In this research, we developed an ordinary least-squares (OLS)-based causal inference method, which does not involve the inverse weighting of the individual propensity scores. We first considered the scenario of homogeneous treatment effect. We proposed a two-stage estimation procedure, which leads to a model-free estimator of average treatment effect (ATE). At the first stage, two summary scores, the propensity and mean scores, are estimated nonparametrically using regression splines. The targeted ATE is obtained as a plug-in estimator that has a closed form expression. Our simulation studies showed that this model-free estimator of ATE is consistent, asymptotically normal and has superior operational characteristics in comparison to the widely used IPW-type methods. We then extended our method to the scenario of heterogeneous treatment effects, by adding in an additional stage of modeling the covariate-specific treatment effect function nonparametrically while maintaining the model-free feature, and the simplicity of OLS-based estimation. The estimated covariate-specific function serves as an intermediate step in the estimation of ATE and thus can be utilized to study the treatment effect heterogeneity. We discussed ways of using advanced machine learning techniques in the proposed method to accommodate high dimensional covariates. We applied the proposed method to a case study evaluating the effect of early combination of biologic & non-biologic disease-modifying antirheumatic drugs (DMARDs) compared to step-up treatment plan in children with newly onset of juvenile idiopathic arthritis disease (JIA). The proposed method gives strong evidence of significant effect of early combination at 0:05 level. On average early aggressive use of biologic DMARDs leads to around 1:2 to 1:7 more reduction in clinical juvenile disease activity score at 6-month than the step-up plan for treating JIA.

Submitted by Rafael Burjack (rburjack@fgvmail.br) on 2012-10-11T20:27:44Z No. of bitstreams: 1 file_0_Tese.pdf: 550931 bytes, checksum: 38cf0263f5b9d60f39a3ea5ce971d077 (MD5) / Approved for entry into archive by Janete de Oliveira Feitosa (janete.feitosa@fgv.br) on 2012-10-24T18:12:32Z (GMT) No. of bitstreams: 1 file_0_Tese.pdf: 550931 bytes, checksum: 38cf0263f5b9d60f39a3ea5ce971d077 (MD5) / Made available in DSpace on 2012-10-30T10:37:14Z (GMT). No. of bitstreams: 1 file_0_Tese.pdf: 550931 bytes, checksum: 38cf0263f5b9d60f39a3ea5ce971d077 (MD5) Previous issue date: 2012-03-01 / In this paper, we analyze the impact of hosting the Summer Olympics on macroeconomic aggregates such as GDP, consumption, government consumption and investments per capita. The data is in panel structure and includes the period of ten years before and ten years after the event containing the Olympic Summer Games between 1960 and 1996. The sample countries comprise only candidates to host the games. This sampling strategy allows us to estimate the average treatment effect consistently, because it is assumed that these countries are comparable to each other, including those that ultimately hosted the games. The impact of hosting the Olympic games is measured by Fixed Effect and First Difference regressions. Moreover, we do a structural break test developed by Andrews (1993) to identify if hosting the Olympic Games creates anticipation effects for demand changes that stimulate current GDP, consumption, government consumption and investments. The results indicate a positive effect of the Summer Olympics in all variables of interest. However, the distribution in time and anticipation of these effects is unclear in the tests, changing significantly depending on the model and the significance level used. / Neste artigo e analisado o impacto de sediar as Olimp adas de Ver~ao em agregados macroecon^omicos como PIB, consumo, gastos do governo e investimentos per capitas. Os dados utilizados est~ao em painel e cont^em o per odo entre dez anos antes e dez anos depois do evento contendo os Jogos Ol mpicos de Ver~ao entre 1960 e 1996. Os pa ses da amostra s~ao apenas os candidatos a sediar os jogos, essa estrat egia amostral permite estimar os efeitos m edios de tratamento de forma consistente, pois assume-se que esses pa ses s~ao compar aveis entre si. Para avaliar o impacto dos jogos e lan cado m~ao de t ecnicas em painel como o Efeito Fixo e a Primeira Diferen ca e, al em disso, faz-se um teste de quebra estrutural desenvolvido por Andrews (1993) entre os pa ses sede. Os resultados indicam um efeito positivo e robusto dos Jogos Ol mpicos de Ver~ao em todas as vari aveis de interesse. No entanto, a distribui c~ao no tempo e antecipa c~ao desses efeitos e amb gua nos testes mudando de forma signi cativa dependendo do modelo e n vel de signi c^ancia utilizados.

