ARE U.S. GOVERNMENT ENFORCEMENT ACTIONS EFFECTIVE AT IMPROVING BUSINESS ETHICS?

Abstract

This dissertation examines the impact and determinants of government enforcement action related to compliance and corruption. Study I assesses whether Health Care Compliance (HCC) related government enforcement actions are effective at improving firms’ corporate social responsibility (CSR) scores. In a study of 37 enforcement actions, I find that the corporate social responsibility (CSR) score significantly increases during the three years after the enforcement settlement, compared to the period before the enforcement action. In particular, I find that the Diversity and Community CSR sub-scores improve following the enforcement settlement. However, there is little evidence that firms with poorer CSR ratings are more likely subject to enforcement action. Study II expands the reach of enforcement actions beyond the healthcare industry to all firms subject to Foreign Corrupt Practice Act (FCPA) enforcement actions. Based on 88 cases of such enforcement actions, I document increases of the overall CSR scores as well as several sub-scores post enforcement settlement, which is consistent with the results in Study I. I also find that firms start to improve their CSR ratings as soon as the identification of the misconduct occurs, which is on average four years prior to the settlement of the enforcement action. Furthermore, the improvement in CSR rating post misconduct year tends to be greater and more significant among firms that face larger monetary sanctions. This implies that the materiality of enforcement monetary penalties plays a significant role in shaping a firm’s timely response to government investigation of misconduct. Although I find that CSR ratings in Employee Relations and Product are significantly negatively related to the chance of being subject to enforcement action investigation, there is little evidence that firms with poor CSR ratings are more likely to subject to enforcement actions. This finding is consistent with the results from Study I.