Accounting Ethics

PERCEPTIONS OF THE EFFECT OF

SARBANES-OXLEY ON EARNINGS

MANAGEMENT PRACTICES

John E. McEnroe

ABSTRACT

A key objective of the Sarbanes-Oxley Act (SOA) was the restoration of

public confidence in the integrity of audited financial statements. One sec-

tion of SOA (Section 302) requires the chief executive officer(s) and the

principal financial officer(s) to certify in each quarterly or annual report

filed with the Securities Exchange Commission (SEC) that the financial

statements fairly present the financial condition and results of operations

for the periods presented in the reports. An important distinction is that the

SEC explicitly states that fair presentation is not limited to a reference that

the statements have been presented in accordance with generally accepted

accounting principles (GAAP). As such, it would follow that this aspect of

SOA would place a higher standard of quality on the financial information

than in the past and that GAAP can no longer be used as a safe harbor

defense for earnings management practices. I surveyed CFOs of the For-

tune 500 firms and audit partners for the 33 largest audit firms by revenue

as to whether they perceived that SOA significantly reduced various earn-

ings management practices in audited financial statements in general. The

results suggest that the respondents perceived that SOA reduced earnings

management in only 4 of 15 cases, and as such, contribute to the body of

survey research involving earnings management.

Research in Accounting Regulation, Volume 19, 137–157

Copyright r 2007 by Elsevier Ltd.

All rights of reproduction in any form reserved

ISSN: 1052-0457/doi:10.1016/S1052-0457(06)19007-4

137

JOHN E. MCENROE138

The Sarbanes-Oxley Act (SOA) of 2002 (P.L. No. 107–204, 2002) primarily addresses corporate governance and oversight of the accounting profession. President George W. Bush signed the legislation on July 30, 2002. Congress passed the SOA in large measure as a response to Enron and other cor- porate scandals. Several prominent individuals involved with government oversight of capital markets have cited SOA as the most far-reaching federal securities regulation since the Securities Acts of the 1930s.1

An important provision of SOA is Section 302(a), which requires the principal executive officer(s) and the principal financial officers(s) to certify in each quarterly or annual report submitted to the SEC, among other things, that (P.L. No. 107–204, 2002, p. 65):

y based on the officer’s knowledge, the financial statements and other financial infor-

mation included in the report fairly present in all material respects the financial condition

and results of operations of the issuer as of, and for, the periods presented in the report.2

The language contrasts with that contained in the standard United States audit report, which states that the financial statements present fairly, in all material respects, the financial information of the entity in conformity with generally accepted accounting principles (GAAP).

As a result, it would follow that this aspect of SOA would place a higher standard on the quality of the financial information than in the past, and concomitantly, GAAP can no longer be used as a safe harbor defense against charges of creative accounting practices. In fact, as will be explained in more detail in a later section, the SEC (2002, p. 7) specifically states that fair presentation is not limited to GAAP. This requirement engenders a certain tension that is designed and expected to result in reduced earnings management practices.

Dechow and Skinner (2000, p. 236) state that academic research has not demonstrated that earnings management has had a large impact, on aver- age, on reported earnings. As a result, they assert that academics perceive that investors should not be concerned with its existence. The authors state that the reason for this belief is, as the research of Healy and Wahlen (1999) and others indicate, academic research is limited in its ability to identify earnings management in large samples due to measurement issues. However, they state that in contrast, practitioners and regulators observe earnings management on a regular basis because they have different objectives than earnings management researchers.

Given this background, this research examines the perceptions of Chief Financial Officers (CFOs) and audit partners of the impact of SOA on earnings management in audited financial statements. Nelson, Elliot, and

Perceptions of the Effect of Sarbanes-Oxley 139

Tarply (2003, p. 18) state that very little systematic research exists concern- ing specific attempts of earnings management. Furthermore, they cite the benefits of a better understanding of extant earnings management practices to several parties, including regulators and standard setters, auditors, CEOs, CFOs, audit committees, investors, managers, and educators. As such, this study extends such research by identifying specific areas of earnings man- agement perceived to be affected by SOA. At this point, the work of Buckmaster (2001) should be cited for it is related to the contribution of this study. His work represents the most comprehensive analysis of the earnings management literature, which he refers to as ‘‘income smoothing.’’ Buckmaster studied three taxonomies of earnings management literature over the period from 1893 to 1998. His work indicates that there will always be a debate over the best approach to study the subject. With that point made, the contribution of the results of this study involve the literature involving perceptions of earnings management versus a study attempting to measure actual observations. The paper begins with a discussion of the quality of earnings, especially within a GAAP framework, followed by the research question, the research method, discussion of results, and the sum- mary and conclusions.

