Cornell Law Library Scholarship@Cornell Law: A Digital Repository Cornell Law Faculty Publications Faculty Scholarship 2004 Was Arthur Andersen Different? An Empirical Examination of Major Accounting Firm Audits of Large Clients Theodore Eisenberg Cornell Law School, [email protected] Jonathan R. Macey Cornell Law School, [email protected] Follow this and additional works at:http://scholarship.law.cornell.edu/facpub Part of theAccounting Law Commons,Corporation and Enterprise Law Commons, and the Securities Law Commons Recommended Citation Eisenberg, Theodore and Macey, Jonathan R., "Was Arthur Andersen Different? An Empirical Examination of Major Accounting Firm Audits of Large Clients" (2004).Cornell Law Faculty Publications.Paper 968. http://scholarship.law.cornell.edu/facpub/968 This Article is brought to you for free and open access by the Faculty Scholarship at Scholarship@Cornell Law: A Digital Repository. It has been accepted for inclusion in Cornell Law Faculty Publications by an authorized administrator of Scholarship@Cornell Law: A Digital Repository. For more information, please [email protected]. Journal of Empirical Legal Studies Volume 1, Issue 2, 263-300, July 2004 Was Arthur Andersen Different? An Empirical Examination of Major Accounting Firm Audits of Large Clients Theodore Eisenberg andJ onathan R Macey* Enron and other corporate financial scandals focused attention on the accounting industry in general and on Arthur Andersen in par- ticular. Part of the policy response to Enron, the criminal prosecu- tion of Andersen eliminated one of the few major audit firms capable of auditing many large public corporations. This article explores whether Andersen's performance, as measured by fre- quency of financial restatements, measurably differed from that of other large auditors. Financial restatements trigger significant neg- ative market reactions and their frequency can be viewed as a measure of accounting performance. We analyze the financial restatement activity of approximately 1,000 large public firms from 1997 through 2001. After controlling for client size, region, time, and industry, we find no evidence that Andersen's performance sig- nificantly differed from that of other large accounting firms. This article is about accounting firm performance and its implications for the state of product differentiation and corporate governance in the accounting industry. We investigate whether a serious sign of accounting error, the substantive restatement of earnings by large publicly held com- *Eisenberg is Henry Allen Mark Professor of Law, Cornell Law School and Visiting Professor of Law, Harvard Law School; Macey is Professor of Law, Yale Law School. We thank Ann M. Eisenberg, Thomas P. Eisenberg, Nick Germain, Erica Miller, and Kathy Wimsatt for data-entry assistance. We are extremely grateful to Mark W. Nelson for providing comments and advice and for his gracious welcome of our incursion onto his intellectual turf. Ian Ayres, Alvin Klevorick, Roberta Romano, and Shyam Sunder also provided extremely useful comments and suggestions. Errors remain our own. 263 HeinOnline -- 1 J. Empirical Legal Stud. 263 2004 264 Major Accounting Firm Audits of Large Clients panies, which are discovered by the issuer, the Securities and Exchange Com- mission, or third parties, occurs with similar frequency across the major accounting firms, or whether restatements tend to be isolated in particular miscreant firms, particularly the late Arthur Andersen (Andersen). Some of the governmental responses and initiatives to the wave of Enron-style corporate collapses reflect the view that Andersen, Enron's auditor, was an outlier among accounting firms. Other government actions, however, reflect the view that the major accounting firms are more or less the same. For example, Andersen's internal corporate governance, moni- toring, and control systems may have been weak relative to those of the other large accounting firms, and this lack of internal controls may have made Andersen particularly vulnerable to "capture" by unscrupulous clients. Alter- natively, it may be the case that all large accounting firms are similar in the sense that all are equally susceptible to internal corporate governance prob- lems, and all have, more or less, the same quality of internal controls. If so, then all firms should be equally susceptible to the possibility that client firms will capture the teams of auditors assigned to prepare financial statements and cause these accountants to acquiesce in inappropriately aggressive accounting treatment, or even to pro-actively participate in the design of materially misleading accounting statements. The stakes of accurately describing Andersen's performance are high. Remarkably few large accounting firms audit many large public corpora- tions. Removing a major firm from such a thin market has implications for the public securities markets as well as for the accounting industry. Inde- pendent of Andersen's particular performance, the stakes of understanding the