The spend-and-tax or tax-and-spend: Further evidence for the Brazilian imperial period

This article tests the flows of rents during the Brazilian Imperial period. To achieve this goal, a Vector of Error Correction Model (VECM) was employed to test longrun and short-run relationships between government revenues and expenditures. The VECM was applied for the entire imperial period with data available (1836-1889) and for the period after the Law Alves Branco (1844-1889), which more than doubled tariffs on imports. A trivariate causality test fails to show a casual relationship among the variables in any direction, regardless of the period tested. When the augmented granger causality test is employed for the entire period, results show a unidirectional causality from government expenditures to revenues, a spend-to-tax model, and a bi-causality relationship for the 1844-1889 period.