Abstract:
This study estimates the effect of economic, political, and institutional factors on Ecuador's budget deficit during the period 1990-2022. A distributed lag autoregressive model is used, given the non-stationarity of the variables and their different orders of integration. The limit test and the analysis of the adjustment speed coefficients validate the existence of a long-term relationship. The long-term coefficients (p-value in parentheses) show that: (i) a 1 pp increase in the real GDP growth rate reduces the deficit by 0.24 pp (p = 0.012); (ii) a 0.1 in the V-DEM index of electoral democracy reduces it by 2.3 pp (p =0.003); and (iii) inflation is only relevant in the short term (+0.025 pp for each additional point, p = 0.078), while political corruption and public debt do not reach statistical significance. The bounds test validates the existence of a cointegration relationship, and the estimated speed of adjustment (-0.92, p < 0.01) implies that 92 % of any deviation is corrected within one year, evidencing a rapid return to fiscal equilibrium. The study expands the Latin American literature by demonstrating that democratic quality not only moderates political-budgetary cycles but is also the strongest quantitative determinant of the deficit in the long term, surpassing the effect of classic macroeconomic variables. In addition, it provides evidence that dollarization acts as a disciplinary mechanism by eliminating monetary financing.
