Análise da eficiência das universidades federais brasileiras: uma aplicação da análise envoltória de dados

Abstract

This dissertation had as main objective to analyze the efficiency level of the Brazilian Federal Universities. As specific objectives it was defined to verify the IFES considered benchmarks, as well as to determine the goals that non-efficient Universities should reach to reach the production frontier, in addition to an analysis of the IFES performance by Brazilian region. To achieve the proposed objectives, the Data Envelopment Analysis - DEA model CCR (constant returns to scale) was used, using management indicators for the year 2017 stipulated by the Federal Court of Auditors as inputs and outputs. In order to homogenize the Institutions that invest more in the graduate courses of the others, two groups were created through Cluster Analysis to later calculate the relative efficiency of each group. The results achieved from the standard frontier defined that of the 31 universities in Group 1, eleven were considered efficient and twenty were considered inefficient. Group 2, with twenty-two Federal Higher Education Institutions, seven were considered efficient and fifteen were considered inefficient. Also according to the model, efficient Institutions resulted in possible benchmarks for non-efficient IFES. For inefficient Institutions, goals or targets were proposed to be reached to reach the production frontier according to each group, with a view to improving their indicators Graduation Success Rate and Capes Concept. In analyzing the performance of Universities by region, it was observed that the majority of Institutions considered efficient are concentrated in the South and Southeast Regions.