The Nature and Disclosure of Fees Paid to Auditors
An Analysis Before and After the Sarbanes-Oxley Act

By Ariel Markelevich, Charles A. Barragato, and Rani Hoitash

E-mail Story Print Story

In the wake of accounting firms’ transformation, regulators became increasingly concerned about the interplay between auditor independence and the provision of nonaudit services (NAS) to audit clients. In his highly publicized testimony before the U.S. Senate on September 28, 2000, then–SEC chairman Arthur Levitt expressed his concern that “as auditing becomes an ever-smaller portion of a firm’s business with an audit client, it becomes harder to assume that the auditor will challenge management when he or she should, if to do so might jeopardize a lucrative consulting contract for the auditor’s firm.” This view, coupled with Enron’s failure, WorldCom’s malfeasance, and the collapse of Arthur Andersen, led to the eventual passage of the Sarbanes-Oxley Act of 2002 (SOA).

The analysis that follows focuses on the market for audit and nonaudit services by examining fees paid to auditors during the period 2000 to 2003. This timeframe is of particular interest because this period saw sweeping changes in auditors’ business, regulatory, and professional environment.

Regulatory Background

In recent years, the SEC and Congress have promulgated a variety of rules that are grounded in the notion that auditor independence is vital to the production of high-quality audits and that fees paid to auditors for both audit and nonaudit services may impair such independence. In November 2000, the SEC issued a directive requiring public companies to disclose audit and audit-related fees paid to their outside auditors. These disclosure rules became effective for proxy statements filed after February 5, 2001 (SEC Final Rule S7-13-00). Following SOA, the SEC expanded (and in some instances redefined) these disclosure requirements, and now requires that fees paid to auditors be broken down into the following categories: 1) audit fees; 2) audit-related fees; 3) tax fees; and 4) all other fees. One of the more significant changes under the expanded guidelines is a change in how audit fees are defined. The initial rule adopted by the SEC (for proxies filed in 2000) required that companies disclose fees paid for audits and quarterly reviews in the “audit fees” category. The expanded rule requires companies to include any fees for services performed to fulfill the accountant’s responsibility under GAAS. Additionally, audit firms are now prohibited from providing such services as financial information system implementation and design, internal auditing, and a number of other services.

Data and Results

The study comprised a sample consisting of 2,507 public companies that have disclosed audit fee information from 2000 to 2003, as reported in the Standard & Poor’s Audit Fee Database. Starting in 2003, companies were required to report fees paid to their auditors under the new disclosure rule. The new rule also mandated that companies present their fiscal 2002 fees under the new rule for comparison purposes. Consequently, the sample consists of fees reported under the old rules for 2000 and 2001, and fees reported under the new rule for 2002 and 2003. The descriptive statistics for the additional fee categories are limited to 2002 and 2003.

Analysis

Exhibit 1 presents the full sample descriptive statistics for fees paid for audit and nonaudit services during the period under study. For ease of exposition, and to mitigate the impact of extreme observations, the discussion focuses on median fees (illustrated in Figure 1).

As noted in Exhibit 1, total fees increased from $602,369 in 2000 to $683,618 in 2003, an increase of roughly 13%. In contrast to the changes in the definition of audit fees and nonaudit fees (as described above), the definition of total fees remained consistent over time. Some critics contend that large fees paid to auditors make auditors more economically dependent on their clients, possibly creating a relationship in which the auditor becomes reluctant to make appropriate inquiries during the audit for fear of losing highly profitable fees. Overall, there has been a slight increase in total fees from 2000 to 2003. Such a modest change in total fees over the sample period makes it difficult to make reasonable inferences concerning the assertion that auditors can become economically dependent upon clients, or how SOA may have affected this.

Audit fees increased almost 80%, from $239,000 in 2000 to $430,000 in 2003. This increase is substantial and is likely attributable to a number of factors, including: 1) increased risk of litigation; 2) changes in the scope and complexity of audit engagements; 3) transition from the Big Five to the Big Four marketplace (the demise of Arthur Andersen); 4) reactions to new regulatory restrictions forbidding auditors from rendering certain nonaudit services; and 5) the changing definition of the audit fees category (the revised SEC disclosure requirement). Additional analyses address the potential impact of the new definition of audit fees; because fees paid by companies in 2002 were reported under both the old and new rules, there is a unique opportunity to study the effects of this reclassification. As a result of the reclassification, median audit fees increased by about 10%, while nonaudit fees decreased by a similar amount. This suggests that any change above 10% results from actual variations in the services provided by auditors (or the fees charged), rather than from the change in classification.

