Nonaudit Services and Independence

---

As federal and state legislators and regulators study again perceptions of auditor independence, conflict of interests, and nonaudit services, this 1998 exchange of letters to the editor of The CPA Journal between Harvey Pitt and Douglas Carmichael reflects concerns at the core of these issues since the Cohen Commission Report in the 1970s. Although impossible to resolve once and for all, a profession has a duty to manage such conflicts responsibly in response to changing conditions and public expectations.

-- Bob Colson

---

From The CPA Journal, June 1998

Our firm prepared the recent White Paper on auditor independence on behalf of the AICPA. We therefore were heartened by Douglas Carmichael's positive reaction in the March 1998 CPA Journal to the White Paper's suggested approach--the adoption of core principles of independence by the Independence Standards Board coupled with the development of individual CPA firm codes containing safeguards to assure adherence to those core principles. As Mr. Carmichael stated, "[a]n approach that would put emphasis on achieving the policy objectives of independence in fact rather than mechanical compliance with overly detailed rules is a real breakthrough."

Mr. Carmichael does appear to disagree, though, with one point in the White Paper--our conclusion that there is no empirical basis for the proposition that the provision of nonaudit services for audit clients leads to audit failure. As explained in the White Paper, numerous studies, including a recent and comprehensive survey of the literature by the General Accounting Office, confirm this understanding.

In particular, Mr. Carmichael takes issue with our reliance on one of those studies, the 1978 report of the Cohen Commission. However, the Cohen Commission's research did not find any instances in which an audit failure had a demonstrated connection to the provision of nonaudit services. As Mr. Carmichael notes in this regard, the Cohen Commission considered three instances in which nonaudit services brought to light a material misstatement or other information relevant to a completed audit, the information was called to the attention of the auditors, and the auditors failed to disclose promptly what had been discovered. That is an entirely different issue from whether nonaudit services compromise audit quality. Indeed, to the contrary, based upon these cases, the Cohen Commission found that management advisory services could increase audit effectiveness. There was only one instance raised with the Cohen Commission in which it had even been alleged that nonaudit services affected independence in a manner that reduced audit quality, the 1968 Westec case, where an independent auditor's provision of accounting advice in combination with involvement in the company's mergers and acquisitions program was claimed to have impaired the auditor's ability to independently audit the resulting transactions.

As discussed in the White Paper, the issue of nonaudit services should also be evaluated in light of the accounting profession's actual experience over the last two decades. In 1986, for example, the Public Oversight Board concluded that there was no known "instance in which it can be demonstrated that the provision of [management advisory services] to an audit client interfered with independence in performing the audit function." A recent and comprehensive study conducted for the ISB by Paul R. Brown and Jeanne A. Calderon of NYU similarly found that since 1980 there have been "a surprisingly small number of cases where auditor independence was a major issue," and, moreover, identified only one reported case (in 1986) in which an auditor's provision of nonaudit services was even raised as an issue. We would welcome additional empirical research in this area because, as the White Paper argued, rational regulation of auditor independence must be predicated on empirical data rather than instinct.

Former Commissioner Steven Wallman maintained, while at the SEC, that the "unstated preference among many that auditors should do little in the way of nonaudit work for audit clients" underlying the current independence regime "fails in a number of respects and may well be contrary to the public interest." The White Paper's ultimate conclusion on this issue, that nonaudit services are essentially unrelated to any risk of audit failure, is consistent with Commissioner Wallman's suggestion that independence policy should focus on empirical results, and finds real-world support in the fact that neither insurance brokers nor insurance companies view the performance of nonaudit services for audit clients as increasing the liability risk of accounting firms who perform such services. We believe, as did the Cohen Commission, that performance of nonaudit services can improve the audit function, and, further, that the ability of firms to draw upon broader sources of revenue generally should make firms less dependent on any single client.

--------------------------------------------------------------------------------

A DIFFERENT PERSPECTIVE

The response of Harvey Pitt and David Birenbaum, similar to the White Paper itself, misses the point about the lack of empirical evidence as it relates to nonaudit services, independence, and audit failure.

The Cohen Commission studied the issue closely and evaluated the evidence that existed at the time. I served as the director of research for the Cohen Commission; attended all its meetings; participated in all its deliberations; along with the Acting Chairman, Lee Seidler, wrote the report; and personally drafted the chapter related to audit independence.

Mssrs. Pitt and Birenbaum state that nonaudit services were alleged to have affected independence in only one instance. That is not my understanding. One of the three cases the Cohen Comission considered involving after the fact nonaudit services was Yale Express. In that instance, the Cohen Commission believed there was a lack of independence. The failure of the audit firm to promptly inform the investing public of a misstatement in prior audited financial statements was a violation of the auditor's public trust and clearly not in the interests of the investing public. The Cohen Commission did conclude, however, that the lack of independence could not have caused an audit failure with respect to financial statements issued before the nonaudit services were provided. It also concluded in the Westec case that providing nonaudit services had impaired independence.

The fact that the Cohen Commission and others have concluded that the evidential record does not warrant blanket prohibitions concerning nonaudit services for audit clients should not be translated into a sweeping value judgment that the issue should be closed to further study, the feeling I got from reading the White Paper. I am pleased to note that their letter explicitly states the desirability for further research in this area.

A plaintiff in a civil action has to prove liability, causation, and damages. A civil plaintiff in an action against an auditor must, in this process, prove that the audited financial statements were materially misstated and that the auditor would have detected the misstatement by performing the audit with due professional care. Unless there have been clear violations of the accounting profession's own ethics rules or interpretations on independence, the plaintiff's efforts are usually directed to establishing the existence of material misstatement and inadequate audit performance, and auditor's independence is a side issue at best. In the case of a large CPA firm, nonaudit services are often provided by a separate department or even an office that does not participate in the audit engagement. There is normally no documentation in the audit workpapers of the nature or extent of nonaudit services. In these circumstances, the plaintiff might not even learn that nonaudit services had been provided to the audit client.

Thus, I urge the Independence Standards Board to carefully examine all available evidence firsthand in evaluating what activities could impair audit independence.*