Can Auditors be Independent? Experimental Evidence on the Effects of Client Type

Abstract

Recent regulatory initiatives stress that an independent oversight board, rather than the management board, should assume the role of auditors' client. In an experiment, we test whether the type of client affects auditors' independence. Unique features of the German institutional setting enable us to realistically vary the type of auditors' client as our treatment variable: we portray the client either as the management preferring aggressive accounting or the oversight board preferring conservative accounting. We measure auditors' perceived client retention incentives and accountability pressure in a post-experiment questionnaire to capture potential threats to independence. We find that the type of auditors' client affects auditors' behaviour contingent on the degree of the perceived threats to independence. Our findings imply that both client retention incentives and accountability pressure represent distinctive threats to auditors' independence and that the effectiveness of an oversight board in enhancing auditors' independence depends on the underlying threat.

Acknowledgments

We are grateful to Ann Vanstraelen, the associate editor, and two anonymous reviewers for helpful comments. The comments of Jannis Bischof, Thomas Carrington, Holger Daske, Aasmund Eilifsen (discussant), Michael Favere-Marchesi (discussant), Don Finn (discussant), Anna Gold, Frank Hartmann, Christopher Humphrey, Kathryn Kadous (discussant), Hansrudi Lenz, Jason MacGregor, Ernst Maug, Martin Richter, Steven Salterio, Ulrike Stefani (discussant), Jeroen Suijs, Hun-Tong Tan, and workshop participants at the Erasmus University of Rotterdam and the University of Mannheim, the 2007 Symposium of the European Auditing Research Network in Aarhus, Denmark, the 2008 Auditing Section Midyear Conference in Austin, TX, the 2008 Annual Congress of the European Accounting Association in Rotterdam, the Netherlands, the 2008 Annual Meeting of the German Academic Association for Business Research in Berlin, Germany, the UF 2009 International Conference on Assurance and Governance in Gainesville, FL, the 2009 Annual Conference of the Accounting Section of the German Academic Association for Business Research (Best Paper Award) are greatly appreciated. Financial support from the German Research Foundation (DFG Grant SFB504) is acknowledged. Finally, we thank two of the Big Four audit firms for supporting us and we thank the participants for their time in completing the research study.

Notes

We combine both the audit committees and supervisory boards under the umbrella term ‘oversight board’.

German supervisory board members face mainly internal liability threats with the company itself being the claimant. The overall level of their liability threat is comparable to those of UK non-executive members (Lederer; 2006). Given the internal nature of the liability threat and the lack of empirical studies on the subject, we argue that there is no compelling reason why liability threats should not play a role in the context of German private companies.

In total, 84 subjects participated. We excluded 12 participants who indicated IT-auditing, tax consulting or consulting as their main field. Our main results hold, nevertheless, when we include all 84 subjects in our analysis. Among the 72 subjects that indicated auditing as their main field, 52 were male and 15 were female auditors; five participants did not indicate their gender.

This applies to stock corporations (Aktiengesellschaften). It also applies to private limited corporations (Gesellschaften mit beschränkter Haftung) with more than 500 employees who are required to form a supervisory board under section 1, paragraph 3 of the One-Third Codetermination Act (Gesetz über die Drittelbeteiligung der Arbeitnehmer im Aufsichtsrat, Drittbeteiligungsgesetz).

Section 52 Gesetz betreffend die Gesellschaften mit beschränkter Haftung, GmbHG (Private Limited Liability Company Act). Furthermore, private limited companies are free to pass corporate bylaws that can even give management the power to directly choose the auditor.

Sections 267, 316 Handelsgesetzbuch (German Commercial Code) requires a company to be audited when two of the following three requirements are fulfilled in two consecutive financial years: more than 50 employees based on an annual average; a balance sheet total of more than €4,840,000; sales of above €9,680,000.

An alternative explanation for the higher level of doubt expressed by auditors employed by management could be that the description of the management preferences as favouring aggressive accounting influences auditors towards high fraud risk. If this were the case, then auditors would be expected to be especially doubtful about evidence provided by management. We do not find support for this explanation. Auditors hired by management show the same level of scepticism for evidence provided by management as for evidence provided by other sources. This behaviour is indicated by a non-significant interaction effect between client type and a dummy variable for evidence type (one indicates that a piece of evidence is provided by management; and zero otherwise) on evidence evaluation (untabulated). Furthermore, auditors do not differ by client type in their written justifications of their decisions. Auditors hired by the management state even slightly more often that they have relied on evidence provided by management compared to auditors hired by the oversight board (full sample: 39 compared with 32 times; subsample of managers/partners: 14 compared with 11 times).

Additional information

Notes on contributors

Christopher Koch

Paper accepted by Salvador Carmona