CPAs Perceptions of Auditor Independence: An Analysis of Views Before and After the Collapse of Enron

Transcription

1 CPAs Perceptions of Auditor Independence: An Analysis of Views Before and After the Collapse of Enron Deborah L. Lindberg, D.B.A. Associate Professor Department of Accounting Illinois State University Frank D. Beck, Ph.D. Assistant Professor Department of Sociology & Anthropology Illinois State University November 2002 Direct all correspondence to: Deborah L. Lindberg, Illinois State University, College of Business, Department of Accounting, Campus Box 5520, Normal, IL, USA ; Telephone: (309) ; Fax: (309) ; The authors are grateful for financial support from the American Accounting Association and Illinois State University; research support from the Illinois Society of Certified Public Accountants; helpful comments and suggestion from Dr. Dennis Patten on an earlier draft of this manuscript; and work performed by Illinois State University Research Assistants Vicki Lavery, Nick Nourse, Glen Steiner, Laura Vahling, and Yan Zheng.

2 CPAs Perceptions of Auditor Independence: An Analysis of Views Before and After the Collapse of Enron Deborah L. Lindberg Illinois State University Frank D. Beck Illinois State University ABSTRACT: Auditor independence helps to ensure quality audits and contributes to financial statement users reliance on the financial reporting process, thus increasing capital market efficiency (McGrath, Siegel, Dunfee, Glazer & Jaenicke 2001). Hence, auditor independence is often referred to as the cornerstone of the auditing profession, since it is the foundation for the public s trust in the attest function (Caswell & Allen 2001; The Economist February 9, 2002). By contrast, Byrnes, McNamee, Brady, Lavelle & Palmeri (2002) and many others (e.g., Greising 2002; The Economist January 19, 2002; The Economist February 9, 2002) have stated that the collapse of Enron has had a very negative impact on the perception of auditor independence. However, we do not need to rely on anecdotal evidence to determine the effect of the Enron bankruptcy on perceptions of auditor independence prior to Enron declaring bankruptcy, we had developed an extensive survey instrument on auditor independence that was sent to 1,500 CPAs in October 2001; after Enron declared bankruptcy, the survey was sent to another 1,500 CPAs drawn from the same database, asking their views on auditor independence. Consistent with our predictions, the results of our research indicate that CPAs perceptions of the effects of non-audit services on auditor independence are more negative after the bankruptcy of Enron, compared to CPAs perceptions of such services on auditor independence before the bankruptcy of Enron. In addition, CPAs hold a more conservative view of whether a material transaction or event detrimentally affects auditor independence after the bankruptcy of Enron, compared to before Enron declared bankruptcy. Further, our findings suggest that auditors believe non-audit services and other issues that threaten auditor independence detrimentally affect the public s perception of independence to a greater extent than they adversely influence actual independence. Key Words: Auditors; CPAs, Independence In Fact ; Independence In Appearance ; Perceptions of Independence; Enron. 1

3 CPAs Perceptions of Auditor Independence: An Analysis of Views Before and After the Collapse of Enron A Discussion of Auditor Independence INTRODUCTION Auditor independence has been termed the cornerstone of the auditing profession, since it is the foundation for the public s trust in the attest function (see, e.g., Caswell & Allen 2001; The Economist February 9, 2002). Auditor independence helps to ensure quality audits and contributes to financial statement users reliance on the financial reporting process. McGrath, Siegel, Dunfee, Glazer & Jaenicke (2001, 40) argue that when independent auditors render unbiased audit decisions, the broader goal of auditor independence, namely to support user reliance on the financial reporting process and to enhance capital market efficiency, is accomplished. However, several major instances of misstated earnings have been reported over the last several years (e.g., Greising 2002; King 1999; Levitt 2000; Loomis 1999; MacDonald 2000). These misstatements have led many to question the effectiveness of various aspects of the audit function, especially auditor independence, previously prompting the Securities and Exchange Commission (SEC), in 2000, to adopt rules identifying nine non-audit services that are deemed inconsistent with auditor independence. Further, the SEC is considering additional measures to strengthen actual and perceived auditor independence, especially in light of the Enron debacle. Auditor independence has long been couched in terms of independence in fact and independence in appearance. An auditor is in fact independent if he or she has the ability to make independent audit decisions even if there is a perception of lack of independence or if the auditor is placed in a potentially compromising position (e.g., Guy, Alderman & Winters 1999; 2

