Вы находитесь на странице: 1из 17

AUDITING: A JOURNAL OF PRACTICE & THEORY Vol. 22, No. 2 September 2003 pp. 53ñ69

Audit Fees, Nonaudit Fees, and Auditor Reporting on Stressed Companies

Marshall A. Geiger and Dasaratha V. Rama

SUMMARY: The SEC and legislators have expressed concerns that independence may be negatively impacted if auditors perform significant nonaudit services for their audit clients, and that providing lucrative nonaudit services to clients may make it more likely that auditors will ìsee things the clientís way.î Such concerns are particularly salient in the context of issues that involve significant auditor judgment, as in the case of reporting decisions related to going-concern uncertainties for financially stressed clients. In this study we examine the association between the magnitude of audit and nonaudit fees and auditor report modification decisions for financially stressed manufacturing companies. In our analysis we control for financial stress, company size, reporting lag, default status, audit committee effectiveness, and management plans. The results indi- cate a significant positive association between the magnitude of audit fees and the likelihood of receiving a going-concern modified audit opinion, but we find no significant association between nonaudit fees and audit opinions. Additional analyses also find no significant relationship between the ratio of nonaudit service fees to audit fees and reporting decisions, and indicate that our results are robust across alternative model, variable, and sample specifications. We also control for the potential endogeneity of audit opinions, audit fees, and nonaudit fees, and find the same positive association of audit fees with opinions, but no association between nonaudit fees and audit opinions. Overall, we find no evidence of a significant adverse effect of nonaudit fees on auditor reporting judgments for our sample of distressed companies.

Keywords: nonaudit fees; audit fees; auditor independence; audit reporting.

Data Availability: A list of sample companies is available from the first author. Other data are available from the sources identified in the paper.

INTRODUCTION T his study examines the association between audit and nonaudit service (NAS) fees received by an auditor and the auditorís decision regarding the type of opinion to render to a finan- cially stressed company. Motivation for this study comes from the interest of legislators and

regulators about the impact of auditorsí fees on audit judgments (SEC 2000a; U.S. GAO 2002; U.S.

Marshall A. Geiger is a Professor at the University of Richmond, and Dasaratha V. Rama is a Professor at Florida International University.

We thank the two anonymous reviewers, Mark DeFond (associate editor), Bill Messier (editor), Arnie Wright (former editor), Brendan OíConnell, Kannan Raghunandan, Srini Sankaraguruswamy, and participants at the 2002 AAA Annual Meeting for their helpful suggestions on earlier drafts of this paper.

Submitted: February 2002 Accepted: January 2003

53

54

Auditing, September 2003

House of Representatives 2002), the continued interest regarding the importance and frequency of auditors issuing going-concern modified (GCM) opinions to stressed companies (U.S. House of Representatives 1985, 1990; Weil 2001; Dietz 2002; NASDAQ 2002), and the recent actions of the U.S. Congress and the Securities and Exchange Commission (SEC) to prohibit auditors from supply- ing certain types of nonaudit services to their audit clients and to require publicly traded companies in the U.S. to disclose the type and amounts of fees paid to the external auditor (Sarbanes-Oxley Act 2002; SEC 2000b). In this study, we examine audit reports for a sample of manufacturing companies in financial stress that filed their proxy statements after February 5, 2001, and test for the association between the type of audit opinion issued (going-concerned modified or not) and the level of audit and NAS fees. The auditing profession has come under increased scrutiny over the past several years about the growing amount of nonaudit fees received from audit clients and the possible negative impact of such fees on auditor independence (Levitt 2000). The SEC and legislators (SEC 2000a, 2000b; U.S. House of Representatives 2002; U.S. Senate 2002) have asserted that significant nonaudit service fees can adversely impact auditor independence and impair auditor decision making, especially when those decisions involve a substantial amount of professional judgment. Such concerns over auditor independence and the level of NAS performed for audit clients led to the SECís recent enactment of new rules related to nonaudit services supplied by auditors for SEC registrants. The new rules restrict certain types of nonaudit services and also include a fee disclosure requirement (SEC 2000b). Beginning February 5, 2001, SEC registrants must disclose the magnitude of audit fees, financial information systems fees, and all other NAS fees. Subsequent to the Enron failure, legislative concerns related to nonaudit services led to the recently enacted Sarbanes-Oxley Act (2002). This Act seeks to prohibit certain types of non- audit services, and also requires that a companyís audit committee explicitly approve any other nonaudit service purchase from the incumbent auditor. This study empirically examines assertions that auditors may act more favorably toward those clients from whom they receive higher NAS fees. We examine the audit reports rendered to finan- cially stressed manufacturing companies and the relative magnitude of nonaudit fees paid by such companies to their auditors. If nonaudit fees can be argued to have an impact on audit judgments, then it is also plausible that the magnitude of audit fees could influence auditor judgments. Hence, we also examine if there is an association between the magnitude of audit fees and audit opinions. Our analyses of stressed companies find no significant association between receiving a going- concern modified (GCM) audit report and the magnitude of NAS fees. However, we do find a positive association between audit fees and GCM opinions. While the small sample size in our study may limit the statistical power of our tests, our results are extremely robust across numerous alterna- tive models and variable and sample specifications. In addition, as part of our ìFurther Analyses,î we control for the possible endogeneity of audit opinions with audit fees and NAS fees and obtain the same results. The consistent findings obtained from numerous additional analyses give further credibility to our results. Our findings are also consistent with contemporary research on NAS fees in the U.S. (DeFond et al. 2002) and lend additional evidence to prior research on NAS fees and auditor independence/ reporting decisions in other countries (Craswell 1999; Lennox 1999). Our results are also consistent with audit pricing studies that have found modified audit opinions require additional audit work and therefore are associated with increased audit fees (Simunic 1980; Palmrose 1986; Francis and Simon 1987; Barkess and Simnett 1994; Bell et al. 2001).

Auditing: A Journal of Practice & Theory, September 2003

Geiger and Rama

55

BACKGROUND AND RESEARCH QUESTIONS Nonaudit Services and Auditor Judgments The association between NAS fees and the possibility of impaired external auditor indepen- dence has been discussed in the literature for quite some time (c.f., Mautz and Sharaf 1961; Hylton 1964; U.S. Senate 1976). At issue is whether the revenues obtained from these additional services might create a situation where the external auditor becomes too closely aligned with the client and begins to lose a true sense of independence, which in turn affects the auditorís judgment. Auditors who perform significant NAS for audit clients have an additional economic incentive to retain the client, possibly at the risk of deciding difficult issues in the clientís favor so as not to present a disagreement with the management that might ultimately lead to dismissal (Levitt 2000). The growth in NAS and the resultant independence-related concerns have been instrumental in the SECís recent action to limit or prohibit certain types of NAS and to require disclosure of nonaudit and audit fees paid by registrants to their external auditors. The former Chairman of the SEC criticized audit firms for using auditing as ìa loss-leader retained as a foot in the door for higher-fee consulting servicesî (Levitt 1996). He also noted that consulting and other services may ìshorten the distance between the auditor and managementî and that ìindependenceóif not in fact, then certainly in appearanceóbecomes a more elusive proposition.î In its rule proposal, the SEC (2000a) stated:

Our concern is Ö as auditors become involved in a broad array of business arrangements with their clients, they come to be seen by themselves, their firms, their clients, and investors less as exacting, skeptical professionals who must be satisfied before signing off on the financial state- ments, and more like any other service vendor who must satisfy the client to make the sale.

