The tax avoidance literature identifies dozens of variables that are related to corporate tax avoidance, but it lacks a framework for understanding the relative importance of these variables and the theories that drive these relations. We distill decades of tax avoidance research into key theories and create a unified sample that contains 31 independent variables from prior literature to quantify the relative and total importance of proposed explanations. Using Shapley value variance decomposition, we show that investment-related factors and operating metrics dominate effective tax rate (ETR) measures and consistently rank high across a range of tax avoidance proxies. When we examine the relative importance of time-invariant effects, we find that manager, firm, and segment fixed effects explain the greatest variation in tax avoidance. We offer several actionable opportunities for future research and offer several key takeaways for researchers and policymakers. This paper was accepted by Eric So, accounting. Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2024.04839 .
Belnap et al. (Tue,) studied this question.