4 McGrath et al. 2001). However, even when the auditor is in fact independent, one or more factors may lead the public to believe the auditor does not appear to be independent. This may in turn cause users of financial statements to believe they cannot rely on financial information (Lowe & Pany 1995, 1996). Thus, the perception of investors and other users of financial information is also an important consideration in the auditor independence debate. Because the goal of auditor independence is to support user reliance on the financial reporting process, auditors must be independent both in fact and in appearance. Actual auditor independence is a mental state, and is in essence embodied in an individual auditor s mind. Accordingly, it is impossible for investors and other users of financial statements to accurately assess actual auditor objectivity (e.g., McGrath et al. 2001). Therefore, users of financial statement information can only evaluate an auditor s appearance of objectivity. Thus, even when an auditor in fact acts independently and issues an unbiased audit opinion, if investors and other users of the financial statement information do not believe that the auditor was independent, investor confidence is eroded and capital market efficiency will suffer. Further, given recent changes to the rules and guidelines governing independence, the definition of factual independence has changed. It is evident then, that independence is socially defined 1. While both independence in fact and in appearance are required to achieve the goal of independence, specific expectations may have been altered by the Enron debacle and the negative publicity the auditing profession has received. Therefore, both facets of auditor independence are explored in our study. In order to address both independence in fact and in appearance, we conducted individual interviews and focus groups with both auditors and users of the auditors opinions. Specifically, we determined from auditors: 1) how auditors view independence, and 2) how the SEC rules or 3

5 the former Independence Standards Board (ISB) guidelines affect how they do their work. Similarly, we asked users of financial statement information: 1) how they view independence, 2) what activities on the part of auditors would raise questions about independence in their minds, and 3) how SEC rules or former ISB guidelines shape their interpretation of audit reports. In addition, we sent an extensive survey that solicited perceptions of numerous aspects of auditor independence to CPAs before Enron declared bankruptcy, and to CPAs after Enron declared bankruptcy. 2 Our research indicates that CPAs perceptions of the effects of non-audit services on auditor independence are more negative after the bankruptcy of Enron, compared to CPAs perceptions of such services on auditor independence before the bankruptcy of Enron. In addition, CPAs hold a more conservative view of whether a material transaction or event detrimentally affects auditor independence after the bankruptcy of Enron, compared to before Enron declared bankruptcy. Further, the results of our survey suggest that auditors believe non-audit services and other issues that threaten auditor independence detrimentally affect the public s perception of independence more so than they adversely influence actual independence. The remainder of this paper is organized as follows. The next section addresses Enron and other recent developments. A discussion of prior research and the development of our hypotheses, the methodology we used, and the results of our analyses follow. The final section summarizes our conclusions, implications of our findings, and opportunities for further research. ENRON AND OTHER RECENT DEVELOPMENTS Several major instances of misstated earnings have been reported over the last several years (e.g., Bazerman, Morgan & Loewenstein 1997; Greising 2002; King 1999; Levitt 2000; Loomis 1999; MacDonald 2000). The Securities and Exchange Commission (SEC) and many other 4

6 interested parties have expressed concern about companies overreaching in terms of permissible accounting treatments, as well as about alleged accounting abuses involving such matters as accounting for business combinations, revenue recognition and restructurings, along with alleged misuse of the materiality concept. This overreaching often is characterized as earnings management (O Malley, Chookaszian, Kolton, Longstreth, Lowenstein, Palmrose, Peters & Saul 2000, p. 2). In most cases, the auditors expressed an unqualified ( clean ) opinion on the financial statements before the irregularities were brought to light, prompting many, including former SEC Chairman Arthur Levitt (Levitt 2000), to question the effectiveness of certain aspects of the audit function, particularly issues related to auditor independence. Currently, the SEC is considering additional measures to strengthen auditor independence, especially in light of the Enron debacle. Many people have stated that Enron has had a very negative impact on the perception of auditor independence (e.g., Byrnes, McNamee, Brady, Lavelle & Palmeri 2002; Greising 2002; The Economist January 19, 2002; The Economist February 9, 2002). In sum, investor confidence in audited financial statements and the financial reporting process has recently been eroded by numerous accounting irregularities (Freehling 2002; Kemper 2002), most notably the bankruptcy of Enron. Byrnes et al. (2002) as well as many others (e.g., Chicago Tribune April 28, 2002; The Economist January 19, 2002; Sloan 2002) have criticized Arthur Andersen, Enron s former auditors for, among other things, lacking independence, since the accounting firm earned more revenue from non-audit services than from audit services. Thus, auditor independence is an important, unresolved issue, and one that was further explored in our study. PRIOR RESEARCH AND DEVELOPMENT OF HYPOTHESES 5

