EVS data-based analysis of tax evasion: Descriptive vs. regression modelling

Abstract

Perception of tax evasion by individual citizens is of considerable interest to politicians, since people’s perceived attitudes affect the approach to tax compliance throughout the society. It is thus worth identifying personal characteristics that are related to a higher degree of tolerance and justification for tax evasion. Based on the 2008 European Values Survey data and using descriptive statistics, the paper discusses the relationship between the respondents’ characteristics and their tendency to justify tax evasion. The study finds a strong relationship between this tendency and age, educational attainment and economic activity, the two other variables (parenthood and income) indicating only a weak relationship. Moreover, the current issue allows us to convincingly argue against the regression analysis stereotypes which often yield biased and confl icting results. The paper confirms our constructive criticism, thus opening up space for an extended discussion of a more balanced use of both descriptive statistics and regression models.