Conselho Nacional de Desenvolvimento Científico e Tecnológico / The concept of poverty has evolved from traditional analyzes, based on income, for analyzes that try to encompass the various dimensions of people's lives. Because of this, it has been made a several attempts to poverty measurement methods with several variables beyond monetary income. One of multidimensional methodology applied the IMP (Multidimensional Poverty Index) based on the Alkire-Foster method (2009). The main advantage of this method is the use of two cutting lines, and its decomposition in dimensions, indicators, regions, etc. This study aimed to apply this methodology to all municipalities from the state of Rio Grande do Sul between 2000 and 2010. The results indicate that between the years analyzed the multidimensional poverty, in general, has decreased in the state of Rio Grande do Sul. The most deprived are the indicators type of sewer sanitation facilities and years of study, there was a reduction in the percentage of deprivation between 2000 and 2010, but there is still a large portion of which is private. Except from literacy indicator, all indicators had declines in the percentage of deprivation. It was observed that there was a decrease of 0.59 percentage points in literate people from 9 years old between 2000 and 2010. The results indicated a decrease in the incidence of multidimensional poverty in Rio Grande do Sul from 24.16% to 14 71% of the sample. Likewise, there was a decrease in mean that multidimensionally poor indicators are deprived of 41.71% in 2000 to 37.67% of the indicators. The adjusted incidence in the state of Rio Grande do Sul fell from 10.71% in 2000 to 5.54% in 2010. It should be noted that the amplitude between the municipalities with the highest and lowest poverty incidence also had a fall in the years analyzed. Cities with lower poverty incidence in 2010 were Porto Alegre, Dois Irmãos, São Leopoldo, Esteio and Caxias do Sul. Municipalities with the highest poverty incidence in 2010 were Chuvisca, Barão do Triunfo, Jari, Barão Xavier and Capão Bonito Sul. / A concepção da pobreza tem evoluído de análises tradicionais, via renda, para análises que tentam abarcar as várias dimensões da vida das pessoas. Por conta disso, tem sido feito várias tentativas de métodos de medição de pobreza com diversas variáveis além da renda monetária. Uma das propostas metodológicas multidimensionais é o IPM (Índice de Pobreza Multidimensional) baseado no método Alkire-Foster (2009). A principal vantagem deste método é o uso de duas linhas de corte e sua decomposição em dimensões, indicadores, regiões e etc. O presente estudo teve o objetivo de aplicar essa metodologia para todos os municípios gaúchos nos anos 2000 e 2010. Os resultados encontrados indicam que entre os anos analisados a pobreza multidimensional, de maneira geral, tem diminuído no Rio Grande do Sul. As maiores privações estão nos indicadores tipo de escoadouro das instalações sanitárias e anos de estudo. Houve uma queda na porcentagem de privações entre 2000 e 2010, mas ainda há uma grande parcela da população que se encontra privada. Com exceção do indicador alfabetização, todos os indicadores obtiveram quedas na porcentagem de privações. Observou-se que houve uma queda de 0,59 pontos percentuais nas pessoas alfabetizadas a partir dos 9 anos de idade, entre 2000 e 2010.Os resultados indicaram uma queda da incidência de pobreza multidimensional no Rio Grande do Sul de 24,16% para 14,71% da amostra. Da mesma forma, houve uma queda na média de indicadores que os pobres multidimensionalmente encontram-se privados de 41,71%, em 2000, para 37,67% dos indicadores. A incidência ajustada no Rio Grande do Sul caiu de 10,71%, em 2000, para 5,54% em 2010. Ressalta-se que a amplitude entre os municípios com maiores e menores incidências de pobreza também teve uma queda entre os anos analisados. As cidades com menores incidências de pobreza, em 2010, foram Porto Alegre, Dois Irmãos, São Leopoldo, Esteio e Caxias do Sul. Já os municípios com maiores incidências de pobreza, em 2010, foram Chuvisca, Barão do Triunfo, Jari, Gramado Xavier e Capão Bonito do Sul.

