Whose Bread I Eat, His Song I Sing? Not If I Am an Auditor

Should the gatekeepers of a firm be hired and paid by the very firm they are to watch over? A new study evaluates the common belief that there is a fundamental conflict of interest in the hiring and firing of auditors by clients.

Across the world, it is common for auditors to be hired and paid by clients. The same is true for credit rating agencies. Although pervasive, this practice has fuelled a debate about the likely conflict of interest it creates. Corporate accounting scandals such as Enron and, more recently, WorldCom, have further intensified this discussion. As a consequence, to ensure auditor independence, there has been rising pressure to formulate policy frameworks that would remove control over hiring and firing of auditors from the client and place it with a regulator or stock exchange instead. Yet, there are also those who believe that the auditor’s concern for reputation and professional identity, coupled with the fear of legal liability, would serve as deterrents against unethical practices.

Despite the divergence of opinion, the practice of auditors being selected by clients is ubiquitous, resulting in little or no variation to be found in auditor appointment methods across the board. This further adds to the complexity of studying the practice in actual audit settings. As a result, past research on this topic is sparse except for experimental evidence from a laboratory study conducted by Mayhew and Pike in 2004, which found audit quality to be higher when auditors were hired by investors instead of managers, and a field experiment conducted by Duflo, Greenstone, Pande and Ryan on environmental auditors in India in 2013, which found that experimenter-assigned auditors reported more truthfully than client-selected auditors. However, given the limitations of these settings, should the findings be generalised to the auditing profession as a whole?

A policy experiment in Korea presented Professor Sanjay Kallapur and his co-authors with an opportunity to evaluate this question in a real-world financial audit setting. In the late 1980s, audit quality was perceived to be low in Korea as clients would hire and fire their auditors at will. With stiff competition among auditors, there was further inducement to compromise audit quality by bending to the client’s wishes. In a bid to improve auditor independence and thereby audit quality, in 1991, the regulator, Financial Supervisory Service (FSS), introduced an auditor designation system for firms. This meant that the FSS would designate or choose the auditor for certain firms. The FSS had the authority to designate auditors for client firms in a particular year and then decide one year at a time whether or not to continue with the designation. Thus, for those years, clients were not allowed to choose their own auditors. As a result, a client firm had no way to influence or determine who its designated auditor would be. At the same time, it also ensured that auditors were not obligated to client managers for having appointed them.

The audit statements for these firms were then tested by the authors for their quality by using two proxies, namely, discretionary accruals and restatements. Since cash is hard to manipulate, earnings management is generally reflected in accruals. For instance, firms can artificially inflate sales by extending more credit, or increase earnings by making a smaller provision for bad debts. The second audit quality measure used was restatements, that is, whether the firm used subsequent restatements of financial reporting due to misstatements.

The hiring and firing of auditors by the very clients they audit is commonplace, although it creates a fundamental conflict of interest. This begs the question — does audit quality improve when auditor choice is independent of the client? Using a real-world financial audit setting in Korea, Professor Sanjay Kallapur and his co-authors seek evidence on this proposition.

For both proxies of audit quality, the researchers failed to find evidence that audit quality was higher under the designation system. To the contrary, during the designated period, firms had both higher discretionary accruals and a higher probability of having their financial statements subsequently restated. These findings indicate that the auditor designated period was associated with lower audit quality as opposed to higher audit quality.

Thus, although there is a strongly held view that there is a conflict of interest in the hiring of auditors by their clients, the evidence from the study opposes it. The results failed to support the proposition that audit quality would improve if auditor choice were removed from the client’s hands. Further, while past research findings had been experimental in nature, these findings are from an actual audit setting and therefore most closely related to the question of the effect of client hiring and firing on audit quality.

The study suggests that policy change recommendations to remove managers’ control over hiring and firing of auditors and giving it to regulators might be premature. Therefore, these findings should make both researchers and policy makers more cautious about their beliefs on the issue.

ABOUT THE RESEARCHERS

Gil S. Bae, Professor in the Accounting area at the Korea University Business School.

Sanjay Kallapur, Professor in the Accounting area, Chair – Faculty Performance Management, Director of the Fellow Programme in Management (FPM) and Director of the Initiative for Emerging Market Studies (IEMS), at the Indian School of Business (ISB).

Joon Hwa Rho, Professor in the Accounting area at the College of Business, Chungnam National University, Korea.

ABOUT THE RESEARCH

Bae, Gil S., Kallapur, Sanjay, and Rho, Joon Hwa. Does Hiring and Firing by Clients Compromise Auditors? Evidence from Regulator- Selected Auditors in Korea. Working paper.

ABOUT THE WRITER

Unnati Ved, Academic Associate and a freelance writer with the Centre for Learning and Management Practice, at the Indian School of Business (ISB).