Tax Fraud and Investigation: A Framework for a Knowledge Based Methodology

Abstract

Tax Fraud is a criminal activity done by a manager of a firm or a tax payer who intentionally manipulates tax data to deprive the tax authorities or the government of money for his own benefit. Tax fraud is a kind of data fraud and fraudsters in business manipulate balance sheets, revenues and expenses data. Such fraud happens every time and everywhere. Fraudsters may be individuals, households, firms, foundations or political parties. Even religious communities are not free of betraying. Tax fraud betrayers manipulate bookkeeping figures and tax declarations either by increasing expenses or decreasing revenues. There is no, and never will be not a clear boundary between fraud and the balance sheet policy of firms, especially if we consider enterprises where accounting and valuation latitude is utilized. Tax fraud investigation performed by a tax fraud authority can be embedded into approaches called “Knowledge-based systems”. There exists a great lack of data due to the singularity of a single fraud event. Therefore missing data must be substituted by assumptions, experience and ideas. In this paper we focus on the Bayesian Learning Theory. In this theory hints, belief, investigation, evidence, integration of partial information, subjective probabilities, and belief updates are the main elements of the methodology.