In the final rules, the SEC (2000b) expressed its concern that ìthe rapid rise in the growth of nonaudit services has increased the economic incentives for the auditor to preserve a relationship with the audit client, thereby increasing the risk that the auditor will be less inclined to be objective.î The Commission suggested that significant NAS fees could increase the pressure on auditors to ìsee it the way the client doesî (SEC 2000b). The SEC also noted that such pressures on auditor indepen- dence may be greatest in the gray areas, where the rules may not be very clear or when a considerable amount of the auditorís professional judgment must be applied. Along these lines, Chairman Levitt (2000) noted:

It is not the bright line of right and wrong that the lack of auditor independence often implicates as much as it is that gray area where the answers arenít so clear; where the temptation to ìsee it the way your client doesî is subtle, yet real.

Auditor Reporting Independence and NAS Fees DeAngelo (1981) and Watts and Zimmerman (1986) have defined auditor independence as the joint probability of detecting and reporting errors. DeAngelo (1981) and Craswell (1999) note that the first element of the definition relates more to auditor technical competence or capability, and the second relates more directly to the issue of auditor independence. Thus, auditor independence depends not only on the ability of the auditor to identify problems, but also to report those problems appropriately. Since prior research has indicated that auditors typically do not have difficulty identi- fying cases involving going-concern issues for financially stressed clients (Kida 1980; Mutchler 1984, 1986; Simnett and Trotman 1989), the independence issue lies more directly with auditor reporting of these difficulties. Thus, Firth (1997), Craswell (1999), and Lennox (1999) have argued that a direct test of the effects of NAS fees on auditor independence is the examination of auditor reporting decisions, particularly on stressed companies.

Auditing: A Journal of Practice & Theory, September 2003

56

Auditing, September 2003

The decision on what type of opinion to render on the clientís financial statements is the final, cumulative audit decision and is subject to a considerable amount of professional judgment and negotiation with client management. SAS No. 59 (AICPA 1988) requires auditors to assess the continued viability of their clients in each audit engagement, and if there is substantial doubt about the ability of the entity to continue as a going concern, then the auditor must evaluate managementís plans and mitigating circumstances. Both the initial going-concern assessment and the evaluation of the appropriateness and probability of success of managementís plans involve highly subjective judgments. 1 Negotiations about the type of audit opinion to be issued are particularly sensitive in the case of a financially stressed client. If auditors defer to the wishes of client management, they would be less likely to issue going-concern modified audit opinions, since audit clients do not welcome the receipt of a going-concern modified audit opinion (Kida 1980; Mutchler 1984). Further, the issuance of a going-concern modified audit opinion involves costs to both the auditor and the client (Kida 1980; Mutchler 1984; Geiger et al. 1998; Blay and Geiger 2001; Weil 2001). Receiving a going-concern modified audit opinion can lead to adverse consequences for the client in terms of stock price declines (Loudder et al. 1992; Blay and Geiger 2001; Weil 2001) and increased risk of business failure (Geiger et al. 1998). Auditors have long been cognizant of the possibility of losing a disgruntled client after issuing them a going-concern modified opinion (Kida 1980; Mutchler 1984). Prior research has documented that auditors are less likely to issue a going-concern modified audit opinion for large companies after controlling for other factors (McKeown et al. 1991; Mutchler et al. 1997). Audit fee research has also documented client size is an important determinant of audit fees (Simunic 1980, 1984; Francis 1984), and recent research indicates that the relative magnitude of nonaudit fees also is higher for larger clients (Abbott et al. 2002). Together, these results suggest that audit opinions may be influenced by the magnitude of nonaudit (and audit) fees received from clients. Some recent studies have examined the association between nonaudit fees and clientís account- ing accruals or their meeting analystís forecasts, arguing that accruals and earnings figures are a proxy for auditor judgments about client actions. However, the results of these studies have been mixed. For example, Frankel et al. (2002) find that nonaudit fees are positively associated with accruals, but Ashbaugh et al. (2002), Chung and Kallapur (2002), and Reynolds et al. (2002) do not find such an association. The Frankel et al. (2002) study finds a positive association between nonaudit fees and meeting analyst earnings forecasts, while the studies of Ashbaugh et al. (2002) and Francis and Ke (2002) do not. One disadvantage of focusing on accounting accruals or meeting earnings forecasts is that auditors have only an indirect effect on these client-oriented outcomes. In contrast, auditorsí decisions and actions directly affect the type of audit opinion they issue. The relationship between NAS fees and auditor reporting behavior has been examined in Aus- tralia where NAS fees have been publicly available for some time. However, the results of such research are contradictory. Wines (1994) finds evidence of a negative association between nonaudit fees and the issuance of any type of qualified report in a sample of 76 Australian companies for the period 1980ñ89. However, Barkess and Simnett (1994) and Craswell (1999) in two large sample studies find no significant association between the level of NAS fees and audit report qualifications. Most recently, Sharma (2001) and Sharma and Sidhu (2001) examined 49 bankrupt companies and conclude that higher NAS fees were negatively associated with going-concern modified reports. These disparate results make it difficult to make strong inferences regarding the affect of NAS levels on auditor reporting in Australia. Additionally, transference of these overseas results to the U.S.

1 Mutchler et al. (1997) and Behn et al. (2001) find that decisions about going-concern modified opinions are related to the evaluation of managementís plans and mitigating circumstances.

Auditing: A Journal of Practice & Theory, September 2003

Geiger and Rama

57

audit market is problematic in that U.S. auditors might respond differently to litigation and reputa- tion concerns than Australian auditors causing U.S. auditors to report differently when faced with similar NAS fee issues. In a contemporaneous U.S. study, DeFond et al. (2002) conclude that there is no association between audit and NAS fees and audit opinions. However, given their large sample size and the difficulty of obtaining default and other nonquantitative data, DeFond et al. (2002) do not include default status, management plans, or audit committee composition as control factors in their study. These variables have been shown by prior researchers to be associated with audit opinions in the U.S. Given our smaller sample and the use of the matched-pair approach, we are able to control for these additional nonfinancial audit report-related factors in our analysis. Further, a recent study by Geiger and Raghunandan (2002) finds that the likelihood of a finan- cially stressed company receiving a going-concern modified report declined in the U.S. during the 1990s. They note that this empirical regularity may be due to the change in the nature of the liability regime facing U.S. auditors due to the enactment of the Private Securities Litigation Reform Act in 1995 and the Securities Litigation Uniform Standards Act in 1998. However, an alternative, but not mutually exclusive, explanation of the decline in going-concern modified reports is that the relative importance of nonaudit fees to the accounting firms has increased significantly in recent years. Hence, a plausible suggestion is that the propensity to issue going-concern modified audit opinions declined because of the increase in the relative magnitude of nonaudit fees. Accordingly, a direct test of the influence of the level of NAS fees on auditor independence would be to assess audit opinions on financially stressed companies and examine how they vary with relative levels of NAS fees.