pattern of financial restatements are also high. Restatements have been shown to trigger substantial drops in restating firms' market value. Examin- ing restatements, which reflect on the quality of a major product of account- ing firms-their public company audits-may also foster insights about the relation between large companies and their auditors and about the internal governance of auditors. Analysis of about 1,000 large public firms from 1997-2001 yields no evidence that accounting profession problems that lead to restatements were unique to Andersen. Andersen's clients did not restate their financial results at a significantly different rate than the other major accounting firms during this period. During the period of our study, private plaintiffs and govern- ment regulators began to focus more intensively on accounting irregulari- ties and the percentage of public companies restating their financial results HeinOnline -- 1 J. Empirical Legal Stud. 264 2004 Eisenberg and Macey 265 increased dramatically. Interestingly, however, we do not find a significant rise in Andersen's share of the increased number of restatements. Rather, the distribution of restatements among the largest accounting firms remained roughly the same. Thus, by the restatement-rate measure, the vilified and now-defunct Andersen was not objectively different from the other major accounting firms. If financial restatement rates are a key measure of audit firm per- formance, or if other measures of performance yield similar results, one may question the government's decision to end Andersen's existence by bring- ing criminal charges against it. Our findings also call into question the value of market competition in generating successful product differentiation. Our results do suggest that the accounting reforms implemented by the Sarbanes- Oxley Act of 2002, which apply to all accounting firms, can be defended on the ground that accounting profession problems are industrywide and not linked to any particular firm. To the extent that financial restatement rates capture an important component of accounting firm quality, we find little evidence that the top accounting firms compete with one another on the basis of quality. Part I of this article discusses the law and economics of accounting and provides the relevant background to our analysis. Part II describes the data and reports the empirical results. Part III discusses the results. A conclusion follows. I. THE LAW AND ECONOMICS OF ACCOUNTING A. The Paradigm The law and economics literature has a simple and elegant explanation for the importance of independent accounting firms.1 Generating information about financial results is important for internal as well as external reasons. 1The classic article is Ross L. Watts & Jerold L. Zimmerman, Agency Problems, Auditing, and the Theory of the Firm: Some Evidence, 26J.L. & Econ. 613 (1983); see also Sung K. Choi & Debra C. Jeter, The Effects of Qualified Audit Opinions on Earnings Response Coefficients, 15J. Acct. & Econ. 229 (1992); Rick Antle, Auditor Independence, 22J. Acct. Res. 1 (1984); George J. Benston, Accountant's Integrity and Financial Reporting, Financial Executive 10 (Aug. 1975). HeinOnline -- 1 J. Empirical Legal Stud. 265 2004 266 Major Accounting Firm Audits of Large Clients Management needs to know how the firm is performing, and therefore needs accurate, reliable information on financial performance. However, all the information management needs can be (and is) generated by firms inter- nally. Demand for the services of independent outsiders to audit this infor- mation exists because firms need a credible signal that they can send to outside investors that the financial results being reported are, in fact, accurate. Thus, outside auditors do not perform any services for a company that the company does not already perform for itself. The role of the auditor is not to prepare financial reports for clients (that is the role of the account- ant). Rather, the auditor's role is to provide a reliable verification of the company's financial reports. Thus, auditors' reputations are central to the standard economic theory of auditing. Only auditors with reputations for honesty and integrity are valuable to audit-clients. The idea is that, absent a reputation for honesty and integrity, the auditor's verification function loses its value. In theory, then, auditors invest heavily in creating and maintaining their reputations for performing honest, high-quality audits. High-quality audits by inde- pendent auditors who have good reputations are assured. The quality assur- ance is derived from the fact that performing poor-quality audits diminishes the value of the audit firm's investment in reputation.2 The pre-Enron law and economics view of the accounting industry pre- dicted that accounting firms compete in a "race-to-the-top" that provides them with incentives to strive to produce high-quality audits. There was a time that the audit function was carried out in a market envi- ronment that induced high