Distribution of tax revenue among municipalities - which are constitutionally official federal entities in Brazil - is highly unequal vis-à-vis their demands of public services. Given conurbation processes and intense urbanization in the second half of the past century, some municipalities concentrate resources and quality public services whereas neighboring municipalities harbor low-income workers who have to commute back to employment opportunities. As a result, urban landscape in Brazil is far from homogeneous displaying ruptures in level of access to transportation networks and leisure activities, education and safety, among other public goods. Given this context, this paper has a threefold objective. Firstly, it identifies the magnitude of these tax revenue inequalities by municipalities using exploratory spatial analysis and detailed fiscal information for 5,212 municipalities out of 5,565 ones in Brazil. Thus, this study describes and characterizes neighboring municipalities with high and low tax revenue vis-à-vis their GDP per capita, population and other indicators. Secondly, it tests the efficiency of public services offer using Data Envelopment Analysis (DEA). In order to do so, level of expenditures in health services per capita is used as input compared to quantity of health basic attention level services provided. Simultaneously, level of basic education expenditure per capita is compared to educational achievement by municipality. This is done for the sample of municipalities with high and low tax revenue per capita. Finally, we use econometric analysis to infer what the main determinants of the observed efficiency are. Thus, this paper highlights two cumulative issues that municipalities as institutions have to face in order to provide quality service to its citizens: raw resources availability compared to need of inhabitants on the one hand, and efficiency and its determinants to transform those resources into services on the other hand. Results indicate that there are 70 municipalities which concentrates income resources significantly higher than their neighbors. This so-called high-low group shows efficiency lower than other groups with worse achievement in health and education. Municipalities classified as low-high - which are in rich regions but have comparatively lower income available - fare better than the other groups with the highest efficiency rate of the sample, although with lower absolute results. In sum, the efficiency analysis confirms that there are heterogeneities in the ability to provide public services with results that differ from the Southeast/Northeast (rich/poor) typical pattern of empirical analysis for Brazil. The determinants of efficiency are in accordance to the literature, confirming that excessive transfers to small municipalities significantly reduces efficiency. Therefore, we can provide practical public policy recommendations to be enforced based on the results.
