ABSTRACT
Tax avoidance holds immense importance due to its substantial implications for government revenues and the fair allocation of resources. Consequently, understanding the factors that shape tax avoidance is critically important. Exploiting a cutting-edge measure of corporate integrity derived from state-of-the-art machine learning algorithms and textual analysis, we explore the effect of corporate integrity on tax avoidance. Our text-based measure is based on a textual analysis of earnings conference call transcripts. Our findings show that companies with greater corporate integrity are significantly less involved in tax avoidance. Further analysis corroborates the results, i.e., propensity score matching, entropy balancing, and an instrumental variable analysis. Our findings are especially noteworthy as they demonstrate that corporate culture, although intangible in nature, exerts a substantial influence on corporate behavior.