The Bolsa Família Program unified income transfer programs and facilitated control, oversight and provided the possibility of integration with other programs as stated Sanson (2008). Weissheimer (2006) stresses that the improvements were not only to rationalize and expand existing programs, not limited to the greater transfer of income, but also to the repercussions of the school attendance requirement of the children of beneficiary families, compliance with the vaccination schedule and follow-up of gestation and prohibition of Child labor. Based on the understanding that a policy to combat poverty must act in areas other than income transfer, Amartya Sen's theory allows a better basis for analyzing these effects and needs. The capabilities approach developed by Sen (1999) considers that the expansion of personal skills should be the goal of the development process and not the expansion of wealth or income, wealth should be seen as a means and not as an end of the process. In this dissertation, the policy analyzed was the Bolsa Família Program, using the theoretical basis of the training approach, 18 variables were distributed among four dimensions: housing, work, education and health. At the end of the analysis of all dimensions, the best gross results were found, for the majority of the variables, in the southeast region, mainly São Paulo State, indicating that this region has more infrastructure compared to the other regions of the country; Among the worst results, the northern region had the worst gross results, especially in variables related to access to durable goods and local infrastructure (sewerage, general grid, energy). The comparative result that indicates the local deprivation of the beneficiaries in the north and northeast region was positive, since it presented smaller differences between beneficiaries and non beneficiaries. Another important result was that variables such as: bathroom, refrigerator, no portfolio, no education, Micropc, sewage and garbage, control of observable characteristics was not enough to explain the difference between groups, with positive results concentrated in non-beneficiaries; Already for the variables: busy, job search, working hours and general network had the difference explained in great part by the characteristics, with approximation of the results between groups after the pairing. For the variables that have conditionality directly linked to the PBF, the results were expressively higher among the beneficiaries, indicating that this counterpart method is efficient; The results indicated that the beneficiaries of the program have deficiencies in all dimensions, with difficulties of access to basic housing, work, education and health conditions that in some part could be explained by the observable characteristics, demonstrating that this deficiency is related to the geographic region And income, color, age and sex, in other variables, beneficiaries have less access even compared to individuals with similar characteristics. As the PBF is applied without distinction in all regions of the country, it has bottlenecks that can be improved by observing local needs and infrastructure and access difficulties in each region, allowing a broadening of benefits beyond conditionalities. / O Programa Bolsa Família unificou programas de transferência de renda e facilitou o controle, fiscalização e proporcionou a possibilidade de integração com outros programas como afirma Sanson (2008). Weissheimer (2006) destaca que as melhorias não foram apenas racionalizar e expandir os programas existentes, não se restringindo apenas à maior transferência de renda, mas às repercussões da exigência de frequência escolar das crianças de famílias beneficiadas, cumprimento do cronograma de vacinação e acompanhamento de gestação e proibição do trabalho infantil. A partir do entendimento de que uma política de combate à pobreza deve agir em outras áreas que não apenas a transferência de renda, a teoria de Amartya Sen permite maior embasamento para análise desses efeitos e necessidades. A abordagem das capacitações desenvolvida por Sen (1999) entende que a ampliação das capacitações pessoais deve ser o objetivo do processo de desenvolvimento e não a ampliação da riqueza ou renda, a riqueza deve ser vista como um meio e não como um fim do processo. Nesta dissertação, a política analisada foi o Programa Bolsa Família, utilizando a base teórica da abordagem das capacitações, foram distribuídas 18 variáveis entre quatro dimensões: condição de moradia, trabalho, educação e saúde. Ao fim da análise de todas as dimensões, os melhores resultados brutos foram encontrados, para a maioria das variáveis, na região sudeste, principalmente estado de São Paulo indicando que esta região possui maior infraestrutura comparada as demais regiões do país; entre os piores resultados, a região norte possuiu os piores resultados brutos, principalmente em variáveis relacionadas a acesso a bem duráveis e infraestrutura local (esgoto, rede geral, energia). O resultado comparativo que indica a privação local dos beneficiários a região norte e nordeste apresentaram destaque positivo, pois apresentou menores diferenças entre beneficiários e não beneficiários. Outro resultado importante foi que variáveis como: banheiro, geladeira, sem carteira, sem instrução, Micropc, esgoto e lixo, o controle das características observáveis não foi suficiente para explicar a diferença entre os grupos, permanecendo resultados positivos concentrados em não beneficiários; já para as variáveis: ocupado, procura emprego, horas de trabalho e rede geral tiveram a diferença explicada em grande parte pelas características, com aproximação dos resultados entre grupos após o pareamento. Para as variáveis que possuem condicionalidade diretamente ligada ao PBF, os resultados foram expressivamente superiores entre os beneficiários, indicando que este método de contrapartida é eficiente; os resultados indicaram que os beneficiários do programa possuem deficiências em todas as dimensões, com dificuldades de acesso a condições básicas de moradia, trabalho, educação e saúde que em alguma parcela puderam ser explicadas pelas características observáveis, demonstrando que essa deficiência está relacionada à região geográfica e a renda, cor, idade e sexo, em outras variáveis beneficiários possuem menos acesso mesmo comparados a indivíduos com características semelhantes. Como o PBF é aplicado sem distinção em todas as regiões do país, possui gargalos que podem ser melhorados observando carências locais e dificuldades de infraestrutura e acesso de cada região, permitindo uma ampliação dos benefícios para além das condicionalidades.