7 Economic Bonding Hypothesis In 2000 the Securities and Exchange Commission adopted amendments to the rules governing relationships between independent auditors and their SEC clients. 3 These SEC rules identified nine non-audit services that are deemed inconsistent with an auditor s independence: 1) Bookkeeping or Other Services Related to the Audit Client s Accounting Records or Financial Statements. An audit firm cannot maintain or prepare the audit client s accounting records or prepare the audit client s financial statements that are either filed with the Commission or form the basis of financial statements filed with the Commission. (An exception includes providing services in emergency situations, provided the accountant does not undertake any managerial actions or make managerial decisions). 2) Financial Information Systems Design and Implementation. The auditor cannot operate or supervise the operation of the client s IT systems. However, the auditor could provide IT consulting services provided certain criteria are met. 3) Appraisal or Valuation Services or Fairness Opinions. Restrictions on these services apply only where it is reasonably likely that the results of any valuation or appraisal would be material to the financial statements, or where the accountant would audit the results. 4) Actuarial Services. Actuarial-oriented advisory services are limited only when they involve the determination of insurance company policy reserves and related accounts. 5) Internal Audit Services. An audit firm is allowed to perform up to 40 percent (measured in terms of hours) of an audit client s internal audit work. (This rule provides an exception for smaller businesses by excluding companies with less than $200 million in assets). 6) Management Functions. An auditor s independence is considered impaired when the accountant acts, temporarily or permanently, as a director, officer, or employee of an audit client, or performs any decision-making, supervisory, or ongoing monitoring function for the client. 7) Human Resources. An auditor is not able to recruit, act as a negotiator on the audit client s behalf, develop employee testing or evaluation programs, or recommend, or advise that the client hire, a specific candidate for a specific job. 8) Broker-Dealer Services. An auditor cannot serve as a broker-dealer, promoter or underwriter of an audit client s securities. 9) Legal Services. An auditor cannot perform services for an audit client in which the person providing the services must be admitted to practice before the courts of a U.S. jurisdiction. It should be noted that while the provisions of the independence rules allow audit firms to offer consulting services, such as information technology (IT) and internal auditing, the services are subject to certain restrictions. Among other things, the rules now require clients of audit 6

8 firms to disclose annually the fees they paid for auditing, IT consulting, and all other services (McGrath 2001). Although these disclosures are intended to give users of financial statements information to assist them in assessing auditor independence, many interested parties did not believe that such disclosures significantly impacted the audit function. Hence, as previously noted, there are many issues surrounding auditor independence that remain unanswered. Enron declared bankruptcy on December 2, 2001, which, at the time, was the largest bankruptcy in U.S. history. Anecdotally, many people CPAs and non-cpas alike have stated that the Enron bankruptcy has forever altered the public s view on auditor independence, since Enron s auditors had previously issued unqualified opinions on Enron s financial statements (e.g., Byrnes et al. 2002; Greising 2002; The Economist January 19, 2002; Largay 2002; The Economist February 9, 2002). As previously noted, many have criticized Enron s former auditors for lacking independence, since the accounting firm earned more revenue from non-audit services than from audit services (e.g., Byrnes et al., 2002; Chicago Tribune April 28, 2002; The Economist January 19, 2002; Sloan 2002). Further, regulators and critics of the accounting profession view an audit firm s provision of consulting services to an audit client as a conflict of interest (Manor 2002; Morgenson 2002). Bazerman et al. (1997) and other researchers (e.g., Simunic 1984; Swanger & Chewning 2001) contend that the provision of most non-audit services threatens auditor independence, since an economic bond, which the auditor doesn t want to lose, develops between the client and the accounting firm. Regardless of whether non-audit services affect actual independence, certainly the appearance of auditor independence has been adversely affected by such services, so much so that after the bankruptcy of Enron, several of the Big 5 CPA firms have sold their consulting practices (e.g., Chicago Tribune May 3, 2002; Public Accounting Report February 15, 2002). 7

