Introduction

This paper is about tax cheating by firms. Our aim is to compare the audit strategies available to a tax authority attempting to collect indirect taxes from sellers prone to evade. We model indirect tax evasion as follows: firms choose between taxable sales on the regular market and unreported sales in an informal market. The latter will typically be for cash. There are three options open to each firm: specialising in legal sales (honesty), selling on both markets (diversifying) or specialising in informal sales (submerging). We call firms that choose the third option ‘ghosts’. Section 2 of the paper outlines our model.

As in Allingham and Sandmo (1972), tax evasion is a gamble. If a firm is audited and found to be making irregular sales, it is fined and made to pay back the evaded tax. Which firms should the authority audit? One possible strategy is to audit randomly, with some fixed probability that any firm is investigated. An alternative enforcement policy is to tailor the audit strategy to take into account what the authority knows about each firm. A simple form of this approach is where the authority conditions audit probability solely on reported turnover via a cut-off rule: those reporting less (no less) than a certain amount are always (never) audited. In Section 5 we compare this cut-off policy with random audit and investigate which is likely to yield the authority more revenue.

The authority may also have access to information which is likely to be correlated with a firm's turnover. Such information will define its audit class in the sense used by Scotchmer (1987).