Research Questions Audit opinions for financially stressed companies have been an ongoing issue of concern to legislators and the public, and have been the subject of the business press (Weil 2001; Coffee 2002; Dietz 2002) and several congressional hearings and Bills over the years (U.S. House of Representa- tives 1985, 1990, 2002; Carmichael and Pany 1993). Further, the NASDAQ has recently required companies listed on their exchange to disclose the receipt of a GCM opinion in a separate news release to the financial press (NASDAQ 2002). The SEC (2000a, 2000b) has asserted that auditor judgments, particularly those that are highly subjective, may be influenced by the magnitude of nonaudit fees. As discussed earlier, audit reporting related to going concern is an area that involves considerable auditor judgment. Hence, it is worth empirically examining whether nonaudit services have an impact on audit opinions issued for companies in financial stress. This leads to the first research question examined in this paper (in the null form):

RQ 1 :

There is no association between the magnitude of nonaudit fees and audit opinions received by financially stressed companies.

Auditors and others have noted that the magnitude of nonaudit fees cannot be viewed separately from the level of audit fees received from clients (SEC 2000b). Specifically, if the argument is that the fees paid by clients compromise audit judgments, the same argument can be applied to both audit fees and nonaudit fees. Consequently, we also examine the association between audit fees and audit opinions. This leads to the second research question examined in this paper (in the null form):

RQ 2 :

There is no association between the magnitude of audit fees and audit opinions received by financially stressed companies.

Auditing: A Journal of Practice & Theory, September 2003

58

Auditing, September 2003

METHOD We use the following logistic model to examine our research questions:

GC = a 0 + a 1 *SIZE + a 2 *PROB + a 3 *DFT + a 4 *DEBT + a 5 *EQUITY+ a 6 *COSTRED

+ a 7 *ASSETSALE + a 8 *AC + a 9 *REPORTLAG + a 10 *BIG5 + a 11 *AUDFEE

+ a 12 *NASFEE.

The variables are defined as follows:

GC

=

1 if audit opinion was modified for going concern, else 0;

SIZE

=

natural log of total assets (in millions of dollars);

PROB =

probability of bankruptcy from Hopwood et al.ís (1994) model;

DFT

=

1 if the company was in default, else 0;

DEBT

=

1 if company issued new debt during the year, else 0;

EQUITY

=

1 if company issued new equity during the year, else 0;

COSTRED

=

1 if company entered into significant cost reduction program, else 0;

ASSETSALE =

1 if company announced sales of significant assets, else 0;

AC

=

1 if company exhibited a strong audit committee, else 0;

REPORTLAG =

number of days from the end of the year to the audit report date;

BIG5

=

1 if audited by a Big 5 firm, else 0;

AUDFEE

=

natural log of audit fee; and

NASFEE

= natural log of nonaudit fees.

In the model above, AUDFEE and NASFEE are the variables of interest. Consistent with prior research, we use natural log of audit and nonaudit fees in our logistic regression (DeFond et al. 2002). 2 We include the following types of control factors in our analysis: (1) financial variables, (2) management plans and mitigating factors, (3) audit committee type, (4) audit firm size, and (5) audit reporting lag. We explain below the rationale for including the specific variables. Prior research suggests that there is (1) a positive association between the likelihood of a going- concern modified audit opinion, and financial stress and default on debt obligations, and (2) a negative association between the likelihood of a going-concern modified audit opinion and company size (Mutchler 1985; Hopwood et al. 1989, 1994; Chen and Church 1992; Carcello et al. 1995; Geiger and Raghunandan 2001). Hence, we include company size (SIZE), financial stress (PROB), and default status (DFT) as control factors. We measure client size (SIZE) using log of total assets (in millions of dollars) 3 and financial stress (PROB) using the coefficients given in Hopwood et al. (1994). 4 We classify a company as in default (DFT) if it is either in payment default or technical default of loan covenants. SAS No. 59 requires auditors to evaluate management plans when there is substantial doubt about the ability of an entity to continue as a going concern. Behn et al. (2001) found that plans to issue equity or debt were negatively associated with the likelihood of receiving a going-concern opinion. Conversely, Behn et al. (2001) note that auditors may view plans to significantly reduce spending or dispose of assets negatively. In our analysis, we include control factors related to (1) plans to issue debt (DEBT) or equity (EQUITY), (2) initiate major cost-reduction efforts ( COSTRED), or (3) dispose of significant assets (ASSETSALE).

2 As discussed in the ìAdditional Analysesî section, use of alternative measures of audit and NAS fees do not substantially alter the results related to either fee variable.

3 If we use natural log of sales, square root of sales, raw net sales, raw total assets, or square root of total assets as alternative specifications of company size, our results do not differ substantively from those presented.

4 Geiger and Raghunandan (2001) note that the correct value of the intercept provided in of Hopwood et al. (1994, Table 3) should be ñ7.322 (as opposed to 5.565). We use the Hopwood et al. (1994) model to be consistent with prior research in this area. As part of sensitivity tests, we also used the model from Zmijewski (1984) to calculate the probability of bankruptcy and obtained similar results to those presented.

Auditing: A Journal of Practice & Theory, September 2003

Geiger and Rama

59

The Blue Ribbon Committee (BRC 1999) and the SEC (1999), among others, have noted that a strong audit committee can strengthen the auditor in negotiations with management. Since audited financial statements and the audit report are ultimately the products of negotiation between the auditor and the client, it is likely that a strong audit committee would make it easier for the auditor to issue a GCM opinion. Along these lines, Carcello and Neal (2000) found that a GCM was more likely to be issued to financially stressed companies that also had an audit committee made up of solely outside members. Hence, we include audit committee (AC) composition as another control variable. Based on the recommendations of the BRC and the SEC (as also reflected in the subsequent rule changes by the NYSE, AMEX, and NASDAQ about listing requirements related to audit committees), we use two criteria to classify an audit committee: (1) if the audit committee had solely independent directors, and (2) the audit committee had at least one member with financial expertise. We consider companies exhibiting both of these criteria to have strong audit committees. 5 Prior research suggests that audit reporting lag (i.e., the time between the companyís fiscal year- end and the date of the audit report) is associated with the type of audit report given to financially stressed companies (Chen and Church 1992; Carcello et al. 1995; Behn et al. 2001). Going-concern modified opinions typically take longer for the auditor to issue than nonmodified audit reports. Thus, we include an audit report lag variable (REPORTLAG) to control for the timeliness of audit opinions on stressed companies. Audit opinions may also be associated with auditor type. Specifically, Big 5 auditors may have more to lose from litigation related to a particular audit and hence may be more conservative in their reporting decisions. Conversely, even when considering only SEC audit clients, the Big 5 derive a much higher proportion of their revenues from nonaudit services than do the non-Big 5 firms (SEC 2000a). Hence, the association between audit firm size and audit opinion is an empirical question. We added a proxy for auditor size by including a BIG5 dummy variable in our regression.