9 Therefore, our first hypothesis is that after the bankruptcy of Enron, CPAs will perceive the provision of non-audit (consulting) services to auditing clients as weakening auditor independence more so than they did before the bankruptcy of Enron. In hypothesis form, this can be stated as: H1: When an audit firm provides non-audit services to its audit clients, CPAs perceptions of the effect of such services on auditor independence will be more negative after the bankruptcy of Enron, compared to CPAs perceptions of such services on auditor independence before the bankruptcy of Enron. Perception as Important as Reality As previously discussed, an auditor is considered independent in fact if he or she has the ability to make independent audit decisions even if there is a perception of lack of independence or if the auditor is placed in a potentially compromising position (e.g., Guy, Alderman & Winters 1999; McGrath et al. 2001). However, even when the auditor is in fact independent, one or more factors may lead the public to believe the auditor does not appear to be independent. Thus, the perception of users of financial information is also an important consideration of auditor independence (Lowe, Geiger & Pany 2001; Olazabal & Almer 2001). Many people who have written about the Enron debacle note that perception becomes reality where auditor independence is concerned (e.g., Byrnes 2002; The Economist Jan ). Therefore, our next two hypotheses concern independence in fact and in appearance and predict that there will be no significant difference in CPAs perceptions of independence in fact and in appearance either before or after Enron declared bankruptcy. H2a: There will be no significant difference between CPAs perceptions of independence in fact and independence in appearance before Enron declared bankruptcy. H2b: There will be no significant difference between CPAs perceptions of independence in fact and independence in appearance after Enron declared bankruptcy. 8

10 The Role of Materiality Materiality is an issue that accountants and auditors have long grappled with. A financial statement item or event is deemed material if it would have an effect on the decision-making of an informed user of financial statement information (Feldstein 2002; Hilzenrath 2001). In other words, materiality is a criterion used by accountants and auditors to determine the importance of an item to the presentation of financial information (Guy et al. 1999). However, there are no pronouncements that help accountants and auditors calculate a dollar amount to be used as a materiality cutoff, or to specifically determine whether or not an item is material. Rather, professional standards state that the determination of materiality is a matter of professional judgment, and should be based on such factors as the size of an item, the nature of an item, the expected users of financial information, and the cumulative effect of items under consideration (see, e.g., AICPA 1983; FASB 1975; FASB 1980). Thus, materiality judgments are subject to a great deal of subjectivity (Kieso, Weygandt & Warfield 2001). After the bankruptcy of Enron, materiality issues have been extensively covered in the financial press (see, e.g., Byrnes et al. 2002; Feldstein 2002; Hilzenrath 2001; Krantz 2002; Tweeton 2002). Due to the negative publicity that the public accounting profession has received, and because of the subjectivity of materiality determinations, we expected that CPAs will take a more conservative view of what constitutes a material transaction or event after Enron declared bankruptcy compared to before Enron declared bankruptcy. In other words, ceteris paribus, the materiality threshold has been lowered as the result of the Enron debacle. Therefore, our next hypothesis is stated as: H3: CPAs will have a more conservative view of whether a material transaction or event detrimentally affects auditor independence after the bankruptcy of Enron, compared to 9

11 before Enron declared bankruptcy. METHODOLOGY Development of Survey Instrument Our survey instrument was constructed in three phases. These phases included: 1) individual interviews, 2) focus groups, and 3) two surveys one sent before Enron declared bankruptcy, and the other sent after Enron declared bankruptcy. We began our research by gathering data through structured, yet open-ended interviews with auditors and users. This was followed by focus group discussions. The interviews and focus groups aided us in our construction of a selfadministered questionnaire that was mailed to CPAs. Such an approach is commonly termed triangulation (Babbie 1995). In short, we attempted to gather thorough and useful data from a randomly sampled and generalizable group of respondents. Open-ended Interviews Because there has already been much written about auditor independence, qualitative research, in particular in-depth interviews, can produce novel and fresh insights into auditor independence issues. In other words, we believe that in-depth interviews helped us glean insight into what the average CPA or user of financial statement information believes constitutes auditor independence. As discussed previously, we interviewed both auditors and users of financial statements. The interviews were designed to determine views on auditor independence issues, especially the factors affecting the appearance of independence. We concentrated on issues related to the appearance of independence; even if auditors are in fact independent, the lack of the appearance of independence compromises overall independence. 10