Fundação de Amparo a Pesquisa do Estado da Bahia / This dissertation examines poverty in the state of Bahia under the multidimensional approach to the study of poverty, emphasizing the need to enlarge the focus of research beyond the lack of income and to consider the multidimensionality of poverty in developing policies that aim to combat it. Initially, discuss, the evolutionary path of the study of poverty, emphasizing the dominance of unidimensional monetary approach in the work on poverty in Brazil and Bahia. Subsequently, performs an empirical analysis of poverty in Bahia in 1995, 2001 and 2006, based on two approaches to study of multidimensional poverty: the basic needs approach and the capabilities approach. From the concepts employed by these approaches, defines poverty as the dissatisfaction of basic human needs which deprives the individual to develop and expand its capabilities. The methodology used in the empirical analysis is to apply the multivariate method of the Multiple Correspondence Analysis in the calculation of a multidimensional indicator of poverty (IMP) per unit of population studied. This indicator is composed of a set of qualitative variables, called the primary indicators of deprivation. The results of applying this method also allow for a multidimensional poverty line, used as a criterion for identification of poverty. Based on this poverty line and the indicator, was possible to calculate the traditional indices of poverty FGT (0), FGT (1) and FGT (2) and then measure the poverty in Bahia. The results show that the proportion of poor people was high, especially among the rural population, but this proportion decreased between 1995 and 2006, except in the metropolitan area. Among the individuals considered poor, poverty was more intense and more severe in rural areas. The decomposition of poverty by race and color or by type of family indicated that individuals in black or brown, and the families of the "mother with children under 14 years", had higher risk of poverty. The breakdown by sex showed no significant differences between men and women regarding the extent and risk of poverty. Regarding census areas, the less populous municipalities were those that had a higher percentage of poor. When compared with the income poverty, the multidimensional poverty is had higher (years 2001 and 2006) and stable. While this first there was a significant fall in the period 1995/2006, the decline occurred in the second was much less significant. Thus, by focusing the analysis only on the inadequacy of income, it is not possible to investigate other types of deprivation that affect the people of Bahia. Although in the period examined to identify improvements in the primary indicators, living conditions these people were still short of what could be considered a adequate situation, revealing that a significant population of Bahia had not their basic needs met. Moreover, the set of deprivation cannot be disregarded in the planning and implementation of policies to combat poverty, because, otherwise, it undermines the effectiveness of such policies. / Esta dissertação analisa a pobreza no estado da Bahia, sob o enfoque multidimensional de estudo da pobreza, enfatizando a necessidade de se ampliar o foco de investigação para além da insuficiência de renda e de se considerar a multidimensionalidade da pobreza na elaboração das políticas que objetivem o seu combate. Para isso, discute, inicialmente, a trajetória evolutiva do estudo da pobreza, dando destaque ao predomínio da abordagem unidimensional monetária nos trabalhos sobre a pobreza no Brasil e na Bahia. Posteriormente, realiza uma análise empírica da pobreza na Bahia nos anos 1995, 2001 e 2006, fundamentada em duas abordagens multidimensionais de estudo da pobreza: a abordagem das necessidades básicas e a abordagem das capacitações. A partir dos conceitos empregados por estas abordagens, define a pobreza como a insatisfação das necessidades humanas básicas que priva o indivíduo de desenvolver e expandir as suas capacitações. A metodologia empregada na análise empírica consiste em aplicar o método multivariado da Análise de Correspondências Múltiplas no cálculo de um indicador multidimensional de pobreza (IMP) por unidade da população analisada. Este indicador é composto por um conjunto de variáveis qualitativas, denominadas de indicadores primários de privação. Os resultados da aplicação desse método também permitem estabelecer uma linha de pobreza multidimensional, utilizada como critério de identificação da pobreza. Com base nesta linha de pobreza e no indicador, foi possível calcular os tradicionais índices de pobreza FGT(0), FGT(1) e FGT(2) e, então, mensurar a pobreza na Bahia. Os resultados demonstram que a proporção de pobres era elevada, principalmente entre a população rural, mas que essa proporção diminuiu entre os anos 1995 e 2006, exceto na área metropolitana. Entre os indivíduos considerados pobres, a pobreza era mais intensa e mais severa na área rural. A decomposição da pobreza por cor ou raça e por tipo de família apontou que os indivíduos de cor preta