DATA We first identify a group of firms that received a GCM audit opinion. Consistent with prior research, we restrict our analysis to companies that received a first-time GCM opinion. As noted by Kida (1980) and Mutchler (1984), deciding to render an initial going-concern modified opinion to an audit client is particularly difficult for the auditor. Issuing a modified opinion in subsequent years is less risky for the auditor and constitutes a different decision model that may not include the per- ceived risk of losing a disgruntled client (Geiger et al. 1998; Carcello and Neal 2003). The new SEC rules require companies to include data about audit and nonaudit fees in proxy statements filed on or after February 5, 2001. Hence, we identified companies with fiscal year-ends between September 30, 2000 and February 28, 2001 from the September 2001 version of the Compact Disclosure-SEC (CD-SEC) database. This was done to maximize the likelihood of obtain- ing the relevant fee data in subsequent proxy statements filed with the SEC. Consistent with prior studies (Mutchler and Williams 1990; Behn et al. 2001; Blay and Geiger 2001) we restricted our analysis to companies in the manufacturing sector (SIC 20 to 39) to eliminate confounding industry effects and to keep the financial ratios consistent with the development of the financial stress models. After identifying companies receiving a going-concern modified audit opinion during our sample period from the CD-SEC database, we then collected data on the financial variables, the audit opinion, and the auditor. Next we eliminated companies receiving a going-concern modified opinion in the prior year. We then searched the EDGAR database and confirmed data about the auditor and opinion and obtained information regarding default status and managementís plans from the relevant 10-Ks filed with the SEC.

5 As noted in the ìAdditional Analysesî section, alternative specifications of this audit committee variable do not substan- tively alter the results presented.

Auditing: A Journal of Practice & Theory, September 2003

60

Auditing, September 2003

We obtained the audit and nonaudit fee data, and audit committee information from proxy statements. Not all public companies are required to file a proxy statement, and, although required by the new SEC rules, some companies did not include the audit and nonaudit fee data, or information on their audit committees in their post-February 5, 2001 proxy statements. Results of our screening and data requirements yielded 66 companies receiving a first-time going-concern modified audit opinion (i.e., the GCM sample) that had all necessary financial, nonfinancial, audit committee, and fee data.

Control Sample Prior researchers have noted that auditors do not generally issue going-concern modified audit opinions for nonstressed companies that suddenly fail (McKeown et al. 1991; Hopwood et al. 1994). Hopwood et al. (1994, 412) noted that ìinvestigations of auditorsí going-concern opinion decisions should be conducted on samples that have been partitioned into stressed and nonstressed categoriesî because ìnonstressed, bankrupt companies are likely to have experienced management fraud leading to misstated financial statements.î Consistent with prior research (McKeown et al. 1991; Hopwood et al. 1994), we classify a company as being in financial stress if it exhibited at least one of the following financial stress signals: (1) negative working capital at the end of the year, (2) negative retained earnings at the end of the year, and (3) bottom line loss for the year. 6 We use a matched-pair design for our analysis. A matched-pair design affords a better selection of a control sample, and reduces the difficulty of collecting data on management plans and other mitigating factors (required to be evaluated by the auditor under SAS No. 59), audit committee, and the related firm-specific variables considered in our analyses. We identified a sample of financially stressed manufacturing companies (based on the stress criteria discussed previously) that did not receive a going concern-modified audit opinion (i.e., the NGC sample) from the same CD-SEC database. We then collected data on the financial statement variables and calculated the companyís bankruptcy probability based on the Hopwood et al. (1994) model (PROB). We used these probabilities of bankruptcy along with net sales data and SIC code to match these stressed, non-going-concern companies (NGC) to the first-time going-concern modified companies (GCM). We matched the NGC control sample companies to the GCM experimental sample based first on PROB, next on size (in net sales dollars), and finally on two-digit SIC classifications. If audit and NAS fee data were not available for the selected NGC company, we replaced it with the next best available match for which all data were available. This procedure ensured that we included similar companies with respect to financial stress and size in our experi- mental and control samples of manufacturing companies.

RESULTS Table 1 provides descriptive data about the samples. The univariate results indicate that the GCM companies were more likely to be in default, to have entered into a significant cost reduction plan, and to have plans to sell significant assets. The GCM companies were also less likely to have been audited by a Big 5 accounting firm and had longer reporting lags. The nonsignificant differ- ences between the two groups with respect to net sales and total assets, as well as the nonsignificant difference in the probability of bankruptcy measure (PROB) lend some evidence to the efficacy of our matching procedures. The audit fees and nonaudit fees are not significantly different between the two groups in the univariate analyses. Additionally, the GCM sample was more likely to have paid no NAS fees to their auditor than the NGC sample companies.

6 McKeown et al. (1991) and Hopwood et al. (1994) classify a company as being stressed if the financial stress signals were exhibited in the fiscal year of the audit opinion being examined or in the preceding two years. We use only the current year financial measures; hence, our sample selection uses a more stringent screen for financial stress.

Auditing: A Journal of Practice & Theory, September 2003

Geiger and Rama

61

Table 1 indicates that the companies in our sample are relatively small (average sales of $117.0 and $118.5 million for the GCM and NGC groups, respectively) with relatively small audit fees (average of $211.7 and $169.5 thousand for the GCM and NGC groups, respectively) and NAS fees (average of $171.4 and $238.4 thousand for the GCM and NGC groups, respectively). This may be partially due to our sample selection procedures that began with identifying GCM companies. McKeown et al. (1991) and Hopwood et al. (1994) document a GCM opinion size bias in favor of not issuing going-concern modified reports to large companies. This size bias could have limited the size of the companies included in our GCM experimental sample, and therefore also in our matched control sample. The small size of the companies in our sample raise the question of whether we have

TABLE 1

Descriptive Data

Going-Concern Sample (n = 66)

Non-Going-Concern Sample (n = 66)

 

Mean

 

Mean

Range

Mean

Range

Difference

Variable

(std. dev.)

[median]

(s.d.)

[median]

p-value

Net Sales (in millions)

117.0

.001ñ2,249.9

118.5

.740ñ873.3

.974

(229.9)

[18.3]

(202.5)

[26.5]

Total Assets

108.8

.191ñ2,343.0

128.2

1.38ñ908.0

.678

(in millions)

(311.2)

[22.3]

(212.9)

[38.0]

PROB

.70

.01ñ1.00

.71

.01ñ1.00

.999

(.41)

[.99]

(.41)

[.99]

DFT (%) DEBT (%) EQUITY (%) COSTRED (%) ASSETSALE (%) AC (%) Big 5 (%) REPORTLAG (in days)

47

ó

23

ó

.003

30

ó

39

ó

.277

26

ó

35

ó

.259

44

ó

14

ó

.001

33

ó

9

ó

.001

61

ó

62

óñ

.859

67

ó

91

óñ

.001

72

12ñ156

54

21ñ102

.001

 

(29)

[70]

(21)

[47]

Audit fee ($000)

211.7

26ñ2,551

169.5

30ñ800

.357

(347.6)

[107]

(128.2)

[129]

Nonaudit fee ($000)

171.4

0ñ2,708

238.4

11ñ1,855

.321

(426.9)

[38.7]

(340.8)

[90.4]

Companies w/ no NAS fees (%)

17

ó

0

ó

.001

PROB = Probability of bankruptcy from Hopwood et al.ís (1994) model; DFT = 1 if the company was in default, else 0; DEBT = 1 if company issued new debt, else 0; EQUITY = 1 if company issued new equity, else 0; COSTRED = 1 if company entered into significant cost reduction program, else 0; ASSETSALE = 1 if company announced sales of significant assets, else 0; AC = 1 if company exhibited a strong audit committee, else 0; BIG5 = 1 if audited by a Big 5 firm, else 0; REPORTLAG = number of days between the year-end and the audit report date; Audit fee = fees paid by the company for the audit; and Non-audit fee = fees paid by the company for nonaudit services.