12 Using an interview schedule, we gathered data from selected auditors and users of financial statement information. The questions were designed to encourage a conversational tone with the respondents, yet still obtain answers to the same questions from each respondent. The interviews were audio taped, transcribed, and analyzed. The transcriptions and analysis allowed us the ability to find patterns or contradictions in the responses. The data gathered through these openended interviews helped us frame the other two steps in the research process, the last of which will be more generalizable. Focus Groups We conducted two focus group discussions. One focus group consisted of only auditors and the other included both auditors and users of audited financial information. The latter focus group provided us with useful data about how auditors and users of audit documents view auditor independence issues differently. While structured, the focus groups served as brainstorming sessions on auditor independence in fact (according to the rules and guidelines), as well as the appearance of auditor independence. The focus groups were video taped and analyzed to attempt to determine patterns or contradictions that existed between these responses and those of the initial interviews. Focus group members were used as information sources for developing our survey questions. 11

13 Survey The data from the in-depth interviews and focus groups was used to construct a questionnaire. Babbie (1995) contends that the construction of such a survey is best done with input from the use of such qualitative techniques; otherwise, the questions asked would be guided only by assumptions of what is important. Survey Participants Prior to Enron declaring bankruptcy, we had developed an extensive survey instrument on auditor independence that was sent to 1,500 CPAs in October After taking into consideration surveys that were returned as undeliverable, we received 347 usable responses, representing a 23.2% response rate. After Enron declared bankruptcy, the survey was sent to another 1,500 CPAs drawn from the same database, asking their views on auditor independence. After taking into consideration surveys that were returned as undeliverable, 4 and removing one potential respondent from the mailing list because the CPA had previously participated in one of our focus groups, we received 379 usable responses, representing a 26.0% response rate. The mailing lists of CPAs were provided by the Illinois Society of Certified Public Accountants, and were randomly generated from their membership records. The majority (69 percent) of auditors in both samples are employed by local CPA firms or are sole proprietors. Fifteen percent worked for one of the Big 5 firms and the remaining respondents are split between regional and national firms. We also asked the CPAs about their primary responsibility at the firm. Approximately 45 percent work on audits and 54 percent do tax work. Some respondents indicated more than one primary responsibility, so 18 percent of respondents indicated that they work in consulting, another 6 percent work in quality assurance, and 2 percent work in recruiting. Another 11 percent indicated Other on the survey instrument; 12

14 a majority of these have all of the above responsibilities, from audits to recruiting. Forty-nine percent of the CPAs who responded to the survey have performed audits or been members of audit teams for 10 or more years. Another 18 percent have worked in that profession for five to nine years and 20 percent for one to four years. Eleven percent of the respondents have less than one year of audit experience or no experience performing audits at all. The majority (76 percent) of the CPAs who responded to our survey are men. Survey Tasks Survey participants were asked to provide a multitude of opinions about issues that may impair auditor independence. Respondents were also asked to give their view about factors that could mitigate independence concerns. The majority of questions on the survey were structured so that the participants gave their perceptions of how various issues affected both independence in fact and independence in appearance. Definitions of independence in fact and independence in appearance were provided at the beginning of the survey, to help assure that respondents were consistently using these terms. 5 For one set of questions, survey participants were instructed to provide their opinions about specific non-audit services on a scale anchored by Definitely Weakens Independence in Fact ( 1 ) and Definitely Strengthens Independence in Fact ( 5 ). Respondents were also asked to provide their perceptions of how the same non-audit services affected independence in appearance. In a similar vein, participants were asked to provide their perceptions, in several different formats, of how a multitude of other issues affects auditor independence, including but not limited to, such factors as: Materiality Disclosure of nonaudit fees Size of the client 13

