Mudanças no perfil de pobreza no Brasil: uma análise multidimensional a partir da Pesquisa de Orçamentos Familiares (POF) 2002-2003 e 2008-2009

• Ano de defesa: 2014
• Autor(a) principal: Rodrigues, Cristiana Tristão
• Orientador(a): Helfand, Steven M.
• Banca de defesa: Braga, Marcelo José , Salvato, Márcio Antônio , Neder, Henrique Dantas , Coelho, Alexandre Bragança
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Viçosa
• Programa de Pós-Graduação: Doutorado em Economia Aplicada
• Departamento: Economia e Gerenciamento do Agronegócio; Economia das Relações Internacionais; Economia dos Recursos
• País: BR
• Palavras-chave em Português: Pobreza; Consumo; Orçamento familiar
• Palavras-chave em Inglês: Poverty; Consumption; Family budget
• Área do conhecimento CNPq: CNPQ::CIENCIAS SOCIAIS APLICADAS::ECONOMIA::ECONOMIA DO BEM-ESTAR SOCIAL
• Link de acesso: http://locus.ufv.br/handle/123456789/160

For a long time, there has been controversy in the poverty literature about which indicator would be most appropriate to measure the well-being of families. Consumption, for various reasons, has become identified as a more accurate measure of well-being than income. Given these facts, there is a need to define a measure that captures welfare more consistently. Therefore, in the first part of this research the aim is to estimate a new poverty profile for Brazil and analyze the changes that have occurred in the period 2003-2009, using consumption from the Household Budget Survey (POF) as a measure of well-being. The paper presents the methodology for the estimation of a food poverty line and an adjusted poverty line for non-food items, as well the methodology used to construct appropriate summary measures of poverty. In addition, it presents and imputation methodology to correct measurement error for food consumption from POF. The conclusions from this study corroborate the general results from other studies in Brazil: poverty has been decreasing over the last decade. However, the big difference is that the results from this paper are based on consumption as a welfare measure, which has not yet been presented in the literature. By using this measure of well-being it is noted that the incidence of poverty is lower than when monetary income is used as a measure of well-being. The decline in poverty is also less pronounced when consumption is used. Therefore, the use of monetary income tends to overestimate the incidence of poverty and generates a steeper decline than when consumption is used. This result becomes relevant due to the fact that monetary income has traditionally been used to calculate poverty rates, despite consumption being pointed in the literature as a measure of well-being more accurate than monetary income. Furthermore, it is important to note that the poverty rates estimated based on the poverty lines of food and non-food items are bigger than the poverty rates that are estimated based on the government administrative xxlines used for Bolsa Família. Thus, the government by using the administrative lines may be failing to assist significant portion of the population that needs assistance. Although superior to income in many regards, recent studies have emphasized that, consumption expenditure still only captures a single dimension of living standards and is thus also insufficient for measuring poverty. Therefore, the replacement of income by consumption in defining poverty lines does not solve the problem of insufficiency of simple indicators for the analysis of quality of life and therefore poverty. We provide estimates of poverty based on a broader set of variables that capture other important dimensions of living standards. These include housing conditions, education, nutrition, health and employment, as well as subjective assessments of living conditions. The methodology is based on Alkire and Foster s (2011) "dual cutoff" approach to measuring multidimensional poverty. The approach is applied to data from the 2002-03 and 2008-09 Household Budget Surveys (POF). The main conclusion that comes from this study is that relative to income poverty, which fell sharply in the last decade, multidimensional poverty has declined more gradually. Multidimensional poverty is a more serious problem in rural areas and in the North and Northeast. Thus, the Multidimensional Poverty Index can contribute to the planning of social policies to avoid poverty, since it reveals the depth of poverty in different population groups, enabling better targeting. This is important because it allows areas with very severe poverty to be prioritized. Moreover, the index also provides information about the dimensions that contribute most to poverty, which can help to maximize the effect of social investments and to guide public policy. Decomposition analysis shows that deprivations in education, basic consumption needs and housing conditions contribute the most to aggregate multidimensional poverty. Improvement in the perceptions that families have of their own living conditions corroborates the decline in multidimensional poverty, which is indicative of improvement in the living conditions of the population. Decompositions, however, paint a somewhat different picture, with income, food and health insufficiency contributing the most to subjective multidimensional poverty. Although poverty has fallen, the objective and subjective multidimensional poverty rates remain high.