62

Auditing, September 2003

ìsignificantî NAS fees included in our analysis. However, as noted earlier, the stressed companies in our sample are of keen interest to legislators, regulators, and auditors and are germane to the examination of NAS fees on auditor reporting decisions. Results from the multivariate logistic regression are presented in Table 2. 7 The overall model is significant (Chi-square = 66.704, 12 d.f., p < .0001). The coefficient for the PROB control factor is

not significant in the multivariate analysis (p = .256), lending further evidence of the efficacy of our matching procedures. However, consistent with prior research, the SIZE variable indicates a contin- ued company size bias (p = .008) in favor of larger companies not receiving a GCM opinion from their auditor. The DFT variable is positive and significant (p = .050) indicating companies in default were more likely to get a going concern modified audit opinion. Consistent with Behn et al. (2001), companies in our study are more likely to receive a modified report if they entered into a cost reduction plan (COSTRED; p = .003) or entered into the sale of significant assets (ASSETSALE; p

= .057) and less likely to receive a modified report if they demonstrated the ability to issue new

securities. However, plans for the issuance of new debt (DEBT; p = .291) or equity (EQUITY; p

= .325) securities, while negatively associated with receiving a going-concern modified report, are

not significantly associated with opinion type. The coefficients for the audit committee variable (AC; p = .675) and reporting lag variable (REPORTLAG; p = .250) are consistent with prior research, but are not significant. 8 The BIG5 variable is negative but not significant (p = .126), indicating no significant Big 5 audit firm effect. The two variables of interest, AUDFEE and NASFEE, indicate that magnitude of the audit fee paid was positively and significantly (p = .043) related to receiving a going-concern audit opinion, but no significant NAS fee and audit report type association was found for these distressed compa- nies (p = .215). Our results support earlier audit fee studies that find a positive audit fee and report modification association (e.g., Simunic 1980; Palmrose 1986; Francis and Simon 1987; Bell et al. 2001). Our results, however, do not indicate a significant adverse effect of NAS fees on auditor reporting decisions on the stressed companies in our study. These findings are further supported by the consistent results obtained and reported in the following section using numerous audit and NAS fee, control variable, and model specifications.

Further Analyses The SEC (2000a, 2000b) has also expressed concern that the relative level of NAS fees com- pared to audit fees derived from the same company is of concern. If auditors receive proportionately more NAS fees than audit fees, auditors may be more swayed to see it the clientís way in order to preserve the more lucrative NAS fees. Accordingly, similar to Firth (1997) we examined the ratio of NAS fees to audit fees and its relationship with audit opinions. Replacing the NASFEE and AUDFEE variables in the logistic regression with the ratio of NAS to audit fees, the coefficient on the ratio variable is positive but not significant (p = .383). Additionally, the SEC indicated a primary concern with ìsignificantî NAS fees clouding auditorís judgments. In an attempt to address ìsignificantî NAS fee companies, we limited the analysis to those companies with higher than the median NAS fee. For these above the median companies, the regression coefficient for the AUDFEE variable is positive but not significant (p = .168), and the NASFEE variable is also positive but not significant (p = .998). These results indicate that limiting

7 Consistent with DeFond et al. (2002) and due to the exploratory nature of this research, we report all model coefficient p- values as two-tail probabilities.

8 These results for the AC variable are in contrast to the findings of Carcello and Neal (2000) that audit committee composition is related to going-concern reporting. One possible explanation is the data in Carcello and Neal (2000) come from the early 1990s and there have been many subsequent initiatives by the SEC, the Independence Standards Board, and the major stock exchanges regarding strengthening audit committee composition. This may have lead to more consistent types of audit committees across the GCM and control group companies.

Auditing: A Journal of Practice & Theory, September 2003

Geiger and Rama

63

Model: GC

TABLE 2 Logistic Regression Results (n = 132)

= b 0 + b 1 *SIZE + b 2 *PROB + b 3 *DFT + a 4 *DEBT + a 5 *EQUITY + a 6 *COSTRED

+ a 7 *ASSETSALE + a 8 *AC + a 9 *REPORTLAG

+ a 10 *BIG5 + a 11 *AUDFEE + a 12 *NASFEE

Variable

Coefficient

Chi-Square

p-value*

Intercept

ñ2.159

0.318

.573

SIZE

ñ0.691

6.970

.008

PROB

ñ0.746

1.289

.256

DFT

1.032

3.847

.050

DEBT

ñ0.550

1.118

.291

EQUITY

ñ0.558

0.970

.325

COSTRED

1.603

8.645

.003

ASSETSALE

1.261

3.628

.057

AC

0.218

0.176

.675

REPORTLAG

0.012

1.327

.250

BIG5

ñ1.069

2.343

.126

AUDFEE

0.887

4.111

.043

NASFEE

ñ0.132

1.537

.215

Model Chi-square = 66.704, 12 d.f., p < .0001; c-stat = .87; Pseudo R 2 = .365

* Two-tailed significance levels. =

SIZE

PROB

=

natural log of total assets (in millions of dollars); probability of bankruptcy from Hopwood et al.ís (1994) model;

DFT

=

1 if the company was in default, else 0;

DEBT

=

1 if company issued new debt, else 0;

EQUITY

=

1 if company issued new equity, else 0;

COSTRED

=

1 if company entered into significant cost reduction program, else 0;

ASSETSALE

=

1 if company announced sales of significant assets, else 0;

AC

=

1 if company exhibited a strong audit committee, else 0;

REPORTLAG

=

number of days between the year-end and the audit report date;

BIG5

=

1 if audited by a Big 5 firm, else 0;

AUDFEE

=

natural log of audit fee; and

NASFEE

=

natural log of nonaudit fees.

the analysis to companies with relatively higher magnitudes of NAS fees reveals no significant association of fees paid to the auditor and report type for our sample. 9 We also separately examined the companies audited by Big 5 audit firms. Results of this modified sample are similar to those presented in Table 2, with the AUDFEE variable positive and significant (p = .062) and the NASFEE variable not significant (p = .337). To ensure that clients of

9 If we limit our analysis to the 37 companies whose ratio of NASFEE to AUDFEE is greater than 1 (i.e., those that paid their auditors more in NAS fees than audit fees), we also find no association between fees and audit reports. For these companies, the coefficients on the NASFEE and AUDFEE variables were both not significant (p = .846 and .491, respectively), as well as the NASFEE/AUDFEE ratio (p= .484) in the modified regressions.