No description available.

O objetivo deste trabalho é analisar como o nível de recursos escolares e a sua distribuição entre os estudantes afetam o desempenho acadêmico dos alunos brasileiros da rede pública de ensino. Para atingir tal objetivo, utilizou-se duas óticas de análise: uma análise descritiva e uma análise econométrica, ambas com base em dados do Censo Escolar e da Prova Brasil de 2013. Sob a ótica descritiva, parte-se da constatação de que as escolas públicas de Ensino Fundamental apresentam desigualdades nos patamares de recursos escolares conforme a região em que a escola se localiza. Em especial, as crianças de regiões mais empobrecidas tendem a enfrentar restrições de oferta de escolas públicas com maiores dotações de equipamentos pedagógicos, de instalações e de qualificação docente. A fim de avaliar apropriadamente a relação entre os recursos escolares e os resultados acadêmicos dos alunos no contexto brasileiro, tal distribuição não aleatória de recursos escolares entre os alunos foi considerada na análise econométrica. Nesta ótica de análise, por sua vez, foi estimado um modelo de correção em dois estágios que identificou os efeitos dos recursos escolares nas notas de Português e de Matemática dos alunos de 5º ano/4ª série do Ensino Fundamental das escolas públicas brasileiras, considerando a restrição de oferta de escolas com maior nível de recursos por meio do emprego de variável instrumental. Os resultados encontrados apontaram para a existência do efeito positivo que escolas mais bem equipadas em termos de recursos escolares têm sobre o desempenho escolar dos alunos, sendo esse efeito maior para as notas de Matemática e para a subpopulação de alunos oriundos de municípios com maior restrição de oferta de escola de alto nível de recursos. / The aim of this study is to analyze how the level of school resources and its distribution among the students affect the academic performance of Brazilian students from public schools. To achieve this goal, we used two analytical forms: a descriptive and an econometric analysis, both based in the Brazilian School Census and Prova Brasil data of 2013. Under the descriptive point of view, it is observed that public schools of Ensino Fundamental have inequalities in school resource levels depending on the region where the school is located. In particular, children from the poorest regions tend to face supply constraints of public schools with larger endowments of teaching equipment, facilities and teaching qualification. In order to properly evaluate the relationship between school resources and academic achievement of students in the Brazilian context, such non-random distribution of school resources among students was considered in the econometric analysis. In this analytical form, in its turn, a two-step correction model was estimated which identified the effects of school resources in the results of test scores in Portuguese and Math of 5th year/4th grade students of Ensino Fundamental education in Brazilian public schools, accounting for the supply restriction of schools with high level of resources using an instrumental variable. The findings pointed to the existence of the positive effect that the best equipped schools in terms of school resources have on school performance of students. This effect is particularly greater in Math scores and for the subpopulation of students from municipalities with the highest supply restriction of high quality schools.

