Pilot CEOs and Tax Avoidance: Evidence from Machine Learning Methods

Abstract

Chief executive officers (CEOs) could use tax avoidance to reduce costs and hence increase firm profits. Nevertheless, the reputation of the firm could be harmed. Prior studies find that firms’ tax avoidance is associated with CEOs’ personality traits. We attempt to predict firms’ engagement in tax avoidance using CEOs’ flying hobby measured by the pilot certificate. Using various advanced machine learning methods, we find that CEOs’ hobby of flying airplanes is significantly associated with increased corporate tax avoidance. In particular, we find CEOs with a student, private, or airline transport pilot certificate are more likely to engage in corporate tax avoidance while CEOs with a commercial pilot certificate are less likely to engage in corporate tax avoidance. This study advances prior literature and helps investors, regulators, and policymakers to predict firms’ tax avoidance strategies using executives’ hobby.

Keywords: Machine Learning; Big Data; Tax Avoidance; Pilot CEO; CEOs’ Hobbies; Sensation Seeking; Personality Trait.