Dissertação
Análise comparativa da eficiência alocativa das despesas públicas dos municípios brasileiros utilizando técnicas de mineração de dados
Fecha
2019-12-13Autor
Paula Guelman Davis
Institución
Resumen
Currently, in Brazil, the financial situation of the local governaments is very complicated. Budgets are smothered by internal costs and debt. This situation undermines the investments demanded by the population. Thus, it is necessary for the government to be able to allocate the available resources more efficiently, that is, with the best cost-benefit relation. The purpose of this study is to analyze the allocative efficiency of public expenditures of Brazilian municipalities using data mining techniques. The municipalities considered efficient were those that had the best relation between their expenses and their results in the socioeconomic indicators, that is, lower expenses and better results when compared to other cities. First, financial and operational data were collected and integrated. These were transformed into indicators and went through a stage of cleaning and selection to be discretized, that is, transformed into nominal attributes that allowed in the classification into ranges for all indicators. From the use of specific algorithms, it was possible to induce decision trees and association rules, which allowed the identification of common characteristics of the municipalities with outstanding performance. Then, the composition of the expenses of the municipalities that made up these groups was analyzed. Efficient municipalities in education spent under R$ 6,000.00 per student in a year and stood out in the Basic Education Development Index (IDEB). The health-efficient municipalities spent under R$800,00 per citizen in a year and achieved the best rankings within the Health Unique System Performance Index (IDSUS). Efficient municipalities in development were within these ranges simultaneously or in only one area and had a high or very high Municipal Human Development Index (IDHm). The study demonstrates the feasibility and importance of using data mining on large data sets. It was possible to identify relevant and unexpected patterns in the areas of education, health and development, such as the importance of spending resources on teacher training and the existing regional differences in terms of the performance of the Unified Health System (SUS).