Identifying the Effect of Earnings Management on Measures of Tax Avoidance: Implications for Empirical Research

IntroductionThere are contradictory results in some areas of tax avoidance literature that raise the question that if the measure of tax avoidance represents truly the concept and what components they are included. The most of measures of tax avoidance use pretax financial accounting income as a benchmark so earnings management methods might affect them. This can bias research results and researchers should pay attention to the effect. They need to separate the effect because it represents different underlying actions of managers.  HypothesisThe purpose of this study is to present a theoretical framework and relevant empirical tests to examine the fact that measures of tax avoidance reflect what types of tax avoidance and when they identify tax avoidance related to or unrelated to earnings management. In this research, four types of tax avoidance will be used to provide a theoretical framework that how measures of tax avoidance reflect these four types of tax avoidance. Any type of tax avoidance represents different underlying actions of managers and might be considered differently by policymakers and researchers.  MethodsComparative statics analysis will be used to determine the role of earnings management on measures of tax avoidance. in empirical tests, the annual data of 153 companies listed on the Tehran Stock Exchange during the period 2012 to 2020 were analyzed. our tests examine the extent to which measures of tax avoidance are affected by earnings management and will be relied on previous research.In the first test, Givoly et al. (2010) will be followed to examine the distributional properties of earnings around the zero earnings growth threshold for firm-year observations that are defined to have above median levels of tax avoidance for each given measure.the second test is based on the sign and magnitude of unexpected accruals for those observations that fall above the earnings threshold. This test follows the intuition of Givoly et al. (2010) that firms that are managing earnings to meet thresholds will be more likely to have positive unexpected accruals without which they would not have met the threshold.   ResultsMost studies use pretax financial accounting income as a benchmark when they are going to measure tax avoidance. That is the reason for the effect of earnings management on measures of tax avoidance. A firm might be considered to have tax avoidance if the firm manages earnings upward without paying additional tax on the managed earnings. The results demonstrate that COF ETR that has been used in prior research - the ratio of cash taxes paid to pretax operating cash flows - captures less likely tax avoidance that is unrelated to earnings management. The results imply that pretax income aggregates over a longer period such as five years do not mitigate the relation with earnings management. The results indicate that the distributions of Cash ETRs and CFO ETRs are significantly different from each other for the one-, three-, and five-year ETR measures.  Discussion and ConclusionThe differentiation of tax avoidance related to or unrelated to earnings management is important because each effect suggests different underlying actions by managers and might be viewed differently by researchers and policymakers. In this study, the recognition of the numerator and denominator effect is a practical concept. For example, when a researcher can find tax avoidance by using Cash ETR and can not find tax avoidance using CFO ETR, it does not mean there is no tax avoidance. this can imply the company avoids tax with increasing income that reflects the denominator effect. If the researcher can find tax avoidance using CFO ETR too, it means the company avoids tax with decreasing tax payments that reflect the numerator effect. Using CFO ETR in empirical studies can help interpret inconsistent results in some studies of literature and avoid attributing earnings management results to tax avoidance. according to this, when CFO ETR is unable to capture tax avoidance, not necessarily imply there is no tax avoidance but it contains the data about the type of tax avoidance. We suggest to future researchers separate types of tax avoidance in their models in vague areas of literature. This can avoid attributing earnings management results to tax avoidance.