Dynamics of Causality Relationships between Fiscal Policy Components: New Evidence from Wavelet Approach

The 2008 global financial crisis and the new coronavirus disease (COVID-19) pandemic have attracted interest in the issue of fiscal policy. Since fiscal policy plays an important role in alleviating the costs of these crises, understanding the relationships between fiscal policy components is crucial and has important implications for choosing fiscal policies in the field of public economics. This study aims to examine the causal links between the fiscal policy components, i.e., government expenditures (current and development) and government revenues (tax and oil) in Iran, using quarterly data for the period of 1990:2-2019:1. For this purpose, first, we employ the time domain Toda-Yamamoto causality test to check the causal relationship among these variables. Then, due to the various characteristics of variables in the frequency bands, we implement a dynamic analysis through wavelet coherence approach and wavelet phase-difference in order to explore the joint time-frequency domain causal relationship between government revenues and expenditure categories. The results of the wavelet analysis show that the linkage between the government revenues and expenditures pairs is not the same across all time horizons and a strong heterogeneity in the revealed interrelationships is detected over time and across scales. Overall, the results reveal various causal effects and confirm the expenditure dominance hypothesis for oil revenue, and revenue dominance hypothesis for tax revenue at different frequencies.