ou parda, bem como as famílias do tipo mãe com filhos menores de 14 anos , apresentavam maior risco de pobreza. A decomposição por sexo não mostrou diferenças significativas entre homens e mulheres quanto à extensão e ao risco de pobreza. Com relação às áreas censitárias, os municípios menos populosos eram os que possuíam maior percentual de pobres. Quando comparada com a pobreza por renda, a pobreza multidimensional se apresentava mais elevada (anos 2001 e 2006) e estável. Enquanto nesta primeira houve uma queda expressiva no período 1995/2006, a diminuição ocorrida na segunda foi bem menos significativa. Assim, ao se concentrar as análises apenas na insuficiência de renda, não se torna possível investigar os outros tipos de privações que atingem a população baiana. Mesmo que no período analisado se identifiquem melhorias nos indicadores primários, as condições de vida desta população ainda estavam longe do que poderia ser considerada uma situação adequada, revelando que uma parcela significativa da população baiana não tinha suas necessidades básicas satisfeitas. Ademais, o conjunto de privações não pode ser desconsiderado no planejamento e execução das políticas de combate à pobreza, pois, do contrário, se compromete a eficácia de tais políticas. / Mestre em Economia

Submitted by Marilene Donadel (marilene.donadel@unioeste.br) on 2018-09-26T00:10:14Z No. of bitstreams: 1 Nadja_Oliveira_2018.pdf: 1690892 bytes, checksum: 2b0670270aa7334aef0ec35c3ba8e630 (MD5) / Made available in DSpace on 2018-09-26T00:10:14Z (GMT). No. of bitstreams: 1 Nadja_Oliveira_2018.pdf: 1690892 bytes, checksum: 2b0670270aa7334aef0ec35c3ba8e630 (MD5) Previous issue date: 2018-06-18 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / The objective of this research is to analyze the poverty feminization process in the Northeast´s Brazilian region, from 2004 to 2015, through a multidimensional approach. For this purpose, was estimated the Multidimensional Poverty Index (MPI) by using the microdata of the National Household Sample Survey (PNAD) for the years 2004 to 2015. the poverty incidence results (H) showed that in Brazil the proportion of poor female heads of household was lower than the male proportion. In 2004, the percentage of poor women who were heads of their families was 55.03%; by 2015, the percentage of poor women fell to 39.31%. The intensity of poverty or average poverty (A) in 2004 revealed that these women presented, on average, 20.07% deprivations of the total indicators considered, which corresponded to 7 of the 37 indicators. In 2015, this proportion represented 18.53%, which means that these women presented on average deprivations in approximately 7 of the 37 indicators. The results of the incidence of adjusted poverty (M0), revealed that in 2004 about 11.05% of Brazilian women were multidimensionally poor, in 2015 the percentage of women in this condition fell to 7.28%. In the Northeast region, unlike that observed for female Brazilian heads of household, it was possible to verify that women in this region are multidimensionally poorer than men, and to confirm the poverty feminization process occurrence in this Region, since poverty in this Region declined more among men than among women. It was also observed that for the female heads of families in the Northeast Region, the dimensions of access to work, availability of resources and consumption of goods were those in which women were more vulnerable in all the years of the analyzed period. In 2015, approximately 53.34%, 48.04% and 34.52% of these women were deprived in these dimensions, respectively. In comparison to male heads of household in the Northeast Region, women were more deprived in terms of family vulnerability, availability of resources until the year 2011, and from the year 2009 on the use of time. There was also a noticeable decrease in multidimensional poverty for female heads of households in the Northeast, whose MPI for the eleven year period has been reduced by 29.65% while the same indicator for female heads of household in Brazil was reduced and 34,12%. The dimensions related to work and income were the ones that contributed the most to the insertion of northeastern women into a condition of poverty and are thus subjects that require more attention from the public power, since the verified distance in the percentage of deprivations observed in these dimensions in relation to Brazil was significant. / A presente pesquisa teve por objetivo analisar o processo de feminização da pobreza na Região Nordeste do Brasil, por meio de uma abordagem multidimensional. Para tal, estimou-se o Índice de Pobreza Multidimensional (MPI) tendo por base os microdados da Pesquisa Nacional por Amostra de Domicílios (PNAD) relativos ao período de 2004 até 2015. Os resultados da incidência da pobreza (H) evidenciaram que no Brasil a proporção de mulheres chefes de família pobres foi inferior à proporção masculina. Em 2004, o percentual de mulheres pobres que eram chefes de suas famílias era de 55,03%, já no ano de 2015 esse mesmo percentual caiu para 39,31%. A intensidade da pobreza ou