Do tax revenues track economic growth? Comparing panel data estimators

Abstract

Determining how economic growth affects tax revenues is crucial for fiscal sustainability, economic stabilization, and policy design. The current literature on tax buoyancy presents contrasting estimates, highlighting the need for a systematic discussion of the trade-offs associated with different estimators. This paper provides new empirical evidence by reviewing a range of panel data estimators in a large sample of 172 countries from 1990 to 2019. We find evidence of lower estimates for tax responses to economic activity in the short term relative to previous literature, suggesting a limited automatic stabilization power of tax systems. The heterogeneity in our results within and across income groups underscores the importance of choosing the appropriate estimator. Our results remain broadly unchanged when we introduce new control variables to disentangle discretionary from automatic tax revenue variations, indicating that economic cycles do not significantly influence the timing of tax policies.