An awareness of the effect of agricultural production on the environment has led to the development of policies to mitigate its adverse effects. This dissertation provides analyses of agri-environmental policies designed to protect environmental assets, as well as analytical decision-making tools useful for conducting policy evaluations. The first essay employs propensity score matching techniques to estimate the additionality of federal agricultural conservation programs for six conservation practices for farmers in Ohio. Additionality is an important measure of the effectiveness of conservation programs in inducing an increase in the conservation effort of farmers. Results suggest that additionality is positive and statistically significant for all six conservation practices. However, while programs achieve positive additionality for all practice types, a comparison between conservation practices reveals that certain practice types achieve higher percent additionality than others. Such results, coupled with information on the environmental benefits obtained per practice, could prove useful to program managers for improving the effectiveness of conservation programs. The second essay develops a new methodology to decompose the additionality measure into the two effects induced by conservation programs: expansion versus the new adoption of conservation practices. To do so, the relative contributions of two types of farmers, prior-adopters and new-adopters, are estimated. Results of the decomposition reveal that the additionality for prior-adopters is not significant for all practice types. Instead, additional conservation effort comes from new-adopters adopting new practices. Second, decomposition estimates suggest that practice types with a greater fraction of enrolled farmers that are new-adopters achieve greater percent additionality than those with greater proportions of prior-adopters. This suggests that a farmers? history in conservation adoption has a significant influence on additionality levels. The final essay analyzes the effect of recent instream flow diversion-guidelines on agricultural water security and streamflows within a decentralized water management regime. Spatially-explicit economic and hydrologic models are integrated to evaluate the tradeoffs between salmon bypass-flows and agricultural water security for three different diversion-guidelines within a northern-California watershed. Results indicate that the most restrictive diversion-guideline provides the greatest protection of bypass-flow days within smaller watersheds; however, within larger watersheds protection is not as significant. Water security, however, decreases sharply under the strict and moderate diversion-guidelines, especially during dry years. Overall, results indicate that greater focus should be given to protecting streamflows in the smallest watersheds, and meeting human water needs during dry years, when agricultural water security is impacted the most.

O objetivo deste trabalho é analisar como o nível de recursos escolares e a sua distribuição entre os estudantes afetam o desempenho acadêmico dos alunos brasileiros da rede pública de ensino. Para atingir tal objetivo, utilizou-se duas óticas de análise: uma análise descritiva e uma análise econométrica, ambas com base em dados do Censo Escolar e da Prova Brasil de 2013. Sob a ótica descritiva, parte-se da constatação de que as escolas públicas de Ensino Fundamental apresentam desigualdades nos patamares de recursos escolares conforme a região em que a escola se localiza. Em especial, as crianças de regiões mais empobrecidas tendem a enfrentar restrições de oferta de escolas públicas com maiores dotações de equipamentos pedagógicos, de instalações e de qualificação docente. A fim de avaliar apropriadamente a relação entre os recursos escolares e os resultados acadêmicos dos alunos no contexto brasileiro, tal distribuição não aleatória de recursos escolares entre os alunos foi considerada na análise econométrica. Nesta ótica de análise, por sua vez, foi estimado um modelo de correção em dois estágios que identificou os efeitos dos recursos escolares nas notas de Português e de Matemática dos alunos de 5º ano/4ª série do Ensino Fundamental das escolas públicas brasileiras, considerando a restrição de oferta de escolas com maior nível de recursos por meio do emprego de variável instrumental. Os resultados encontrados apontaram para a existência do efeito positivo que escolas mais bem equipadas em termos de recursos escolares têm sobre o desempenho escolar dos alunos, sendo esse efeito maior para as notas de Matemática e para a subpopulação de alunos oriundos de municípios com maior restrição de oferta de escola de alto nível de recursos. / The aim of this study is to analyze how the level of school resources and its distribution among the students affect the academic performance of Brazilian students from public schools. To achieve this goal, we used two analytical forms: a descriptive and an econometric analysis, both based in the Brazilian School Census and Prova Brasil data of 2013. Under the descriptive point of view, it is observed that public schools of Ensino Fundamental have inequalities in school resource levels depending on the region where the school is located. In particular, children from the poorest regions tend to face supply constraints of public schools with larger endowments of teaching equipment, facilities and teaching qualification. In order to properly evaluate the relationship between school resources and academic achievement of students in the Brazilian context, such non-random distribution of school resources among students was considered in the econometric analysis. In this analytical form, in its turn, a two-step correction model was estimated which identified the effects of school resources in the results of test scores in Portuguese and Math of 5th year/4th grade students of Ensino Fundamental education in Brazilian public schools, accounting for the supply restriction of schools with high level of resources using an instrumental variable. The findings pointed to the existence of the positive effect that the best equipped schools in terms of school resources have on school performance of students. This effect is particularly greater in Math scores and for the subpopulation of students from municipalities with the highest supply restriction of high quality schools.

