Can Ethics and Economic Incentives Mitigate Tax preparers’ Aggressive Judgements and Decisions?

ABSTRACT

We examine whether the personal attribute of ethics and an external economic incentive in the form of a standard tax preparation fee can mitigate tax preparers’ aggressive tax reporting behaviour. We use a structural equation model (SEM) to identify the influential paths of ethics and economic incentives on the aggressiveness of tax preparers. Based on the results of a quasi-experiment involving tax preparers in Korea, we find that ethical attributes both directly and indirectly mitigate aggressive tax reporting by reducing tax preparers’ fear of client dissatisfaction. We then find that the introduction of a standard tax preparation fee does not directly or indirectly suppress the tax aggressiveness of tax preparers. Our study contributes to an integral examination of the effectiveness of both psychological and economic factors on tax preparers’ judgements and decisions.

KEYWORDS:

• Aggressive tax reporting
• ethics
• self-serving bias
• standard tax preparation fee
• tax preparers

JEL CLASSIFICATION:

• C93
• H26

Acknowledgments

We acknowledge helpful comments and suggestions by Chia-Hsiang Weng from National Chengchi University and all the participants at the 2021 Japanese Accounting Association Annual Meeting.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ According to prior studies (Cuccia, Hackenbrack, and Nelson 1995; Roberts 1998; Vermeer, Spilker, and Curatola 2020), we define aggressiveness as adopting a tax position in a way to reduce tax burdens in the resolution of grey areas.

² For a description of qualified tax preparers in Korea, please refer to ‘2.1 Overview of the Korean tax preparation service market’..

³ National Tax Statistics (https://stats.nts.go.kr).

⁴ KACPTA divides tax services into three high-level categories (i.e. bookkeeping and tax reporting for income taxes; tax reporting for property taxes; and tax consulting services) and further breaks down each service category into specific tasks (e.g. data gathering, preparing qualified evidence for tax deduction, etc.). KACPTA proposes to determine expected service hours by service category and multiply these service hours by standard unit prices to calculate standard tax preparation fees. Different standard unit prices are applied for services with different degrees of difficulty.

⁵ A similar statement is found in Article 1–2 of the CTA Act.

⁶ The experiment in this study was approved by the Institutional Research Board (IRB).

⁷ NICE Data and Research is a Korean company that provides research services. This company discloses the contact information of CTAs on its website with each CTA’s consent.

⁸ We first disseminated questionnaires to CTAs whose offices were registered in the Seoul capital area. Because of the low rate of responses, we next disseminated questionnaires to those located in non-capital areas. For both groups of participants, we collected responses for two weeks. We also sent reminders twice after the initial dissemination of the questionnaires. A t-test showed that there were no significant differences in the responses between the two groups.

⁹ Before conducting the principal component analysis, we conducted a Kaiser-Meyer-Olkin (KMO) test and Bartlett’s test to check sample adequacy. The results of the KMO test (0.502) were slightly poor but marginally acceptable according to Kaiser (1974). The results of Bartlett’s test (41.435; p < 0.001) indicated that our sample is adequate for principal component analysis.

¹⁰ We re-examined the analysis using a latent variable and a summated rank scale variable for ETHICS. Consistent results were obtained, thereby supporting the robustness of the ETHICS variable.

¹¹ Before we conducted the SEM, we performed an analysis of covariance (ANCOVA) with the two independent variables and all demographic variables. We observed no significant effect of any of the demographic variables. Therefore, we include only the two independent variables and the mediating variable in our SEM analysis.