Abstract

The research proposes a new model for measuring abnormal book-tax difference (ABTD) based on specific sources of book-tax differences regulated by the U.S. income tax code and U.S. GAAP, in order to capture non-conforming tax avoidance. The initial validation reveals that tax-shelter firms exhibit a significantly larger mean value of ABTD compared to their matched counterparts. Moreover, the t-test for ABTD yields a critical value (indicating the highest significance) that exceeds the mean values of the other three BTD-based tax measures. Additionally, there is a statistically significant positive association between ABTD and firm performance. Robust tests further demonstrate that the ABTD measure can be effectively used in the tax planning context of multinational corporations. This paper contributes to current tax research by introducing a new ABTD metric designed to quantify tax avoidance. The effectiveness of this metric has been successfully validated through practical examples, thereby establishing a connection between accounting practices and this newly introduced non-conforming tax avoidance. Future accounting researchers can utilize our ABTD measure to explore deeper financial accounting topics, such as examining the relationship between real earnings management and non-conforming tax avoidance.

Key words: Non-conforming tax avoidance, abnormal book-tax differences, financial reporting.