Risco orçamentário nos estados brasileiros e no Distrito Federal: uma proposta de aplicação de técnicas de tracking error sobre as despesas públicas

Abstract

The intervention of the State in the economy has the assumption of market failures and occurs mainly through the budget. However, intervention by the State may not meet the expectations of its main funders the citizens. There are instruments that allow budgetary changes, contingencies and unlinked revenues, allowing for the execution not to occur as set forth in the Annual Budget Law (LOA). Such instruments provide flexibility for public managers to allocate resources. In view of this, what is the level of divergence between the expenses established in the LOAs of the States and the Federal District and the expenses incurred? The main objective of this research was to evaluate the differences between the expenses established in the LOAs and the expenditures executed in the States and in the Federal District. Significant and systematic deviations are indicative of low budget reliability. The specific objectives were to understand the budgeting process and possible budget change needs throughout implementation; to investigate the instruments allowing expenditure to be implemented that diverge from the budget, as well as its applications; to adopt tracking error techniques in public finances as a way of measuring the divergence between budgeted and executed expenses; to apply the PEFA-framework methodology, which is an international standard for evaluating the management of public finances, to provide a basis for international financial-budgetary comparison. In the context of Agency Theory, public managers are supposed to be utility-maximizing agents who can use free resources to privilege particular interests. A similar situation is observed in corporate finance when investment fund managers have some flexibility to act. In this case, investors use the tracking error to analyze whether the fund management has achieved the strategy to which it has committed. Therefore, this research proposed to apply techniques similar to the tracking error in public finances as methodology to measure how much the execution has diverged from the values previously established in the budget and thus, to analyze the budget reliability. Three complementary techniques were applied: percentage variation, linear regression analysis and PEFAframework methodology. It was found that on average the authorized expenditure was higher than the expenditure initially set in the LOA by approximately 15%. Capital expenditures showed greater volatility and greater variation than current expenditures. Expenses allocated to the functions of Education, Health, Special Charges and Public Safety showed the smallest variations, while the expenses allocated to the functions of Industry, Housing, Energy, and Basic Sanitation presented high variations. In the PEFA-framework methodology, expenses classified by function were rated D. In short, expenses that went through rigid control were the ones that presented the least divergence. The budget reliability was critical to other expenses.