15 Employment of auditors by former clients Peer reviews Client outsourcing of internal audit services Risk of litigation Risk of loss of reputation Potential for non-audit revenue Stock ownership of audit clients Training of auditors Auditors signing forms assuring independence Length of the client-auditor relationship Culture of the audit firm An individual auditor s ethical and moral characteristics Board of Directors Audit Committees SEC regulations Dependent Variables The participants responses to the survey questions were used as dependent variables. As noted in the methodology section, numerous dependent measures were taken, generally on a 1 to 5 scale. Independent Variables Data for several independent variables were obtained as a result of our survey questions and procedures, including pre-enron bankruptcy respondents vs. post-enron bankruptcy respondents, firm size of respondents employers, experience level of survey participants, gender of respondents, etc. The specific statistical application of various independent variables will be discussed in the results section. 14

16 RESULTS Tests of Economic Bonding Hypothesis Our first hypothesis predicts that CPAs perceptions of the effects of non-audit services on auditor independence will be more negative after the bankruptcy of Enron, compared to CPAs perceptions of such services on auditor independence before the bankruptcy of Enron. To test H1, we predicted respondent s negativity toward a particular non-audit service on independence in fact using a dichotomous variable for whether the CPA was surveyed prior to or after the Enron bankruptcy, while controlling for the size of company they work for, their job experience performing audits, and gender. The F-values and the slopes of the dichotomy are presented in Table 1. Note that the slope is actually a difference of means test, controlling for the other factors. Those surveyed after the bankruptcy are coded 1, else 0. [Insert Table 1 about here] For the ten non-audit services we asked about in the survey, differences exist between CPAs surveyed before and after the Enron bankruptcy for nine of them. Clearly, audit firms performing non-audit services for clients is more of a concern for CPAs after the Enron bankruptcy than it was before. The largest difference has to do with the effect the provision of internal audit services by an audit firm has on independence in fact. However, an examination of the means for the two groups indicates that the provision of management functions, brokerdealer services, or legal services also concern CPAs, whether measured before or after the Enron bankruptcy. The non-audit service for which no significant difference exists pre- or post-enron, financial information systems design, is perceived as least problematic for the maintenance of independence in fact. We are aware that the R-square values for the models are quite low suggesting that more factors are at work here than we inquired about in the survey. 15

17 We also tested H1 with respect to independence in appearance. The patterns are not as clear, but intriguing nonetheless. CPAs opinions regarding the effect of audit firms providing non-audit services to clients on independence in appearance have grown more negative for only five of the ten non-audit services on which we focused. The largest difference, again, is for the provision of internal audit services. In addition, after the Enron bankruptcy, concern about independence in appearance is seen as in greater jeopardy if the audit firm also provides bookkeeping or legal services, financial information systems implementation services, or appraisal valuation. The differences in opinions for the other five non-audit services are insignificant once gender, amount of experience, and size of the firm for which the CPA works are controlled for. However, the differences are still in the expected direction greater negativity among CPAs surveyed after the Enron bankruptcy. Comparing the means for Table 2 to those of Table 1 also indicates that the provision of non-audit services is more of a concern for independence in appearance than they are for independence in fact. [Insert Table 2 about here] To further test this hypothesis, we performed a factor analysis on the ten non-audit services included in our questionnaire and respondents perceptions of the effect of these services on auditor independence in fact and in appearance. 6 The factor analysis revealed that eight of the non-audit services (actuarial services, internal audit services, bookkeeping, legal services, appraisal or valuation services or opinions, human resources, management functions, and brokerdealer services) are related. We used a principal axis factor analysis procedure where the factor solution was determined using the criteria of eigenvalues > 1, the scree test, and the interpretability of factors (Gorsuch 1983). We used the maximum likelihood extraction method, and the varimax with Kaiser normalization rotation method. The rotation converged 16

