Dynamic Fiscal Policy and Economic Growth in Bangladesh: Short-run and Long-run Evidence from ARDL Cointegration Analysis

Abstract

This study investigated the dynamic short-run and long-run effects of fiscal policy on the economic growth of Bangladesh using time series data spanning from 1999 to 2024. By utilising the Autoregressive Distributed Lag (ARDL) cointegration model, the analysis was conducted to examine the impact of government revenue, government expenditure, and government investment on Bangladesh's GDP growth. The ARDL Bounds Test confirms that among the variables, there is a stable long-term cointegration exists. The findings reveal that government expenditure has a statistically significant positive impact in both the short and long run, supporting the idea of a Keynesian stimulus effect. Conversely, government revenue shows a significant negative impact, indicating a short-term contractionary effect of taxation. A notable and surprising result shows that, in the short run, investment has spurred growth; however, in the long run, it has a significant negative impact, indicating diminishing returns or the crowding-out effect of private investment. However, ECT where the coefficient is suggesting a statistically significant result, which confirms the model's robustness and indicates that nearly 35% of the disequilibrium from the long-run relationship is corrected within one year. The economy required approximately three years (1/0.35) to immediately resume its long-run equilibrium path of growth acceleration after recovering from the shock. All diagnostic tests were performed, including normality, the serial correlation LM test, and heteroskedasticity tests, which confirm the model's robustness. This finding provides policymakers with critical insights for designing balanced fiscal strategies that promote sustainable economic growth in Bangladesh.