Abstract:
This study investigates two capital markets effects of auditor provided tax services (APTS), a particular form of auditor provided non-audit services (APNAS). Firstly, this study examines the influence of APTS on income shifting by United States of America (U.S.) multinational companies and, secondly, this study examines the impact of APTS on default risk of all U.S. companies.
There are two competing hypotheses on the impact of APNAS on the quality of the work of auditors and the empirical evidence is mixed. One strand of literature suggests that APNAS provide knowledge spillover effects and thus improve the quality of the work of the auditor. The other strand of literature suggests that APNAS impair the independence of the auditor and therefore lead to a decrease in the quality of the audit. APNAS may thus increase or decrease the value of audit as a governance mechanism. The U.S. Securities Exchange Commission (SEC) has banned several previously allowed APNAS such as bookkeeping, financial information systems design and implementation, appraisal and valuation, and internal audit. However, the SEC continues to permit auditors to provide tax services.
This study extends the literature on APNAS by examining the effect of APTS on income shifting by multinational companies and on default risk. Using a sample of 10,248 firm-year observations on U.S. multinationals over the period 2002 – 2015 and the income shifting measurement model developed by Dyreng and Markle (2016), this study finds that APTS reduce outbound income shifting, which is consistent with knowledge spillover rather than impairment of independence. The result holds after addressing potential endogeneity concern and is robust to excluding observations from the financial crisis periods. Furthermore, the result holds after including firm-specific characteristics as influences on the income shifting parameters.
Using a sample of 21,364 firm-year observations on U.S. firms over the period 2003 – 2016, this study finds that APTS have a positive relationship with default risk, consistent with impaired independence of the auditor. The result holds after addressing potential endogeneity concern and is robust to excluding the global financial crisis period. The effects of APTS on income shifting and default risk are therefore opposite in direction. However, the positive relationship between APTS and default risk is weaker for firms with high institutional holdings and a strong information environment, indicating that stronger corporate governance mitigates the impact of APTS on default risk. Furthermore, this study finds that the channel for the effect of APTS on default risk appears to be earnings quality. That is, APTS lower audit quality, thereby lowering earnings quality and increasing default risk. Given the cost of default, this is an important finding.
Thus, taking the results on income shifting and default risk in combination, the question of the SEC continuing to permit auditors to provide tax services is left open to question.
