Client Importance, Non-Audit Services, and Abnormal Accruals
Posted: 20 Jun 2003
Several recent studies examine the association between fees for audit and non-audit services, and auditor independence impairment. However, auditors' incentives to compromise their independence are proxied in different studies by different measures. The economic theory of auditor independence (Watts and Zimmerman 1981; DeAngelo 1981) suggests that auditors' incentives to compromise their independence are related to client importance, the ratio of quasi rents specific to the client divided by all other quasi rents. Following previous studies we use the ratio of a client's total fees to audit firm's total revenues as our first proxy for client importance. Because it has been argued that non-audit services have a higher margin than do audit services and therefore generate more quasi rents (revenues minus marginal costs), we also use a second measure of client importance, the ratio of non-audit fees from a client to the audit firm's total US revenues. If auditors care more about revenues of their practice office rather than the whole firm, then client importance should be assessed at the local office level. So we also use two additional proxies for client importance, client total fees and non-audit fees divided by a surrogate measure of practice-office revenues. To develop this surrogate measure we assume that clients are audited through the audit firm's practice office closest in distance to the client's headquarters, and allocate audit firm revenues to practice offices in proportion to the sum of log(sales) of clients of each office. Consistent with prior research we use abnormal accruals estimated using the (modified) Jones model as our measure of earnings, and therefore audit, quality.
After controlling for industry effects and determinants of abnormal accruals based on previous studies, we do not find a significant cross-sectional association between the absolute value of abnormal accruals and any of the client importance ratios in our sample of 1,871 firms for the year 2000. Following suggestions in Greenwald (1975) we show that the lack of significance is unlikely to have resulted from low power of our tests.
Our model development also suggests that independence impairment could be higher in sub-samples of clients that have greater opportunities or incentives to manage earnings, or those with weaker corporate governance structures. In addition, auditor expertise may create client dependence on auditor, and thereby strengthen auditors' independence. We measure earnings management opportunities using business and geographical diversification. We use leverage, market-to-book ratios, sales growth, and the number of shares and options owned by executives as a percentage of shares outstanding as proxies for client incentives. The strength of corporate governance is proxied by whether the chairman and CEO positions are separate, the percentage of non-employee directors, and percentage of shares owned by 5 percent block-holders. Auditor expertise is proxied by whether they are the industry specialist. We fail to find a significant association abnormal accruals and client importance measures for any of the sample partitions.
JEL Classification: M40, M49, M41, M43
Suggested Citation: Suggested Citation