In this paper we analyze a fiscal mechanism used in Italy, which in Italian is called “Studi di Settore” (Sector Studies). This mechanism relies on information gathered on taxpayers to both partition the population into fairly homogeneous clusters and to determine the presumed income they should declare. When this estimated income is announced, before taxpayers fill out their tax returns, their optimal declaration strategies lead the tax- payer population to be naturally split into three homogeneous groups, one of which pays more taxes than are due, the second group comply but bears the audit cost, while the third evades and it is not audited. This result is close to the Italian situation where the greatest number of taxpayers make a tax declaration according to the announced cluster income, but there are always those who declare less and so are audited.