18 in three iterations. Consistent with Ford, MacCallum & Tait (1986), we considered loadings of.40 or higher as significant. The eight non-audit services identified as related were summed together to create the dependent variable of non-audit services. An ANOVA was performed specifying non-audit services as the dependent variable, with pre-enron vs. post-enron respondents, firm size, experience, and gender as between subject factors. The results of this analysis (not shown) indicate a highly significant main effect for pre-enron respondents vs. post-enron respondents (F = 16.03, p <.001 for the in fact model and F = 11.01, p <. 01 for the in appearance model). This result also indicates that the bankruptcy of Enron did significantly affect CPAs perceptions of non-audit services on auditor independence. Tests of Independence Perceptions Hypothesis 2a posits that there will be no significant difference between CPAs perceptions of independence in fact and independence in appearance before Enron declared bankruptcy. Hypothesis 2b posits the same for CPAs after Enron declared bankruptcy. In short, are CPAs more concerned about each item impairing or failing to maintain independence in appearance than they are about each item impairing or failing to maintain independence in fact? We thought they would not be. However, the series of paired difference t-tests in Table 3 clearly demonstrates that CPAs are more concerned about many non-audit services and audit practices affecting independence in appearance than they are about those same issues affecting independence in fact. Twenty-nine of the 31 tests performed on the responses of CPAs surveyed before the Enron bankruptcy are significant and in each case the average score for in appearance concerns is greater than for in fact. Twenty-eight of the 31 tests are significant for 17

19 CPAs after Enron declared bankruptcy. Again, the concerns about each item affecting independence in appearance are greater than the relative concern for independence in fact. [Insert Table 3 about here] As partial support for Hypothesis 2a and 2b, we also compared the relative concern of CPAs for each issue affecting independence in fact to independence in appearance. For all 31 issues tested among CPAs before the Enron bankruptcy, the tau-b is moderately positive and significant. The same holds true for CPAs after the Enron bankruptcy. This means that as concern over a certain issue affecting independence in appearance increases, so does the concern over that same issue affecting independence in fact. In other words, while concern over the effect of these issues on independence in appearance is greater after Enron declared bankruptcy, concern over each issue s effect on independence in fact trends the same way. Further, of interest to us and to be tested in the future, is why a large percentage (29 out of 31) of the tau-b coefficients are lower among the CPAs surveyed post-enron than they were among the CPAs surveyed pre-enron. In other words, the association between what issues affect independence in fact and what factors affect independence in appearance has weakened. We also find it interesting that the average concern for each issue is higher for the CPAs surveyed after Enron declared bankruptcy than it is for CPAs surveyed before the bankruptcy. This is further support for Hypothesis 1, which was previously tested, and Hypothesis 3, to be tested next. Tests of The Role of Materiality H3 predicts after the bankruptcy of Enron, CPAs will have a more conservative view of whether a material transaction or event detrimentally affects auditor independence compared to before Enron declared bankruptcy. One of our survey questions asked participants to indicate, on 18

20 a scale anchored by Definitely Does Not ( 1 ), and Definitely Does ( 5 ), whether the materiality of the amount of non-audit fees impairs independence in fact and in appearance. Therefore, to test H3, an ANOVA was performed specifying the participants responses to the in fact portion of the question as the dependent variable, with pre-enron vs. post-enron respondents, firm size, experience, and gender once again used as between subject factors. The results of this analysis indicate significant main effects for three of the four independent variables: pre-enron respondents vs. post-enron respondents (F = 19.50, p <.001 presented in Table 4); firm size (F = 20.33, p <.001); and gender (F = 8.91, p <.01). Moreover, the results are in the expected direction, since the mean of the dependent variable was 3.44 for the CPAs who participated in our survey after Enron declared bankruptcy, versus only 2.98 for the CPAs who participated in our survey before Enron declared bankruptcy. Therefore, these results indicate that after the bankruptcy of Enron, CPAs had a more conservative view of whether a material transaction or event detrimentally affects auditor independence in fact compared to before Enron declared bankruptcy. Based on a similar factor analysis to that performed for H1 above, we identified nine items that were related to each other, including materiality, dollar amount of waived audit adjustments, client audit revenue, estimation of a valuation allowance, outsourcing of internal audit services, low-balling of initial audit fees, potential for non-audit revenue, compensation for referrals, and contingency fees. Since all these items appeared to us as part of an underlying and unmeasured concern with the effect of dollars on audit firm-client relations and independence, we perform the same test on the other eight items that we performed on concerns over the materiality of non-audit fees. Table 4 presents these results. [Insert Table 4 about here] 19