The prediction of self-reported and hypothetical tax-evasion: Evidence from England, France and Norway

This paper sets out to test the Hessing et al. model of tax paying behaviour (which treats tax evasion as a defective behaviour within a social dilemma) using a new measure of hypothetical tax evasion. Two studies were carried out. In all, 400 questionnaires were distributed in Exeter and London (England), Oslo (Norway) and Paris (France). Two hundred and thirty five questionnaires were returned, 44 from Oslo, 64 from Paris and 127 from England. A wide range of independent variables were measured in the questionnaire including scales assessing equity, opportunity to evade, tax attitudes, personality variables and demographic variables. The dependent variables were self-reported tax evasion (income concealment or the overstatement of deductions) and hypothetical tax evasion. Respondents were classified into three groups: 'non-evaders' are those who answered no to both evasion measures, 'hypothetical evaders' agreed that they might evade tax but did not report ever having done so, and 'self-reported evaders' are self-reported evaders who are also mainly hypothetical evaders. On motivational and attitude measures hypothetical evaders were similar to the group of admitted evaders but different from the non-evader group, as the Hessing et al. model would predict. On perceived environmental constraint measures the opposite was the case, with hypothetical evaders being similar to the non-evader group and different from the admitted evaders. This suggests that the reason hypothetical evaders do not actually evade tax is that they perceive the environmental constraints to be greater than do the self-reported tax evaders. (C) 2001 Elsevier Science B.V. All rights reserved.