pobreza média (A) no ano de 2004 revelou que essas mulheres eram em média privadas em cerca de 20,07% do total de indicadores considerados, o que correspondeu a 7 dos 37 indicadores. Em 2015, essa proporção representou 18,53% dessas mulheres, isto é, elas apresentaram em média privações em aproximadamente 7 dos 37 indicadores. Os resultados da incidência da pobreza ajustada (MPI) apontaram que em 2004 cerca de 11,05% das mulheres brasileiras eram multidimensionalmente pobres, no ano de 2015 o percentual de mulheres nessa condição caiu para 7,28%. Na região Nordeste, diferentemente do observado para as chefes de família do Brasil, foi possível constatar que as mulheres são multidimensionalmente mais pobres do que os homens, o que caracteriza a ocorrência do processo de feminização da pobreza. Também se observou que para as chefes de família da Região Nordeste, as dimensões do acesso ao trabalho, disponibilidade de recursos e consumo de bens se mostraram mais vulneráveis em todos os anos do período analisado. No ano de 2015, aproximadamente 53,34%, 48,04% e 34,52% dessas mulheres eram privadas nessas dimensões, respectivamente. Em comparação aos homens chefes de família da Região Nordeste, as mulheres foram mais privadas nas dimensões referentes à vulnerabilidade familiar e na disponibilidade de recursos até o ano de 2011 e, a partir do ano de 2009, no uso do tempo. Também se observou uma sensível diminuição na pobreza multidimensional para as mulheres chefes de família do Nordeste, cujo MPI para o período de onze anos, fora reduzido em 29,65% enquanto o mesmo indicador para as mulheres chefes de família do Brasil foi reduzido em 34,12%. No entanto, as dimensões relativas ao trabalho e à renda foram as que mais contribuíram para a inserção da mulher nordestina em uma condição de pobreza, e são assim, temas que exigem mais atenção do poder público, pois a distância verificada no percentual de privações observadas nessas dimensões em relação ao Brasil se mostrou significativa.

Gleichzeitig mit der weit verbreiteten Verwendung von traditionellen einkommens- oder konsumbasierten Maßnahmen zur Messung menschlicher Armut und Entwicklung gibt es seit den späten 70er Jahren zunehmendes Interesse an der Ökonomie des Glücks. Ebenso gibt es einen breiten Literaturbereich, der Indizes definiert und diskutiert hat, die „functionings" auf der Grundlage des „Capabilities Approach“ von Sen (1985) auf sich vereinen, welche auf einer Vielzahl von ideologischen Urteilen und Zielen zur Bestimmung des objektiven Wohlbefindens (1984), S. 187) beruhen. Beide Ansätze sind ähnlich in ihrer Prämisse, dass Einkommen oft eine unzureichende Determinante des Wohlbefindens ist – ein latenter Begriff, dass besser mit anderen, breiteren Definitionen- subjektiv oder objektiviert- erfasst wird. Beide Konzepte des Wohlbefindens wurden im Hinblick auf ihr Verhältnis zum Einkommen untersucht, und es wurde ein klarer Unterschied zwischen Einkommen und diesen beiden Maßnahmen festgestellt. Angesichts der relativen Neuheit und Komplexität beider Ansätze sind sie jedoch bislang selten in wissenschaftlichen Arbeiten zusammengebracht worden. Den dritten Aufsatz in dieser Arbeit ist ein Versuch, diese beiden Ansätze zu kombinieren und damit diese Lücke in der Literatur zu erfüllen. Subjektives Wohlbefinden wird mit verfügbaren Daten, die Zufriedenheit messen, festelegt, während das objektive Wohlbefinden durch einen Multidimensional Poverty Index (MPI) operationalisiert wird (Alkire & Santos, 2010). Der MPI ist einer der neuesten Versuche zur Messung des menschlichen Wohlbefindens im Rahmen des Capabilities Approach. Die Entwicklungspolitik erkennt ebenfalls graduell diese Maße als eine genauere Beschreibung des Wohlbefindens an, oder betrachtet diese zumindest als eine sinnvolle Ergänzung zu metrisch-monetären Maßen. Angesichts der vielen nationalen und internationalen Programme, die eine umfassende Verbesserung des menschlichen Wohlbefindens zum Ziel haben, gibt es erstaunlich wenige Arbeiten, die dazu beitragen können, die Auswirkungen eines bestimmten Programms auf das allgemeine Wohlbefinden und nicht nur auf eine bestimmte Dimension zu quantifizieren und zu bewerten. Der zweite Aufsatz in dieser Arbeit beschäftigt sich kritisch mit diesem Ansatz und betrachtet dabei den besonderen Fall Südafrikas. Aufgrund der steigenden Beliebtheit der mehrdimensionalen Armutsmaße besteht ein zunehmender Bedarf an einer Überprüfung ihrer grundlegenden Eigenschaft, ebendies zu erreichen. Eine wachsende Zahl von Forschern hat sich mit den Problemen beschäftigt, die ein zusammengesetztes Maß wie der MPI mit sich bringen kann und dessen Fähigkeit zur Messung multidimensionalen Wohlbefindens beeinträchtigen kann. Ein bestimmter Aspekt ist hierbei die Gewichtung der einzelnen Dimensionen und Indikatoren, um Armut über verschiedene Regionen hinweg zu definieren. Diese Dissertation schafft in Aufsatz eins eine Brücke zwischen den statistischen Methoden und den optimalen Gewichtungsschemata, die speziell zur Messung des multidimensionalen Wohlbefindens in verschiedenen Ländern genutzt werden können.