Auditing: A Journal of Practice & Theory, September 2003

64

Auditing, September 2003

any single Big 5 audit firm were not driving the results, we formed five subsamples by deleting clients of each individual Big 5 firm. In addition, we also included indicator variables for each of the Big 5 firms in the full-sample regression. In each of these modified regressions, the significance of the AUDFEE and NASFEE variables were substantively the same as those presented in Table 2. To assess possible interaction effects with our NASFEE variable, we created five interaction terms for the NASFEE variable and the DEBT, EQUITY, COSTRED, ASSETSALE, and BIG5 vari- ables. Model results when including these interaction terms indicate a positive but not significant AUDFEE variable (p = .167) and a continued nonsignificant NASFEE variable (p = .310), with none of the interaction terms being significant (p > .10). To ensure that our results were not driven by the severe financial stress of our sample firms, we examined only those companies with probabilities of bankruptcy less than .75. 10 It might be argued that these less stressed companies afford the auditor more reporting latitude and may be more likely influenced by NAS fees. The results of this reduced sample are similar to those presented in Table 2, with the AUDFEE variable being positive and significant (p = .054) and the NASFEE variable not significant (p = .349). We performed several sensitivity analyses by using alternative measures for size and financial stress. Specifically, we used log of sales, raw total assets or sales, or square root of total assets or sales as alternative size measures and used the probability of bankruptcy measure based on Zmijewskiís (1984) model. We also used the raw NAS and audit fees, as well as these fees divided by total assets, square root of total assets, and log of total assets to test the sensitivity of our fee measures. The results remain essentially the same as those presented in Table 2óthe NASFEE variable was not significant and the AUDFEE variable was positive and significant (p < .10) in each of these alterna- tive variable specifications. We created alternative specifications for audit committee strength if: (1) the audit committee was comprised solely of outside directors, and (2) the audit committee was comprised solely of outside directors, had at least one person with financial expertise, and met at least four times throughout the year. Results of these alternative specifications were consistent with the nonsignifi- cant results reported in Table 2 for the AC variable. We also reran the model replacing the AC variable with three separate audit committee variables: (1) percentage of outside committee mem- bers, (2) presence of at least one member with financial expertise, and (3) number of committee meetings for the year. None of the three variables were significant (p > .20), and the other variables, in particular the NASFEE and AUDFEE variables, remained largely unchanged.

Endogeneity Issues Auditors and others have noted that audit fees and nonaudit fees may be jointly determined (Simunic 1984; Palmrose 1986; Davis et al. 1993). In fact, the CEOs of the AICPA and Deloitte & Touche have recently asserted that the joint performance of audit and nonaudit services leads to more effective and efficient audits (Melancon 2002; Copeland 2002). Accordingly, we test whether our results are misspecified due to the possible endogeneity of the audit opinion with audit fees and nonaudit fees. We used our logistic regression model as part of a system of equations that included the following two additional fee models:

AUDFEE

=

a 0 + a 1 *SIZE + a 2 *CATA +

a 3 *DEBT + a 4 *EQUITY

+ a 5 *ACQ

 

+

a 6 *BIG5 + a 7 *NUMSUBS + a 8 *FORSUB + a 9 *GC + a 10 *NASFEE

NASFEE

=

a 0 + a 1 *SIZE + a 2 *DIRPERCT + a 3 *INSTPERCT + a 4 *TLTA + a 5 *ROSE

+ a 6 *MKTRET

+

+ a 7 *BIG5 + a 8 *EQUITY + a 9 *NEWCEO +a 10 *RESTROPER

a 11 *ACQ + a 12 *AUDFEE

where the new variables are defined as follows:

10 There were 37 GCM and 38 NGC companies with bankruptcy probabilities (Zmijewski 1984) less than .75.

Auditing: A Journal of Practice & Theory, September 2003

Geiger and Rama

65

CATA

= current assets divided by total assets;

ACQ

= 1 if company acquired another company during the year, else 0;

NUMSUBS

= square root of the number of subsidiaries;

FORSUB

= 1 if company had a foreign subsidiary, else 0;

DIRPERCT

= director and insider ownership percentage;

INSTPERCT

= institutional ownership percentage;

TLTA

= total liabilities divided by total assets;

ROSE

= return on stockholders equity in prior year;

MKTRET

= industry adjusted prior-year stock return;

NEWCEO

= 1 if company had a new CEO, else 0; and

RESTROPER = 1 if company restructured operations during the year, else 0.

The variables for our fee models are based on variables found in prior research to be significant in modeling audit and NAS fees (Simunic 1980, 1984; Francis 1984; Palmrose 1986; Francis and Simon 1987; Beatty 1993; Barkess and Simnett 1994; Firth 1997; Craswell 1999; Lennox 1999). 11, 12 Using a two-stage system of simultaneous equations 13 indicates that audit opinions were endog- enous with audit and NAS fees in our sample. 14 However, correcting for endogeneity by substituting the fitted AUDFEE and NASFEE variables from these models into our GC model produces results that are essentially the same as those presented in Table 2 (Woolridge 2000). Specifically, the fitted AUDFEE variable is positive and significant (p < .0001) while the fitted NASFEE variable is not significant (p = .965). 15 Thus, while we do find that audit opinions, audit fees and NAS fees are jointly determined, we conclude that this endogeneity does not substantially affect the results con- cerning the relationship between GC modifications and fees presented in the paper.

SUMMARY AND CONCLUSIONS The SEC has recently expressed concern that when auditors receive nonaudit service (NAS) fees from their audit clients, audit decisions may be affected. The Commission noted that subtle influences are particularly likely to be prevalent in the gray areas that require significant auditor judgment. In addition, the recently enacted Sarbanes-Oxley Act of 2002 prohibits auditors from supplying certain types of nonaudit services to their audit clients and imposes additional restrictions on nonaudit purchases from auditors. This study examined the association between nonaudit fees and audit reporting for financially stressed companies. Audits of financially stressed companies are the types of engagements in which auditors must exercise a significant amount of professional judgment. Our analysis of audit opinions on companies in financial stress does not find evidence of a significant association between NAS fees and the likelihood of a company receiving or not receiving a going-concern modified audit report. Our sample of financially stressed companies was fairly small in size and paid relatively little to their auditors in the way of audit and NAS fees. Accordingly, our nonsignificant NAS results present a weak test of the effect of NAS fees on auditor decision

11 The AUDFEE model results were robust to the inclusion of an audit committee variable, and replacement of the foreign subsidiary dummy variable with a variable for the number of foreign subsidiaries and a percentage of foreign subsidiaries to total subsidiaries variable.

12 The NASFEE model results were robust to the inclusion of a CEO tenure variable, use of a current year stock return measure in place of the prior year measure, use of a raw stock return measure in place of the industry-adjusted returns measures, inclusion of the audit committee variable, inclusion of a current assets-to-total assets ratio, and replacement of the foreign subsidiaries dummy variable with a variable for the percentage of foreign subsidiaries.

13 See Copley et al. (1994) for an example of the application of a two-stage simultaneous equations approach in an audit fee setting.

14 Both fee models were significant (p < .0001) and results of Hausman tests (Hausman 1978; Woolridge 2000) were also significant (p < .05) for both audit and NAS fees. This is consistent with recent evidence regarding NAS and audit fee determination in Whisenant et al. (2002).