quality financial reporting. In that era, accounting firms were willing to put their seal of approval on the financial records of a client company only if the company agreed to conform to the high standards imposed by the accounting profession. Investors trusted accountants because investors knew that any accounting firm that was sloppy or corrupt could not stay in busi- 'Brian W. Mayhew, Auditor Reputation Building, 39J. Acct. Res. 599 (2001); Brian W. Mayhew, Jeffrey W. Schatzberg & Galen R. Sevcik, The Effect of Accounting Uncertainty and Auditor Reputation on Auditor Objectivity, Auditing: A Journal of Practice & Theory 49 (Sept. 2001); Norman Macintosh, Teri Shearer, Daniel B. Thornton & Michael A. Welker, Financial Accounting as Simulacrum and Hyperreality: Perspectives on Income and Capital, Accounting, Organizations and Society 13 (Jan. 2000); Ronald R. King, Reputation Formation for Reliable Reporting: An Experimental Investigation, 71 Acct. Rev. 375 (1996); Watts & Zimmerman, supra note 1; Antle, supra note 1; George Benston, The Value of the SEC's Accounting Disclosure Requirements, 44 Acct. Rev. 515 (1969). HeinOnline -- 1 J. Empirical Legal Stud. 266 2004 Eisenberg and Macey 267 ness for long. Auditors had significant incentives to "do superior work" because "auditors with strong reputations could command a fee premium, and high fees signaled quality in the auditing market."' High-quality audit services were further assured by the audit firms' independence from their clients, where independence is measured by the percentage of an audit firm's billings that are derived from a particular client.4 In a world in which auditors have both invested in developing high- quality reputations and in which no single client represents more than a tiny fraction of total billings, high audit quality seems assured. Under these con- ditions, any potential gain to an auditor from performing a shoddy audit, much less from participating in a client's fraud, would be vastly outweighed by the diminution in value to the auditor's reputation. In sum, even though companies can (and do) audit themselves, they can justify the expense of hiring outside auditors to enhance their financial reputation and credibility with a wide range of current and prospective claimants on their cash flows, including investors, suppliers, customers, and prospective employees. Under this reputational model, companies need independent audits to attract outside capital because it is widely believed that an auditing firm that discovers a problem would insist on a correction or, ultimately, fire the client. Being fired by an accounting firm has serious implications for the client.5 The accounting firm that dismissed the audit- client, however, would lose only that client, and even this loss could proba- bly be offset as the accounting firm might well gain new clients by virtue of the enhancement of its reputation that followed from firing the client. So, even though companies can and do impose their own financial con- trols and audit themselves, they hire outside auditors to capitalize on the 3Jonathan Macey & Hillary Sale, Observations on the Role of Commodification, Independence, and Governance in the Accounting Industry, 48 Villanova L. Rev. 1167, 1168 (2003). 4For example, Andersen was said to be independent of Enron because Andersen had 2,300 other audit-clients, and Enron accounted for only about 1 percent of Andersen's total revenue from auditing (Andersen's Enron's revenues were reported in 2001 as $100 million as com- pared to $9.34 billion in 2001 audit revenue). Id. at 1176 n.33. Of course, Andersen's inde- pendence as a firm did not extend to the partners responsible for doing the actual audit work for Enron. Id. at 1168. 5See, e.g., Martin Fackler, How Auditors Got the Nerve to Defy Big Japanese Bank, Wall St. J. Aug. 6, 2003, at Al (describing how auditors' failure to sign off on financial projections of a large Japanese bank caused a crisis that forced the bank to seek a $17 billion government bailout that put the financial institution under government control). HeinOnline -- 1 J. Empirical Legal Stud. 267 2004 268 Major Accounting Firm Audits of Large Clients audit firm's reputation. Hiring an auditor, at least in theory, allows the client company to "rent" the reputation of the accounting firm, which rents its reputation for care, honesty, and integrity to its clients. This, in a nutshell, is the economic theory of the demand for the services of independent auditors. In theory, then, accounting firms are willing to put their seal of approval on a company's financial records only if the company agrees to conform to the high standards imposed by the accounting profession. Investors trust accountants because investors know that any accounting firm that is sloppy or corrupt could not stay in business for long. The long-term loss to the reputation of an independent accounting firm that does slipshod or fraudulent work is much greater than any possible short-term gains the accounting firm might get by