Une réforme efficace et efficiente du secteur public constitue un enjeu important pour les gouvernements et les partenaires au développement partout dans le monde. La décentralisation budgétaire ̶ le transfert de compétences en matière de recettes et de dépenses du gouvernement central vers les gouvernements locaux ̶ est souvent considérée comme un moyen d’améliorer l’efficacité de la mobilisation des ressources domestiques et de renforcer l’accès et la qualité des services publics. Dans la poursuite de ces objectifs, la décentralisation budgétaire est devenue un élément essentiel de la politique de réforme du secteur public dans de nombreux pays surtout ceux en développement. Face à l’importance des enjeux liés à cette réforme notamment sur les finances publiques, une importante littérature économique s’est attachée à étudier ses conséquences et les conditions de sa réussite. Cette thèse contribue à cet effort à travers quatre chapitres portant sur les effets de la décentralisation budgétaire dans les pays en développement et plus précisément en Côte d’Ivoire.Le premier chapitre analyse les conséquences de la mobilisation fiscale au niveau municipal sur l’accès aux services publics et sur la réduction de la pauvreté. Les résultats suggèrent que l’augmentation des revenus des municipalités accroît l’accès aux services publics, notamment l’éducation, et, dans une moindre mesure, réduit la pauvreté. Cet effet est accentué dans les régions à faible diversité ethnolinguistique et en milieu urbain. Le second chapitre évalue les effets de l’autonomie fiscale des municipalités sur la distribution de revenus de la population. Empiriquement, les résultats suggèrent qu’une augmentation des revenus prélevés par le gouvernement central (et rétrocédés aux municipalités) réduit les inégalités de revenus, alors qu’une augmentation des revenus directement prélevés par les municipalités augmente les inégalités (taxes locales). Le chapitre trois (3) analyse dans quelle mesure les transferts du gouvernement central aux municipalités affectent la mobilisation des revenus de ces dernières. Les résultats montrent qu’une augmentation des transferts augmente les revenus des municipalités provenant des impôts rétrocédés, et, dans une moindre mesure, les taxes locales. Le quatrième chapitre porte sur les déterminants de la conformité des taxes foncières. Nos résultats concluent que la conformité des taxes foncières est déterminée, entre autres, par les niveaux de services publics du gouvernement central et des municipalités, de la confiance des citoyens envers les municipalités et de la compréhension des citoyens du système de taxation. Les résultats de cette thèse induisent d’importantes implications en termes de politiques de renforcement du secteur public. / An effective and efficient public-sector reform enhances government capability to raise domestic revenue for prioritized spending. It also contributes to strengthening the government capacity to manage public resources more effectively and deliver public services. In pursuing these objectives, fiscal decentralization, the devolution of taxing and spending powers to lower levels of government, has become a key public-sector reform in many countries. Given such high stakes, a growing economic literature is focusing on how to better understand the challenges developing countries face in implementing fiscal decentralization reforms. The present thesis adds to this effort through four contributions shedding light on specific issues related to fiscal decentralization.Chapter 1 analyses whether and how municipal revenue mobilization could enhance citizens’ access to public services and reduce poverty. The results indicate that increasing municipal-raised revenues improve access to public services and reduce poverty. This effect works mainly through enhancing access to education than on access to health, especially in less ethnically diverse localities and in urban zones. Chapter 2 investigates the effect of municipal revenue autonomy on inequalities within sub-national governments. We conclude that higher local revenue autonomy reduces income inequalities within localities, which effect differs between the type of revenue and jurisdiction considered. Chapter 3 analyzes the impact of the transfers from central government to municipalities on the revenue mobilization by municipalities. Our results show that increasing central transfers to subnational government boost revenue mobilization by municipalities for both tax revenue and non-tax revenue. The effect of transfers is found to be higher for tax revenue than for non-tax revenue. Finally, we examine whether the effect of government responsiveness on property tax compliance differs between central government and municipalities in chapter 4. The results conclude that responsiveness by municipalities is found to have a greater effect on property tax compliance than those by central governments. The results of our analysis bear important policy implications to strengthen public sector, especially in sub-Saharan Africa.

