Workers’ tax evasion

We apply a direct method to estimate tax evasion in Italy assuming that tax evaders might consider declaring a closer-to-true income in an anonymous interview. The methodology is applied to employed and self-employed tax-payers, combining the Survey of Household Income (SHIW) by the Bank of Italy and a large random sample of tax forms by SeCIT, both referred to incomes received in 2000. Posing particular attention to the post-stratification of the data, we find that tax evasion is consistently higher for self-employment income than for employment income: the difference ranges from about 7% in lower deciles to 27% around the mode. This analysis shows that a relevant level of tax evasion arises also at low levels of employment income, although some under-sampling problems need to be considered. A description of tax evaders is attempted using standard OLS and quantile regression methods and an evaluation of the redistribution and incidence effects of tax evasion among workers is provided.