Tax enforcement and revenue dynamics under alternating leniency and stringency policies

This paper examines how alternative policy approaches to tax enforcement adopted by governments — leniency (tax amnesties and settlement schemes) and stringency (credible enforcement strategies against tax evasion)— affect the dynamics of personal income tax (PIT) revenues from self-reporting taxpayers in Italy. We exploit a well-defined policy window (2002–2012) characterized by a clear alternation between leniency -oriented and stringency -oriented tax enforcement policies across different government coalitions. Using a detailed chronology of these enacted measures, we construct the corresponding policy classification and estimate cointegration and error-correction models on monthly administrative PIT payments. Robustness checks include a placebo on employees and an out-of-window extension covering the post-2012 “mixed” policy phase. The results show that stringency -oriented policies are associated with significantly higher PIT revenues from self-reporting taxpayers, whereas the leniency approach coincides with lower receipts. No policy effects are detected for employees, supporting the identification strategy, while post-2012 estimates are inconclusive, consistent with blurred policy signals and weaker identification. From a policy perspective, the evidence suggests that a credible and visible stringency approach is more conducive to sustained revenue than recurrent amnesties and settlement schemes, which risk undermining enforcement credibility and tax compliance by shaping expectations of future concessions.