As avaliações vão para a gaveta?: um estudo da utilização de avaliações de impacto no governo de Minas Gerais

Abstract

Although program evaluations constitute subsidies to make decision process more rational, motivated by a number of reliable information on the implementation and performance of government programs, they are not always used as they should. As international authors note, evaluative studies end up being misused, forgotten or even shelved. In Brazil and in the Brazilian state governments, despite growing interest in program evaluation, it is possible to observe shortage of empirical national studies that verify the use of such assessments as a tool for public services course corrections and improvement. In this context, this research aims to analyze the dynamics of the use of evaluation results of structuring projects undertaken since 2007 by the State Government of Minas Gerais. It is a qualitative research strategy that uses the case study and employs the techniques of direct observation, document analysis and, mainly, semi-structured interviews with different stakeholders involved in these assessments. In order to illustrate and discuss the theoretical research, the problem of the evaluation use is studied in cases of evaluations of four structuring projects evaluation: Viva Vida, Gestão lntegrada de Ações e Informações de Defesa Social, Minas Sem Fome and Acelerar para Vencer. In each of the cases are traced back the history of evaluation, since evaluation design and initial expectation, until it completion, with reference to the type of use occurred, using the respondents perception. We conclude that the State of Minas Gerais presents an evaluation system under construction, still fresh in Brazilian public administration and requires improvement in its institutional arrangement. The inability to accommodate the plurality of actors involved in public programs, ignoring the context, circumstances and purpose that involves the deployment of programs are some of the factors responsible for not using the results of evaluative studies.