cutting corners. Companies that refuse to comply with the auditors' demands for transparency and simplicity in report- ing risk being dismissed by their auditors. Being fired by an accounting firm sends a negative signal to investors that often both devastates a company and leads to the dismissal of top management. Outside audits send a strong signal to investors that the company's financial house is in order. From the perspective of audit firms' clients, good audits are good investments because they reduce the cost of capital and increase shareholder wealth. Good audits also increase management's credibility among the investment community. In theory, the capital markets audit the auditors. Public accountants knew they had a lot to lose if their clients' information turned out to be false or misleading. Auditors who did a superior job would reduce the chance of their clients' issuing unreliable information and so reduce their own risk of being sued by aggrieved investors. Such suits are costly to audi- tors; even unsuccessful suits damage their valuable reputations.6 B. What Went Wrong? It is generally thought that something went very wrong in the accounting profession. Several factors may have contributed to a decline in audit quality. The shift of organizational form from the general partnership form to the limited liability partnership form reduced the threat of liability faced by 6Daniel B. Thornton, Financial Reporting Quality: Implications of Accounting Research, Sub- mission to the Senate (Canada) Standing Committee on Banking, Trade and Commerce, Study on the State of Domestic and International Financial System (May 29, 2002). HeinOnline -- 1 J. Empirical Legal Stud. 268 2004 Eisenberg and Macey 269 audit firm partners not directly involved in auditing a particular client. This, in turn, may have resulted in a diminution in the incentives of accounting firm partners to monitor the performance of their colleagues. The removal of aider and abettor liability risk also may have, at the margin, reduced audi- tors' incentives to monitor one another,7 a reduction in incentives that was exacerbated in 1995 by passage of the Public Securities Litigation Reform Act (PSLRA).8 These changes may have been exacerbated by the increasing com- plexity involved in performing audits. Auditing became more complex as new and more sophisticated methods of financing proliferated and as the audit rules themselves became more technical and complex. As a conse- quence, audit firms that were engaged by large public companies found that the "audit engagement teams" they assigned to perform audits had to spend increasingly large percentages of their time performing audit services for that client. Thus, for example, the head of Andersen's Enron audit team spent 100 percent of his time on the Enron account.9 This, in turn, led to the capture of auditors by their clients, since auditors' careers increasingly came to depend entirely on the "care and feeding" of single clients. Thus, just as the danger of client "capture" of auditors was increasing, the incen- tives of accounting firms to develop internal corporate governance struc- 7Central Bank of Denver v. First Interstate Bank of Denver, 511 U.S. 164 (1994) (holding that § 10(b) and SEC Rule 10b-5 prohibit only "the making of a material misstatement (or omis- sion) or the commission of a manipulative act" and do not prohibit the aiding and abetting of such acts). This decision was thought to have alleviated substantially the legal risks to outside advisors such as auditors and lawyers. 'Pub. L. No. 104-67, codified in various parts of Titles 15 and 18 of the U.S. Code, §§ 101-109. The PSLRA established new rules of pleading that require plaintiffs' complaints to "state with particularity all facts giving rise to a strong inference that the defendant acted with the required state of mind" when making a misstatement or omission in financial reporting. The PSLRA also delayed the beginning of discovery until after a court has decided whether to allow the case to go forward on the basis of the heightened pleading standards. Prior to passage of the PSLRA, plaintiffs' attorneys could begin to gather documents and interview witnesses as soon as their complaint was filed. The PSLRA also sharply limited the doctrine of "joint and several liabil- ity," which ensures that victims can recover full damages even if one or more of the parties to the fraud cannot pay. Under the PSLRA, those whose reckless misconduct contributes to the fraud can be held responsible for only their proportionate share of victims' losses. As a result, when the primary perpetrator of the fraud is bankrupt, investors cannot fully recover their losses from other entities, such as accounting firms. 