Nesta tese propõe-se responder a três principais questões: No primeiro ensaio estuda-se o impacto dos investimentos privados de multinacionais do setor automobilístico no Brasil sobre o crescimento da atividade econômica dos municípios. Os resultados obtidos sugerem que as cidades onde se instalam este tipo de investimentos alcançam um desempenho superior aos outros grupos de municípios sem este recurso. Estes resultados apontam para uma subprovisão de bens públicos esperada em todo agrupamento de cidades, e um crescimento da atividade industrial no grupo de municípios de domicílio dos investimentos em detrimento a um crescimento abaixo da média do grupo de controle dos municípios adjacentes. No segundo ensaio proposto nesta tese, levanta-se a questão sobre qual o impacto do investimento em infraestrutura viária sobre a economia dos municípios do estado de Santa Catarina. Para responder esta questão, o método de regressão de dados em painel foi utilizado para identificar e mensurar o efeito do investimento público em rodovias interestaduais sobre o crescimento econômico municipal. Os resultados apontam para um efeito positivo sobre os municípios adjacentes e aqueles cortados pela rodovia federal. O setor agropecuário tende a perder valor agregado dado uma possível realocação de recursos para o setor de serviços e indústria de transformação. No terceiro e último ensaio, sobre a dinâmica da distribuição espacial da população no Brasil, tem-se como ideia central testar a Lei de Zipf tendo por base as estimativas da dinâmica populacional dos municípios brasileiros e seus limites populacionais (K). Recentes trabalhos na área de economia regional e urbana, e da nova geografia econômica (NGE), retomam o tema sobre a distribuição espacial da população entre as cidades. A principal questão neste ensaio é descobrir sobre a existência de uma tendência na distribuição do tamanho dos municípios brasileiros a atender a Lei de Zipf. Após as estimativas para um estado estacionário na dinâmica populacional dos municípios brasileiros e as estimativas por regressão quantílica da lei de potência, rejeitou-se em quase a totalidade das amostras a existência de um expoente igual a -1. / In this thesis propose to reply three principals’ questions. In the first essay examines the impact of private investment of multinational automobile industry in Brazil about the growth of economic activity of municipalities. The results obtained suggest that the municipalities that settle the investments reach a performance superior to other groups of municipalities. These results point to a underprovision expected of public goods in all groups of municipalities and increasing industrial activity in the municipalities of domicile of the investment over a growth below the average of the control group of adjacent municipalities. In the second essay proposed in this thesis, the question is about what the impact of highway infrastructure investment on the economy of the municipalities in the State of Santa Catarina. To answer this question makes use of panel date regression method to identify and measure the effect of public investment on the municipal economic growth. The results does are positive on the neighbouring and crossways interstate highway, the agriculture sector loses added value due a possible relocation of resources to service and manufacture sector. In the third essay, on the dynamics of spatial distribution of population in the Brazil, the core insight this is to test the Zipf’s Law from estimates of population dynamics of Brazilian municipalities and their population limits (K). Lately, studies on the approach of urban and regional economy and of new economic geography (NEG) recovers’ theme on the spatial distribution of population among cities. The central question in this essay, is whether there is a trend in the distribution of sizes of municipalities meet Zipf's law? After estimates for a steady state in the population dynamics of Brazilian municipalities and estimates for regression quantiles power law rejected in almost all of the samples the presence of an exponent equal to-1.

The main contribution of this thesis is a comprehensive analysis of the influence of changes in the population structure on local communities, in particular with respect to the provision of publicly provided goods. The focus is placed on the consequences of two of the major processes of demographic change, namely aging and shrinking. The three main chapters of this contribution consider the effects at the local level from both a theoretical and an empirical perspective. The first model focuses on the influence of population aging on the provision of local publicly provided goods, when the young population may relocate. When aging advances, gerontocracies and social planners substitute publicly provided goods aimed at the mobile young for publicly provided goods for the elderly. However, due to fiscal competition, gerontocracies will provide even more of the publicly provided good for the young than the social planner. The second model considers in a two-period setting, the interaction of a shrinking population when the investments made by the previous generation are long lived. The laissez-faire and welfare maximizing outcomes are computed for two cases; first with no costs of upkeep and second for the case when costs of upkeep accrue. A comparison of the solutions shows that public provision for the first generation is inefficiently low in laissez-faire when there are no costs of upkeep. However, if costs of upkeep accrue, the laissez-faire outcome for the intergenerational publicly provided good may be too high. Chapter four contains an empirical analysis. In a two-stage analysis the efficiency of the provision of child care services in municipalities is evaluated in the German State of Saxony. First, the results of the Data Envelopment Analysis (DEA) show substantial efficiency differences; the median municipality is up to 28% inefficient. In a second stage bootstrapped truncated regression, determinants of the inefficiency are identified. Explanatory variables such as an uncompensated mayor or a larger share of over 65-year-olds significantly increase inefficiency.