Shakespeare is often quoted as having said that death and tax are the only two certainties in life; many people of the modern world would perhaps agree that he would have been at liberty to add tax evasion to the list.

“Tax Evasion with Endogenous Reference Points” was the title of a seminar held by Teimuraz Gogsadze, a Ph.D. candidate of the University of Leicester and an ISET graduate of the Class of 2011. Gogsadze examined how taxpayers make tax compliance decisions and respond to changes in tax policies, such as the tax rate, the penalty rate, and the probability of an audit.

Gogsadze overviewed tax evasion behavior and discussed the inability of mainstream decision theories in economics – namely expected utility theory – to explain the evidence. For instance, it was stressed that people are found to evade tax more when the tax rate increases, while expected utility theory predicts a reverse tax rate-compliance relation. Gogsadze discussed the possible explanations of the theory’s failure, arguing that an individual might not only enjoy the final level of wealth but she/he might also care about the deviations from her/his reference wealth level; the latter being supported by the evidence that falling below the reference point induces a sensation of losses that bites more than the equivalent gains are enjoyed, a feature is known as the loss aversion.

It was underlined that an alternative decision theory, the so-called prospect theory of Kahneman and Tversky (1979), (in which an individual is loss averse and only cares about deviations from the reference point and not the final level of wealth), explains tax evasion behavior and correctly predicts the tax rate-compliance relation if the reference point of the taxpayer is their legal after-tax income. Therefore, the reference point must be ex-ante, exogenously chosen under a prospect theory framework.

The final part of the talk was dedicated to the introduction and the application of yet another decision theory of Koszegi and Rabin (2006, 2007), which also builds upon the reference-dependence and loss aversion, but allows the reference point to be endogenously determined within the framework.

Regardless of the increasing empirical and experimental support that the Koszegi-Rabin theory has recently gained, Gogsadze showed that while the theory is capable of explaining some of the empirical facts – which expected utility theory cannot – it is unable to correctly predict the tax rate-compliance relation.

Gogsadze concluded his presentation by stating that Kozegi-Rabin theory overall performed better than expected utility theory, but worse than prospect theory in explaining the stylized facts of tax