O objetivo deste trabalho é analisar como o nível de recursos escolares e a sua distribuição entre os estudantes afetam o desempenho acadêmico dos alunos brasileiros da rede pública de ensino. Para atingir tal objetivo, utilizou-se duas óticas de análise: uma análise descritiva e uma análise econométrica, ambas com base em dados do Censo Escolar e da Prova Brasil de 2013. Sob a ótica descritiva, parte-se da constatação de que as escolas públicas de Ensino Fundamental apresentam desigualdades nos patamares de recursos escolares conforme a região em que a escola se localiza. Em especial, as crianças de regiões mais empobrecidas tendem a enfrentar restrições de oferta de escolas públicas com maiores dotações de equipamentos pedagógicos, de instalações e de qualificação docente. A fim de avaliar apropriadamente a relação entre os recursos escolares e os resultados acadêmicos dos alunos no contexto brasileiro, tal distribuição não aleatória de recursos escolares entre os alunos foi considerada na análise econométrica. Nesta ótica de análise, por sua vez, foi estimado um modelo de correção em dois estágios que identificou os efeitos dos recursos escolares nas notas de Português e de Matemática dos alunos de 5º ano/4ª série do Ensino Fundamental das escolas públicas brasileiras, considerando a restrição de oferta de escolas com maior nível de recursos por meio do emprego de variável instrumental. Os resultados encontrados apontaram para a existência do efeito positivo que escolas mais bem equipadas em termos de recursos escolares têm sobre o desempenho escolar dos alunos, sendo esse efeito maior para as notas de Matemática e para a subpopulação de alunos oriundos de municípios com maior restrição de oferta de escola de alto nível de recursos. / The aim of this study is to analyze how the level of school resources and its distribution among the students affect the academic performance of Brazilian students from public schools. To achieve this goal, we used two analytical forms: a descriptive and an econometric analysis, both based in the Brazilian School Census and Prova Brasil data of 2013. Under the descriptive point of view, it is observed that public schools of Ensino Fundamental have inequalities in school resource levels depending on the region where the school is located. In particular, children from the poorest regions tend to face supply constraints of public schools with larger endowments of teaching equipment, facilities and teaching qualification. In order to properly evaluate the relationship between school resources and academic achievement of students in the Brazilian context, such non-random distribution of school resources among students was considered in the econometric analysis. In this analytical form, in its turn, a two-step correction model was estimated which identified the effects of school resources in the results of test scores in Portuguese and Math of 5th year/4th grade students of Ensino Fundamental education in Brazilian public schools, accounting for the supply restriction of schools with high level of resources using an instrumental variable. The findings pointed to the existence of the positive effect that the best equipped schools in terms of school resources have on school performance of students. This effect is particularly greater in Math scores and for the subpopulation of students from municipalities with the highest supply restriction of high quality schools.

This thesis studies regression discontinuity designs with the use of additional covariates for estimation of the average treatment effect. We prove asymptotic normality of the covariate-adjusted estimator under sufficient regularity conditions. In the case of a high-dimensional setting with a large number of covariates depending on the number of observations, we discuss a Lasso-based selection approach as well as alternatives based on calculated correlation thresholds. We present simulation results on those alternative selection strategies.:1. Introduction 2. Preliminaries 3. Regression Discontinuity Designs 4. Setup and Notation 5. Computing the Bias 6. Asymptotic Behavior 7. Asymptotic Normality of the Estimator 8. Including Potentially Many Covariates 9. Simulations 10. Conclusion

Dans cette thèse, je me suis interessé à l’identification partielle des effets de traitements dans différents modèles de choix discrets avec traitements endogènes. Les modèles d’effets de traitement ont pour but de mesurer l’impact de certaines interventions sur certaines variables d’intérêt. Le type de traitement et la variable d’intérêt peuvent être défini de manière générale afin de pouvoir être appliqué à plusieurs différents contextes. Il y a plusieurs exemples de traitement en économie du travail, de la santé, de l’éducation, ou en organisation industrielle telle que les programmes de formation à l’emploi, les techniques médicales, l’investissement en recherche et développement, ou l’appartenance à un syndicat. La décision d’être traité ou pas n’est généralement pas aléatoire mais est basée sur des choix et des préférences individuelles. Dans un tel contexte, mesurer l’effet du traitement devient problématique car il faut tenir compte du biais de sélection. Plusieurs