9Macey & Sale, supra note 3. HeinOnline -- 1 J. Empirical Legal Stud. 269 2004 270 Major Accounting Firm Audits of Large Clients tures to combat capture may have been decreasing due to the passage of statutes providing limited liability to accounting firm partners. The danger of client capture appears to have been further exacerbated by the growth in the provision of highly profitable consulting services by auditors. If the consulting services accounting firms offer to their audit- clients have higher profit margins than the auditing services, accounting firms will be tempted to use auditing work either as a loss leader or as a mechanism for "'opening the door' with a client for the purpose of pitch- ing their (higher margin) consulting services." ° Providing consulting ser- vices further erodes auditor independence by shifting the balance of power away from the auditor and in the direction of the audit-client when auditors are discussing audit work and retention issues. Worse, consulting services provide a means by which audit-clients can reward auditors for succumbing to the client's wishes about what accounting treatment should be used to report novel or complex transactions and business practices." Where auditors offer clients only audit services, the client's only option is to fire the auditor if the client does not think that the auditor is being suf- ficiently aggressive. But when the accountants also are peddling consulting services, the client can employ a "carrot and stick" strategy that rewards the accounting firm for being compliant and punishes the firm for being inflex- ible. This pressure is particularly acute in an environment in which the firm is the only client of the engagement partner from the accounting firm that is performing the audit, since a partner's inability to procure lucrative con- sulting work would be reflected in the salary, promotion, and bonuses of the partner. As John Coffee has observed, it is difficult for an audit-client to fire its auditor because such dismissals invite "potential public embarrassment, public disclosure of the reason for the auditor's dismissal or resignation, and potential SEC intervention."12 By contrast, where a company is both an audit- client and a consulting client of a particular accounting firm, "the client can easily terminate the auditor as a consultant or reduce its use of the firm's Id. at 1178. "Id. (discussing conversation with Jeffrey Gordon). 12John Coffee, Understanding Enron: It's About the Gatekeepers, Stupid, 57 Bus. Law. 1403, 1411-12 (2002). HeinOnline -- 1 J. Empirical Legal Stud. 270 2004 Eisenberg and Macey 271 consulting services, in retaliation for the auditor's intransigence."" When the client terminates the high-margin consulting services provided by the accounting firm and retains only the low-margin auditing services, there is no need to make any public disclosure. This means that there is no risk that firing the auditor from a consulting engagement will provoke heightened scrutiny from investors, the SEC, or plaintiffs' class action law firms. Another major complaint about accounting firms has been about the quality of their internal corporate governance. The inability or unwill- ingness of officials in Andersen's head office in Chicago to monitor and control the auditors in the field has been highlighted in the securities class action litigation that has followed in the wake of Enron. Carl Bass, an Andersen partner and member of that firm's Professional Standards Group, was apparently removed from the Andersen Professional Standards Group when he tried to correct accounting errors in Enron's financial reporting. It also appears that accounting partners at Andersen were able "to ignore with impunity the advice provided by higher level, more objec- tive experts within the firm" and that these higher-level officials were unwill- ing or unable to assert their authority by following through to ensure that their recommendations were followed.'4 It appears that the Enron-Andersen relation really was an example of an accounting firm captured by its audit-client. A number of surviving Arthur Andersen documents reveal that Arthur Andersen was concerned about, yet covered up or ignored fraudulent account- ing practices by Enron. For instance, Arthur Andersen Professional Standards Group ("PSG") partner Carl Bass sent an e-mail on 12/19/99 to Defendants Steward and Neuhausen expressing opposition to Enron's accounting for [a special purpose entity (SPE) funded with Enron equity called "LJM2" with whom Enron had a contractual relationship] and urged Arthur Andersen not to support it. Again on 2/4/00 Bass sent another e-mail to Stewart stating that Bass thought that a particular SPE had no real substance and that he was annoyed that Enron would receive appreciation on the Enron stock that had been con- tributed to that SPE. That information was also sent to [other members of the Enron audit team].... Bass had... commented that "this whole deal looks like there is no substance." Later, on March 4, 2001, just before Bass was removed as PSG advisor for the Enron audit team, Bass sent Stewart another e-mail crit- icizing Enron's accounting for the Blockbuster and Raptor transactions, which, 13Id. at 1412. 4Macey & Sale, supra note 3, at 1180. HeinOnline -- 1 J. Empirical Legal Stud. 271 2004