The median fees paid for nonaudit services declined from $312,741 in 2000 to $211,200 in 2003, a decline of 32%. In 2003, nonaudit services accounted for roughly 31% of total fees, compared to almost 52% in 2000. If the relation between auditor independence and the provision of nonaudit services to audit clients is problematic, as the SEC and Congress have argued, then the reduction in the ratio of nonaudit services to total fees should help limit auditor-independence violations.

Tax fees represent the largest category of the nonaudit fee composite. Median tax fees declined by 8%, from $102,000 in 2002 to $93,448 in 2003. Median audit-related fees increased 26% over the same period, from $40,140 in 2002 to $50,500 in 2003.

Analysis by Audit Firm Size

Category 1 includes only the Big Five (Arthur Andersen, Deloitte, Ernst & Young, KPMG, and PwC), category 2 consists of the two largest second-tier firms (BDO Seidman and Grant Thornton), and category 3 contains all other auditing firms.

Descriptive statistics for fees paid to the Big Five are presented in Exhibit 2A. Total fees behavior is similar to that in Exhibit 1 for the whole sample. Median total fees increased by 20%, from $678,000 to $812,000, during the same period. Median audit fees grew substantially, from $264,000 in 2000 to $503,000 in 2003, roughly 91%. Correspondingly, median nonaudit fees decreased approximately 27%, from $364,550 in 2000 to $266,348 in 2003.

Exhibit 2B presents the results for category 2, second-tier firms. The median total fees increased by 33%, from $218,713 in 2000 to $291,450 in 2003. Median audit fees increased from $138,950 in 2000 to $197,900 in 2003, roughly 42%. Median nonaudit fees decreased by 19%, from $80,674 in 2000 to $65,350 in 2003.

Results for the third group, small audit firms, are presented in Exhibit 2C. Median total fees increased from $298,207 in 2000 to $308,919 in 2003, an increase of 4%. Consistent with the trend in the previous two categories, median audit fees increased by roughly 40%, from $149,875 in 2000 to $210,488 in 2003, while median nonaudit fees decreased by 17%, from $114,000 in 2000 to $95,138 in 2003.

Changes in Market Share

Exhibit 3 presents market share data by audit firm category in terms of the total fees received by audit firms and the total number of clients they serve.

Although there is little change in the percentage of total fees received by each of the three auditor groups from 2000 to 2003, it is interesting that the Big Five collected just under 92% of the fees in both years. With respect to changes in the aggregated fees between 2000 and 2003, second-tier firms increased collections from their audit clients by just over 44%. Conversely, the Big Five firms and small firms experienced a decline in total collections of roughly 13%.

The Big Five lost 86 clients (a 4% decrease) to the second-tier (36 clients, a 58% increase) and small firm groups (50 clients, a 15% increase). These results suggest that although both the second-tier and small firms gained clients, the majority of Arthur Andersen’s clients were retained by the remaining Big Four.

Implications

Audit fees increased substantially between 2000 and 2003, with the Big Five experiencing the greatest percentage increase, accompanied by a large decline in nonaudit fees for firms of all sizes. These results are not caused by the changes in the definitions of audit and nonaudit fee classifications but rather by changes in the services provided by auditors, or the fees charged for those services. The net effect of these changes may appear relatively modest given that several of the Big Five spun off their consulting practices during or just prior to the period in question and that SOA now limits the types of consulting services that can be offered to audit clients.

Small audit firms appear to have been more negatively affected during the study period, as evidenced by their relatively flat total fees from 2000 to 2003, as compared to a 20% and 33% increase for Big Five and second-tier firms, respectively. Using total fees as a barometer, second-tier firms experienced a substantial increase in market share from 2000 to 2003, with both the Big Five and small firms giving up ground.

Although the full regulatory impact of SOA remains to be seen, to the extent that Congress and the SEC are correct that the relation between auditor independence and the provision of nonaudit services to audit clients is problematic, then the expanded fee disclosures and restrictions on consulting services should reduce auditor-independence violations. On the other hand, if auditor-independence violations stem more from auditors’ dependency on the total fees received from audit clients, then the relatively small reduction in total fees documented from 2000 to 2003 may require refocusing on other aspects of the auditor-independence issue.