Altering norm perceptions to increase tax compliance

Abstract

This study tested the effects of an intervention designed to increase compliance and honesty in work-related expenses (WRE) and other deduction claims. The research follows on from earlier work documented in Centre for Tax System Integrity (CTSI) Working Papers No. 7 and 8. Overall, the research adds further empirical support to the usefulness of such a norm-based approach to increase tax compliance, identifies possible limiting and facilitating conditions and encourages further attempts to apply such a strategy more broadly. The intervention was based on the assumption that taxpayers overestimate the level of endorsement of tax cheating among Australian taxpayers. People tend to think they themselves have more ethical views about taxpaying than they think most others have. Given the possibility that taxpayers are influenced by what they consider to be the social norm, such an overestimation of tax evasion and its endorsement by fellow Australians could lead to a decline in tax compliance. Therefore, the intervention tested in this study aimed to correct people’s misperception of the social norm and thus increase compliance in tax deduction claims. In this study, 2250 individual taxpayers were randomly assigned to three different experimental conditions. In the norm feedback condition, taxpayers received a letter that informed them about recent survey findings, indicating that people think most others accept tax cheating more than they actually do. Then taxpayers were asked to participate in a survey asking further questions on this finding. The two other conditions served as control groups. In the survey only condition, taxpayers did not receive any information about the previous research findings and were only asked to participate in the survey. In the no survey condition, taxpayers were not contacted at all; no survey data was available. Actual tax return data were compared between all three groups to investigate the effects of the intervention on levels of deduction claims. The survey data showed that the norm feedback had some effects in altering people’s perceptions of the social norm, but these depended on participants’ level of income and their degree of identification as Australians. The tax return data showed that WRE claims were significantly lower in the norm feedback condition than in the two control conditions, which in turn did not differ from each other. This effect was not further moderated by income level. In contrast, other deduction claims (gifts/donations and dividends/interest) showed no differences between all three conditions.