Os determinantes da despesa com pessoal no poder executivo nos estados brasileiros: evidências em dados em painel (2013-2023)

Abstract

This research investigates the determinants of personnel expenditure in the Executive Branch of the 27 Brazilian federative units between 2013 and 2023. Personnel expenditure constitutes the most rigid and voluminous component of subnational public spending, and its control is a central pillar of fiscal health under the aegis of the Fiscal Responsibility Law. Methodologically, the study adopts static and dynamic panel data models, utilizing the fixed effects approach, ratified by the Hausman Test, and the DIFGMM estimator to address endogeneity and the temporal persistence of expenditures. It analyzes the sensitivity of expenditure in relation to human capital (average schooling), the macroeconomic environment (real gross domestic product and unemployment rate), and fiscal constraints (indebtedness). The results from the static models indicate that personnel expenditure is markedly pro-cyclical, presenting a positive and significant elasticity in relation to real gross domestic product (+0.683), which confirms that revenue increases during periods of economic prosperity tend to be absorbed by payroll expansion. Robustly, it is identified that the population's education level exerts the greatest impact on spending increases, presenting an elasticity greater than unity (+1.494), validating Baumol's “cost disease” hypothesis in the state public sector. Robustness analysis via DIF-GMM validated the consistency of structural determinants, overcoming endogeneity limitations. The dynamic model identified significant temporal persistence of expenditures (0.557), ratifying a budgetary rigidity that challenges subnational fiscal management. In this dynamic context, in which monetary data were deflated to isolate real growth, the impact of human capital has remained high (+1.383); however, the GDP variable completely lost its significance, revealing that the apparent pro-cyclicality reflected a nominal price trajectory, which was superseded by the inertia of the administrative structure. Contrary to the hypothesis of immediate rigid budgetary constraints, the indebtedness and unemployment variables did not show statistical significance, evidencing that the high institutional and legal rigidity of the payroll in Brazil prevents market shocks or financial liabilities from acting as effective disciplinary mechanisms in the short term.