Reporting tax evasion

Abstract

I study the anonymous tax evasion reports collected by the website evasori.info for Italy. I find that tax morale, measured by the number of reports per unit of irregular activity, is negatively associated with the median reported amount. If the utility of reporting evasion is increasing in the amount of the transaction, then, for a fixed cost of reporting, there exists a threshold value above which tax evasion is reported. Stronger tax morale implies a bigger utility for the individual that reports the transaction and, therefore, a lower threshold value above which tax evasion is reported. The empirical analysis features a regression of the median reported evaded amount per economic activity on tax morale, controlling for province fixed effects, economic activity fixed effects and time dummies. I also propose a simulation exercise to estimate the threshold below which tax evasion is reported and its dependence on tax morale. The simulation also shows that the empirical results are not the byproduct of oversampling of small transactions. I conclude that tax morale makes it more difficult to overlook petty evasion.