versions paramétriques de ces modèles ont été largement étudiées dans la littérature, cependant dans les modèles à variation discrète, la paramétrisation est une source importante d’identification. Dans un tel contexte, il est donc difficile de savoir si les résultats empiriques obtenus sont guidés par les données ou par la paramétrisation imposée au modèle. Etant donné, que les formes paramétriques proposées pour ces types de modèles n’ont généralement pas de fondement économique, je propose dans cette thèse de regarder la version nonparamétrique de ces modèles. Ceci permettra donc de proposer des politiques économiques plus robustes. La principale difficulté dans l’identification nonparamétrique de fonctions structurelles, est le fait que la structure suggérée ne permet pas d’identifier un unique processus générateur des données et ceci peut être du soit à la présence d’équilibres multiples ou soit à des contraintes sur les observables. Dans de telles situations, les méthodes d’identifications traditionnelles deviennent inapplicable d’où le récent développement de la littérature sur l’identification dans les modèles incomplets. Cette littérature porte une attention particuliere à l’identification de l’ensemble des fonctions structurelles d’intérêt qui sont compatibles avec la vraie distribution des données, cet ensemble est appelé : l’ensemble identifié. Par conséquent, dans le premier chapitre de la thèse, je caractérise l’ensemble identifié pour les effets de traitements dans le modèle triangulaire binaire. Dans le second chapitre, je considère le modèle de Roy discret. Je caractérise l’ensemble identifié pour les effets de traitements dans un modèle de choix de secteur lorsque la variable d’intérêt est discrète. Les hypothèses de sélection du secteur comprennent le choix de sélection simple, étendu et généralisé de Roy. Dans le dernier chapitre, je considère un modèle à variable dépendante binaire avec plusieurs dimensions d’hétérogéneité, tels que les jeux d’entrées ou de participation. je caractérise l’ensemble identifié pour les fonctions de profits des firmes dans un jeux avec deux firmes et à information complète. Dans tout les chapitres, l’ensemble identifié des fonctions d’intérêt sont écrites sous formes de bornes et assez simple pour être estimées à partir des méthodes d’inférence existantes. / In this thesis, I have been interested in the nonparametric (partial) identification of structural potential outcome functions and Average Treatment Effect (ATE) in various discrete models with endogenous selection and treatment. This topic of treatment effect concerns measuring the impact of an intervention on an outcome of interest. The type of treatments and outcomes may be broadly defined in order to be applied in many different contexts. There are many examples of treatment in economics (Labor, health, education, trade, industrial organization) such that Job training programs, surgical procedures, higher education level, research and development investment, being a member of a trade union etc. The decision to be treated or not, is usually not random but is based on individual choices or preferences. In such a context, determining the impact of the treatment becomes an important issue since we have to take into account the selectivity bias. The parametric version of such models has been widely studied in the literature, however in models with discrete variation, the parametrization is a strong source of identification. Then, we don’t know if the empirical results we obtain, are driven by the data or by the parametrization imposed on the model. I propose to look at a fully nonparametric version of those models, in order, to have more robust policy recommendations. The central challenge in this nonparametric structural identification is that the hypothesized structure fails to identify a single generating process for the data, either because of multiple equilibria or data observability constraints. In such cases, many traditional identification techniques become inapplicable and a framework for identification in incomplete models is developing, with an initial focus on identification of the set of structural functions of interest compatible with the true data distribution (hereafter identified set). Therefore, in the first chapter, I provide a full characterization of the identified set for the ATE in a binary triangular system. In the second chapter, I consider a model with sector specific unobserved heterogeneity. I provide the full characterization of the identified set for the structural potential outcome functions of an instrumental variables model of sectoral choice with discrete outcomes. Assumptions on selection include the simple, extended and generalized Roy models. In the last chapter, I consider a binary model with several unobserved heterogeneity dimensions, such as entry and participation games. I provide the full characterization of the identified set for the payoffs in 2 2 games with perfect information, including duopoly entry and coordination games. In all chapters, the identified set of the functions of interest are nonparametric intersection bounds and are simple enough to lend themselves to existing inference methods.