15 Using a PROBIT regression model in place of the logistic regression model produces essentially identical results.

Auditing: A Journal of Practice & Theory, September 2003

66

Auditing, September 2003

making. However, our decision setting, as discussed earlier, is still very germane to perceptions of audit independence issues, and audit reporting on distressed companies has over the years continued to receive disproportionate attention from legislators, the financial press, and financial statement users and the consistent results obtained from the voluminous number of analyses performed in this study give credibility to our results. The results reported in this paper reinforce those of the DeFond et al. (2002) large sample study that finds no association between NAS fees and audit reports. Our results using a different method and sample are consistent with their position that market-based incentives, such as losses from possible litigation and possible damage to reputation, may be strong factors in influencing U.S. auditorsí reporting decisions in the presence of NAS fees. However, while DeFond et al. (2002) find no association of audit fees on auditor reporting, we find a positive association; this finding is consistent with prior audit fee research that finds a positive association between audit fees, audit effort, and the likelihood of receiving a modified audit opinion (Kida 1980; Francis 1984; Francis and Simon 1987; Palmrose 1986; Beatty 1993). Differences in audit fee results between these two studies are likely to be due to differences in sample selection procedures and differing sets of control variables used. In this study we have focused only on the final audit reporting decision on financially stressed companies. We have not examined the subsequent consequences of the reporting decision, namely ìtype I errorsî (a going-concern modified opinion issued for a client that is subsequently viable) and ìtype II errorsî (a clean opinion issued for a client that subsequently fails). A fruitful area for future research would be to examine if the two types of audit reporting ìerrorsî are also associated with audit and nonaudit fees. We have also not attempted to assess how NAS fees from one client might affect individual audit partners under different terms of compensation from their firms (Trompeter 1994). Assessment of the effects of NAS fees on audit decisions in the context of partner compensa- tion would be an interesting avenue for future research. Finally, we have assessed the audit reporting decision only in a limited context and on a limited sample of companies. The examination of the possible effects of NAS fees on other auditor judgments made in the course of an audit (e.g., client acceptance and retention decisions, decisions regarding the extent of controls testing or the proper accounting treatment of questionable transactions), and in cases involving larger amounts of NAS fees paid to the auditor, would extend our knowledge on the effects of auditors providing NAS to audit clients.

REFERENCES

Abbott, L. J., S. Parker, G. Peters, and K. Raghunandan. 2002. An empirical investigation of audit fees, non- audit fees, and audit committees. Working paper, University of Memphis, Tennessee. American Institute of Certified Public Accountants (AICPA). 1988. The Auditorís Consideration of an Entityís Ability to Continue as a Going Concern. Statement on Auditing Standards No. 59. New York, NY:

AICPA. Ashbaugh, H., R. LaFond, and B. Mayhew. 2002. Do non-audit services compromise auditor independence? Further evidence. Working paper, University of WisconsinñMadison. Barkess, L., and R. Simnet. 1994. The provision of other services by auditors: Independence and pricing issues. Accounting and Business Research (Spring): 99ñ108. Beatty, R. P. 1993. The economic determinants of auditor compensation in the initial public offerings market. Journal of Accounting Research (Autumn): 294ñ302. Behn, B. K., S. E. Kaplan, and K. R. Krumwiede. 2001. Further evidence on the auditorís going-concern report: The influence of management plans. Auditing: A Journal of Practice & Theory (March): 13ñ28. Bell, T. B., W. R. Landsman, and D. A. Shackelford. 2001. Auditorsí perceived business risk and audit fees:

Analysis and evidence. Journal of Accounting Research (June): 35ñ43. Blay, A. D., and M. A. Geiger. 2001. Market expectations for first-time going-concern recipients. Journal of Accounting, Auditing & Finance (Summer): 209ñ226.

Auditing: A Journal of Practice & Theory, September 2003

Geiger and Rama

67

Blue Ribbon Committee (BRC) on Improving the Effectiveness of Corporate Audit Committees. 1999. Report and Recommendations of the Blue Ribbon Committee on Improving the Effectiveness of Corporate Audit Committees. New York, NY: The New York Stock Exchange and the National Association of Securities Dealers. Carcello, J. V., D. Hermanson, and F. Huss. 1995. Temporal changes in bankruptcy-related reporting. Audit- ing: A Journal of Practice & Theory (Fall): 133ñ143. óóó, and T. L. Neal. 2000. Audit committee composition and auditor reporting. The Accounting Review (October): 453ñ468. óóó, and óóó. 2003. Audit committee characteristics and auditor dismissals following ìnewî going- concern reports. The Accounting Review 78 (January): 95ñ117. Carmichael, D. R., and K. Pany. 1993. Reporting on uncertainties, including going concern. In The Expecta- tion Gap Standards: Progress, Implementation Issues, and Research Opportunities. Jersey City, NJ:

AICPA. Chen, K. C. W., and B. K. Church. 1992. Default on debt obligations and the issuance of going-concern opinions. Auditing: A Journal of Practice & Theory (Fall): 30ñ49. Chung, H., and S. Kallapur. 2002. Client importance, non-audit fees, and abnormal accruals. Working paper, Purdue University, West Lafayette, Indiana. Coffee, J. C. 2002. Guarding the gatekeepers. New York Times (May 13): B3. Copeland, J. 2002. Statement of James Copeland. U.S. Senate Committee on Banking, Housing and Urban Affairs (March 14). Washington, D.C: Government Printing Office. Copley, P. A., M. S. Doucet, and K. M. Gaver. 1994. A simultaneous equations analysis of quality control review outcomes and engagement fees for audits of recipients of federal financial assistance. The Accounting Review (January): 244ñ256. Craswell, A. T. 1999. Does the provision of non-audit services impair auditor independence? International Journal of Auditing (3): 29ñ40. Davis, L. R., N. Ricchiute, and G. Trompeter. 1993. Audit effort, audit fees, and the provision of non-audit services to audit clients. The Accounting Review (January): 135ñ150. DeAngelo, L. 1981. Auditor independence, ìlow balling,î and disclosure regulation. Journal of Accounting and Economics (December): 113ñ127. DeFond, M., K. Raghunandan, and K. R. Subramanyam. 2002. Do non-audit services impair auditor indepen- dence? Evidence from going-concern audit opinions. Journal of Accounting Research (September):

1247ñ1274.

Dietz, D. 2002. ìAuditors are timidî they failed to warn in most big firm bankruptcies since 1996. Pittsburgh Post-Gazette (April 26): C8. Firth, M. 1997. The provision of non-audit services by accounting firms to their clients. Contemporary Accounting Research (Summer): 1ñ21. Francis, J. 1984. The effect of audit firm size on audit prices: A study of the Australian market. Journal of Accounting & Economics (2): 133ñ151. óóó, and D. Simon. 1987. A test of audit pricing in the small-client segment of the U.S. audit market. The Accounting Review (January): 145ñ157. óóó, and B. Ke. 2002. Further evidence on non-audit services and auditor independence. Working paper, University of MissouriñColumbia. Frankel, R., M. Johnson, and K. Nelson. 2002. Auditor independence and earnings quality. Working paper, Michigan State University, East Lansing. Geiger, M. A., K. Raghunandan, and D. V. Rama. 1998. Costs associated with going-concern modified audit opinions: An analysis of auditor changes, subsequent opinions, and client failures. Advances in Account- ing 16: 117ñ139. óóó, and óóó. 2001. Bankruptcies, audit reports, and the Reform Act. Auditing: A Journal of Practice & Theory (March): 187ñ195. óóó, and óóó. 2002. Going-concern opinions in the ìnewî legal environment. Accounting Horizons (March): 17ñ26.

