Asymmetric Long-run Effects of Tax Revenue on Economic Growth in Indonesia

Indonesia combines one of the lowest tax-to-GDP ratios among major emerging economies with a growth record punctuated by three systemic crises, yet the empirical literature on its tax and growth nexus assumes that revenue increases and decreases carry symmetric effects. This study examines whether the long-run relationship between tax revenue and real gross domestic product (GDP) is asymmetric, using a nonlinear autoregressive distributed lag model on annual data for Indonesia from 1994 to 2024, with domestic credit and the Corruption Perceptions Index as longrun regressors and capital expenditure in a parallel specification. A symmetric benchmark detected no cointegration and failed the functional-form test, whereas the asymmetric model established a level relationship. A 1 percent cumulative increase in tax revenue was associated with a 0.71 percent rise in long-run GDP, cumulative decreases showed no significant association and long-run symmetry was rejected (Wald F (1, 22) = 7.84, p = 0.010). A one-point improvement in the corruption index, measured on a 0 to 100 scale, not in percentage units, was associated with approximately 1.5 percent higher long-run output, while the capital-expenditure specification produced no long-run relationship, locating the asymmetry on the revenue side of the budget. The asymmetry persisted when revenue was measured in real terms and as a share of GDP, under alternative deterministic specifications and after a level shift in 1999 was allowed. These findings characterize revenue mobilization as growth policy rather than as consolidation alone, subject to the associational limits of a small annual sample.