L'objet de cette thèse est de proposer et de développer les différentes mesures multidimensionnelles de la pauvreté. La multidimensionnalité de la pauvreté fait aujourd'hui consensus. Scientifiques, décideurs politiques et professionnels du développement s'accordent pour dire que la seule dimension monétaire (le manque de revenu) ne suffit pas à représenter la pauvreté. En se basant sur les travaux de Sen (Prix Nobel d'Economie) en particulier sur l'approche des capacités, nous proposons quatre mesures différentes de la pauvreté. La première est une mesure monétaire basée sur l'approche utilitaire ; la seconde est une mesure subjective basée sur les expériences des ménages ; la troisième est une mesure multidimensionnelle axiomatique et enfin la dernière est une mesure non axiomatique basée sur la théorie des ensembles flous. Elles sont toutes mises en oeuvre en utilisant les données d'enquêtes EDAM3-IS (Enquête Djiboutienne Auprès des Ménages 2012). Les résultats s'inscrivent dans un contexte de croissance économique que connait Djibouti. Cependant, toutes les mesures utilisées ont montré des grandes disparités régionales entre la capitale et les régions en termes d'infrastructure de base et de bien-être des ménages. Chacune de ses méthodes a fourni des résultats avec différentes interprétations des déterminants de la pauvreté. Cela ne signifie pas qu'il existe une méthode bien meilleure que l'autre, mais chaque approche, dans un contexte particulier, peut-être plus pertinente. Ainsi, l'identification des pauvres en appliquant les différentes mesures de la pauvreté nous a donné un profil différencié. Par conséquent, le décideur doit définir au préalable l'objectif poursuivi dans les politiques de lutte contre la pauvreté. Nous pouvons souligner que l'intégration d'une pondération subjective dans la mesure de la pauvreté est une de nos contributions au développement de mesures multidimensionnelles de la pauvreté. / The aim of this thesis is to propose and develop the various multidimensional measures of poverty. There is a consensus on the multidimensional nature of poverty. Scientists, policy makers and development professionals agree that the monetary dimension (lack of income) is inadequate to represent poverty. On the basis of the work of Sen (Nobel Proze of Economics), particularly on the capability approach, we propose four different measures of poverty. The first one is a monetary measure based on the utilitarian approach ; the second is a subjective measure founded on household experience ; the third is a multidimensional axiomatic measure and the final one is a non-axiomatic measure based on the theory of fuzzy sets. They are implemented using survey data EDAM3-IS (Djiboutian Survey of Households 2012). The esults fall within the framework of economic growth in Djibouti. However, all the measures used have shown great disparities between the capital and the regions in terms of basic infrastructure and household welfare. Each method produced results with different interpretations of the determinants of poverty. This does not mean that there is one method being better than the other but rather each approach, in a particular context, may be more relevant. Thus, identifying the poor by applying the different measures of poverty gave us a clear-cut profile, which implies that the decision-maker must first set the aim in view in the implementation of anti-poverty policies. It can be emphasized that the inclusion of a subjective weighting in the process of measuring of poverty is one of our contributions towards the development of multidimensional measures of poverty.