Auditing: A Journal of Practice & Theory, September 2003

68

Auditing, September 2003

Hausman, J. A. 1978. Specification tests in econometrics. Econometrica (46): 1251ñ1271. Hopwood, W., J. McKeown, and J. Mutchler. 1989. A test of the incremental explanatory power of opinions qualified for consistency and uncertainty. The Accounting Review (January): 28ñ48. óóó, óóó, and óóó. 1994. A reexamination of auditor versus model accuracy within the context of the going-concern opinion decision. Contemporary Accounting Research (Spring): 409ñ431. Hylton, D. P. 1964. Are consulting and auditing compatible? A contrary view. The Accounting Review (July):

667ñ670.

Kida, T. 1980. An investigation into auditorsí continuity and related qualification judgments. Journal of Accounting Research (Autumn): 506ñ523.

Lennox, C. S. 1999. Non-audit fees, disclosure and audit quality. The European Accounting Review (2): 239ñ252. Levitt, A. 1996. The accountantís critical eye. Remarks at the 24th Annual National Conference on Current SEC Developments (AICPA), Washington, D.C., December 10. óóó. 2000. Renewing the covenant with investors. Remarks before the New York University Center for Law and Business, May 10. Loudder, M. L., I. K. Khurana, R. B. Sawyers, C. Cordery, C. Johnson, J. Lowe and R. Wunderle. 1992. The information content of audit qualifications. Auditing: A Journal of Practice & Theory (Spring): 69ñ82. Mautz, R. K., and H. A. Sharaf. 1961. The Philosophy of Auditing. Sarasota, FL: American Accounting Association. McKeown, J. C., J. F. Mutchler, and W. Hopwood. 1991. Towards an explanation of auditor failure to modify the audit opinion of bankrupt companies. Auditing: A Journal of Practice & Theory (Supplement): 1ñ

13.

Melancon, B. 2002. Statement of Barry Melancon. Committee on Financial Services, United States House of Representatives (March 13). Washington, D.C.: Government Printing Office. Mutchler, J. F. 1984. Auditorsí perceptions of the going-concern opinion decision. Auditing: A Journal of Practice & Theory (Spring): 17ñ29.

óóó. 1985. A multivariate analysis of the auditorís going concern opinion decision. Journal of Accounting Research (Autumn): 668ñ682. óóó. 1986. Empirical evidence regarding the auditorsí going-concern opinion decision. Auditing: A Jour- nal of Practice & Theory (Fall): 148ñ163. óóó, and D. D. Williams. 1990. The relationship between audit technology, client risk profiles, and the going-concern opinion decision. Auditing: A Journal of Practice & Theory (Spring): 39ñ54. óóó, W. Hopwood, and J. C. McKeown. 1997. The influence of contrary information and mitigating factors on audit opinion decisions on bankrupt companies. Journal of Accounting Research (Autumn): 295ñ

310.

NASDAQ. 2002. NASDAQ approves rule changes to modify key corporate governance standards. Press Release. May 24, 2002. New York, NY: NASDAQ. Palmrose, Z-V. 1986. The effect of non-audit services on the pricing of audit services: Further evidence. Journal of Accounting Research (Autumn): 405ñ411. Reynolds, J. K., D. R. Deis, and J. R. Francis. 2002. Professional service fees and auditor objectivity. Working paper, Louisiana State University. Sarbanes-Oxley Act. 2002. Public Law No. 107-204. Washington, D.C.: Government Printing Office. Securities and Exchange Commission (SEC). 1999. Audit Committee Disclosure . Release No. 34-42266. Available at: http://www.sec.gov/rules/final/34-42266.htm. óóó. 2000a. Proposed Rule: Revision of the Commissionís Auditor Independence Requirements. Release

Nos. 33-7870; 34-42994. Washington, D.C.: SEC. óóó. 2000b. Revision of the Commissionís Auditor Independence Requirements. Release Nos. 33-7919; 34-

43602. Washington, D.C.: SEC. Sharma, D. S. 2001. The association between non-audit services and the propensity of going concern qualifica- tions: Implications for audit independence. Asia-Pacific Journal of Accounting & Economics 8: 143ñ

156.

óóó, and J. Sidhu. 2001. Professionalism vs. commercialism: The association between non-audit services (NAS) and audit independence. Journal of Business, Finance & Accounting (June/July): 595ñ629.

Auditing: A Journal of Practice & Theory, September 2003

Geiger and Rama

69

Simnett, R., and K. Trotman. 1989. Auditor versus model: Information choice and information processing. The Accounting Review (July): 514ñ528. Simunic, D. 1980. The pricing of audit services: Theory and evidence. Journal of Accounting Research (Spring): 161ñ190. óóó. 1984. Auditing, consulting, and auditor independence. Journal of Accounting Research (Fall): 679ñ

702.

Trompeter, G. 1994. The effect of partner compensation schemes and generally accepted accounting principles on audit partner judgment. Auditing: A Journal of Practice & Theory (Fall): 56ñ68. U.S. General Accounting Office (GAO). 2002. Government Auditing Standards Amendment No. 3: Indepen- dence. Washington, D.C.: Government Printing Office. U.S. House of Representatives. 1985. Hearings before the subcommittee on oversight and investigations of the committee on energy and commerce. (February 20): No. 99ñ17. Washington, D.C.: Government Print- ing Office. óóó. 1990. Hearings before the subcommittee on telecommunications and finance of the committee on energy and commerce. (August 2): No.101ñ196. Washington, D.C.: Government Printing Office. óóó. 2002. Hearings before the committee on financial services: (March 13). Washington, D.C.: Govern- ment Printing Office. U.S. Senate (Metcalf Committee). 1976. The Accounting Establishment: A Staff Study. Subcommittee on Reports, Accounting and Management of the Commission on Government Operations. Washington, D.C.: Government Printing Office. óóó. 2002. Oversight Hearing on Accounting and Investor Protection Issues Raised by Enron and Other Public Companies. Committee on Banking, Housing and Urban Affairs (February 12 and 26, March 19 and 21). Washington, D.C.: Government Printing Office. Watts, R. L., and J. L. Zimmerman. 1986. Positive Accounting Theory. Englewood Cliffs, NJ: Prentice Hall, Inc. Weil, J. 2001. ìGoing concernsî: Did accountants fail to flag problems at dot-com casualties? Wall Street Journal (February 9): C1. Whisenant, S., S. Sankaraguruswamy, and K. Raghunandan. 2002. Evidence on the joint determination of audit and non-audit services. Working paper, University of Houston, Texas. Wines, G. 1994. Auditor independence, audit qualifications and the provision of non-audit services: A note. Accounting and Finance (May): 75ñ86. Woolridge, J. M. 2000. Introductory Econometrics: A Modern Approach. Cincinnati, OH: South-Western College Publishing. Zmijewski, M. E. 1984. Methodological issues related to the estimation of financial distress prediction models. Journal of Accounting Research (Supplement): 59ñ82.

Auditing: A Journal of Practice & Theory, September 2003