Uncertainty about the probability of a tax audit : raising compliance or confusion?

The purpose of this paper is to investigate the governmental use of uncertainty in the application of tax policy. The main query examines if the government is able to raise compliance through creating uncertainty about the probability of audit. The theoretical portion of this paper uses a two-period expected utility maximizing Bayesian learner; a model that is originally postulated by Gollier (2003) for use in the context of the insurance industry. This paper references empirical work in the use of uncertainty in taxes, as well as the legality of withholding the probability of audit under the Freedom Information Act. The goal of this paper is to prove that an expected utility maximizing Bayesian learner reduces the optimal amount of tax evasion in the first period if the ratio of absolute prudence to absolute risk aversion is smaller than two.