La cobertura en salud hacia toda la población peruana es el principal objetivo del sistema sanitario nacional. Su mal manejo en las inversiones destinadas para alcanzar este objetivo, demuestran poca efectividad. Debido a esto, la población multidimensionalmente pobre queda vulnerable ante cualquier eventualidad que afecte su salud y se ve más propensa a incurrir en gastos de bolsillo que finalmente terminarán siendo catastróficos, para acceder a los servicios médicos correspondientes. En este contexto, estos gastos son preocupantes, ya que a largo plazo empeoran el bienestar del hogar. Para hacer un análisis más exhaustivo, primero se elaborará el Índice de Pobreza Multidimensional. Seguido, por un modelo de variable binaria que ayudará a estimar la regresión final y definirá la relación que existe entre pobreza multidimensional y gasto catastrófico en salud. Otras variables usadas que explican este gasto son: género, ocupación, afiliación a algún seguro, estado civil, nivel educativo y presencia de miembros vulnerables a enfermarse. Los resultados obtenidos demuestran que este gasto catastrófico sí es más probable a ocurrir en aquellos hogares multidimensionalmente pobres respecto a los multidimensionalmente no pobres. Esto, resalta la importancia del Estado por elaborar políticas y/o gestionar programas que protejan la salud de la población sobre todo la más vulnerable, con la finalidad de mejorar el crecimiento y desarrollo del país. / Health coverage for the entire Peruvian population is the main objective of the national health system. Their mismanagement of investments destined to achieve this objective, show little effectiveness. Due to this, the multidimensionally poor population is vulnerable to any eventuality that affects their health and is more prone to incur out-of-pocket expenses that will ultimately end up being catastrophic, to access the corresponding medical services. In this context, these expenses are worrying, since in the long term they worsen the well-being of the household. For a more comprehensive analysis, the Multidimensional Poverty Index will first be developed. Followed by a binary variable model that will help estimate the final regression and define the relationship between multidimensional poverty and catastrophic health spending. Other variables used to explain this expense are: gender, occupation, affiliation to some insurance, marital status, educational level, and presence of members vulnerable to becoming ill. The results obtained show that this catastrophic expenditure is more likely to occur in multidimensionally poor households than in multidimensionally nonpoor ones. This highlights the importance of the State to develop policies and / or manage programs that protect the health of the population, especially the most vulnerable, in order to improve the growth and development of the country. / Trabajo de investigación

No description available.

Chapter 1: The Impact of Culture on the Second-Generation Immigrants’ Level of Trust in Canada Trust is one of the main elements of social capital; it determines the extent to which an individual cooperates with others. In this chapter, I assess whether cultural factors influence the level of trust in the population of second-generation immigrants in Canada. This paper is related to two strands of empirical literature. The first analyses the determinants of trust and the second studies the cultural transmission of values, attitudes and beliefs. I follow closely the literature on the cultural transmission and use an epidemiological approach to assess whether trust of second-generation immigrants is affected by their cultural heritage. This approach consists of comparing information about the outcomes of second-generation immigrants with that of the country of origin of their ancestry. We apply this approach using the Ethnic Diversity Survey (EDS), the World Value Survey (WVS) and the European Value Survey (EVS). Estimation results show that the average level of trust in the countries of origin of the ancestors of the second-generation immigrants has a strong significant impact on their level of trust. Thus, individual whose country of ancestry displays a high level of trust, tend to have a high level of trust. This provides evidence that individuals’ level of trust is not only explained by their personal experiences, characteristics, and the environment in which they live; but also by the culture in their country of ancestry. This means that culture does matter! I find that the results remain robust even if certain key countries are omitted or a different data set is used. Chapter 2: Decomposing Health Achievement and Socioeconomic Health Inequalities in Presence of Multiple Categorical Information This chapter presents a decomposition of the health achievement and the socioeconomic health inequality indices by multiple categorical variables and by regions. I adopt Makdissi and Yazbeck's (2014) counting approach to deal with the ordinal nature of the data of the United States National Health Interview Survey 2010. The findings suggest that the attributes that contribute the most to the deviation from perfect health in the United States are: anxiety, depression and exhaustion. Also, I find that the attributes that contribute the most to the total socioeconomic health inequality are ambulation, depression and pain. The regional decomposition results suggest that, if the aversion to socioeconomic health inequality is high enough, socioeconomic health inequalities between regions are the main contributors to the total socioeconomic health inequality in the United States. Chapter 3: Accounting for Freedom and Economic Resources in the Assessment of Changes in Women Poverty in Sub-Saharan Africa This chapter assesses the importance of freedom in women’s wellbeing in twelve Sub-Saharan Africa countries by using data from Demographic Health Surveys. This paper presents a poverty comparison by using the stochastic dominance approach and relies on the economic resources and freedom as the two aspects of wellbeing which evokes the multidimensionality of poverty. This study is related to the following three pieces of literature: the sequential stochastic dominance, the multidimensional poverty, the Sen’s capability approach which is based on freedom. This paper is built on Makdissi et al. (2014) but differs from it in a number of respects. First, it focuses on poverty instead of welfare. Secondly, it applies the Shapley decomposition to determine the contributions of the economic resource distribution and the incidence of the threat of domestic violence to poverty changes over time. Consistent with previous work on the importance of freedom, I find that more freedom, i.e. less threat of domestic violence, affects women’s wellbeing positively since it decreases women’s poverty. The results indicate that women’s wellbeing has improved in Burkina Faso, Ghana, Kenya, Lesotho, Madagascar, Malawi, Rwanda, Senegal, and Zimbabwe and deteriorated in Ethiopia, Nigeria and Tanzania.