Sample size determination for risk-based tax auditing

Abstract

A modern system of Revenue Administration requires an effective and efficient management of compliance which in turn requires a well designed taxpayers audit strategy. The selection of taxpayers to be audited by Revenue Authorities is a non-standard sample size determination problem, involving an initial random sample from the population and, based on the statistical information derived from it, a risk-based auditing scheme whose sole objective is to select for auditing the taxpayers with the highest estimated risk in the population. This paper provides a methodological approach that estimates the initial optimal random sample size such that the Revenue Administration Authority maximises their expected tax revenues. The methodology is illustrated using administrative data from the UK’s Revenue Authority, Her Majesty’s Revenue and Customs (HMRC).