Purpose The purpose of this paper is to examine the effect of corporate governance structure and CEO compensation on the level of tax aggressiveness. Design/methodology/approach This work analyzes a sample of 471 observations of 100 companies listed on the NASDAQ 100 for the period 2008-2012. It uses a fixed-effect panel model to analyze the effect of different model variables on the tax aggressiveness level. Findings The main finding of this study is the great influence of corporate governance structure and CEO compensation on reducing tax aggressiveness. Indeed, it finds a significant negative relation between board size, CEO salary, CEO stock options and tax aggressiveness. In addition, the study reveals that there is a direct negative relation between CEO duality, tax fees and tax aggressiveness. Research limitations/implications The study was conducted using robust methods to test the effect of corporate governance structure and CEO compensation on tax aggressiveness level. The generalized least squares method was used to fit panel data and overcome heteroscedasticity and autocorrelation problems. The aim of the study was to prove the great effect of both corporate governance structure and CEO compensation on reducing tax aggressiveness. As this study was based on data from American companies, the results cannot be generalized to all contexts. Originality/value This paper differs from previous work and tests the effect of corporate governance structure, CEO compensation, CEO characteristics and audit fees on tax aggressiveness. The findings of this study will enrich the literature on tax aggressiveness by suggesting that corporate governance structure and CEO compensation can significantly limit tax aggressiveness behavior. Therefore, shareholders must be aware of these two variables. They need to limit tax aggressiveness behavior, as it is usually accompanied by rent diversion, as reported by Desai and Dharmapala (2006). Therefore, these findings will be helpful to investors, managers and regulators because they have implications for the interactive decision-making process.
Read full abstract