Model of performance indicators for the evaluation of state public health expenditures in Brazil

Objective: To present a proposal on a set of performance indicators that can be used in audits of state audit courts to assess the effectiveness of budget spending on public health. Methodology: This is quantitative research, with the application of statistical techniques related to Data Envelopment Analysis (DEA) in two stages and estimation of linear longitudinal regression models for strongly balanced short panel data, on the budget expenditures made in the period from 2015 to 2019 by the States and the Federal District on public health. Results: It was found that there is no use of performance indicators in the elaboration and execution of Multi-Annual Plans (MAPs) by governments to evaluate the effectiveness of spending on public actions; that the mention of "performance indicators" refers exclusively to percentages to compare what was accomplished in relation to what was predicted; that records of duplicate public actions were found in the MAPs by some entities, and that overvaluation was identified in the records of the resources provided for in most of the MAPs. The degree of statistical explanation of the proposed set of indicators reached an average of 87% of statistical confidence. None of the public entities was 100% effective in the expenditures made. 69% of the auditors of the Courts of Auditors legitimized the set of indicators proposed to evaluate the effectiveness of state public health spending. Finally, it is concluded that the proposed performance evaluation indicators can be used to verify the effectiveness of state public health spending. Contributions of the study: It presents performance indicators, created by means of a statistical method, to evaluate the effectiveness of spending on state public health, since the specialized literature indicates little or no discussion about it. It indicates that the actions included in the MAPs are not audited by the courts of auditors and that the legislative branch is silent in the supervision between what is foreseen and what is carried out, considering that there are repeated and overvalued actions.