BACKGROUND AND RESEARCH QUESTION

Earnings Management and GAAP

The quality of financial information, especially as it relates to the quality of earnings, has attracted much attention in both academic and business jour- nals.3 Perhaps the most often cited reference is then SEC chairman Arthur Levitt’s 1998 speech (Levitt, 1998). Levitt described what he referred to as five popular accounting ‘‘illusions’’: ‘‘big bath’’ restructuring charges, cre- ative accounting acquisitions, ‘‘cookie jar’’ reserves, immaterial misappli- cation of accounting principles, and premature revenue recognition.

Although some of the above practices appear to be in violation of GAAP, other creative accounting practices fall within the letter of GAAP. As Nelson (2003, p. 101) states, precise accounting standards may offer safe harbors through accounting transaction engineering, which in turn, con- strains accounting regulators’ ability to address earnings management which reflects aggressive financial reporting.

As such, GAAP is often referred to as a ‘‘rule dominated’’ financial re- porting system, and as in Nelson’s observation, has been blamed in large

JOHN E. MCENROE140

measure for a number of creative accounting practices. For example, past SEC Chairman Harvey Pitt stated, ‘‘The development of rule-based ac- counting standards has resulted in the employment of financial engineering techniques designed solely to achieve accounting objectives rather than economic objectives’’ (FASB, 2002, p. 2).

The FASB also concedes that there is a widely held perception that rule dominated standards allow for a manipulation of the financial information contained in an entity’s financial statements.4

Many also assert that because much of the detail and complexity in accounting stand-

ards results from rule-driven implementation guidance, the standards allow financial and

accounting engineering to structure transactions ‘‘around’’ the rules, thereby circum-

venting the intent and the spirit of the standards. (FASB, 2002, p. 2)

Given the corporate scandals and attention to earnings management in both academic research and media publications, there is a movement toward a principle-based approach to accounting standard setting. This, in the view of many, would lead to more transparency in financial statements and a reduction in earnings management.

For example, the American Accounting Association Financial Account- ing Standards Committee ‘‘strongly supported’’ the FASB’s effort to eval- uate the feasibility of concept-based standards.5 The Committee stated the reason for its position is that the associated accounting standards applicable to the transaction, in conjunction with its economic substance, should be the reference points for the manner in which it is reported. The Committee concluded that a concept-based system provides the best approach toward that objective (AAA, 2003, pp. 73–74). The Committee acknowledges, however, that the implementation and enforcement of concept-based stand- ards will be difficult, because it will take the joint efforts of management, the board of directors, and the auditors to apply professional expertise and judgment to achieve unbiased financial statements.

Katherine Schipper (2003), a current member of the FASB, on the other hand, argues that extant U.S. GAAP are, in fact, principles-based, however, they contain elements such as scope, treatment, and detailed implementation guidance that make them appear to be rules-based. Sir David Tweedle, Chairman of the International Accounting Standards Board (IASB) offered an international perspective that, in part, shares Schipper’s point of view. Tweedle states that although International Financial Reporting Standards (IFRS) and U.S. GAAP strive to be principles-based since they are derived from a body of concepts, U.S. GAAP is more specific in its requirements and implementation guidance; hence, more rule-based than the IFRS.

Perceptions of the Effect of Sarbanes-Oxley 141

Tweedle asserts that a key difference is that the international approach requires both the firm and the auditor to ‘‘take a step back’’ to determine if the proposed accounting for the transaction is consistent with the under- lying principle (FASB, 2002, p. 7).

The SEC, in its explanation of the SOA Section 302(a) requirement, also expressed its concerns with the limitations of GAAP. The SEC explicitly states in the certification statement that fair presentation is not limited to an assertion that the financial statements and other financial information have been presented in accordance with GAAP. The SEC explained its position in the ensuing paragraph:

We believe that Congress intended this statement to provide assurances that the financial

information disclosed in a report, viewed in its entirety, meets a standard of overall

material accuracy and completeness that is broader than financial reporting require-

ments under generally accepted accounting principles. In our view, a ‘‘fair presentation’’

of an issuer’s financial condition, results of operations and cash flows encompasses the

selection of appropriate accounting policies, proper application of appropriate account-

ing policies, disclosure of financial information that is informative and reasonably re-

flects the underlying transactions and events and the inclusion of any additional

disclosure necessary to provide investors with a materially accurate and complete picture

of an issuer’s financial condition, results of operations and cash flows. (SEC, 2002, p. 7)6

In essence, this position constitutes a principles-based reporting system, es- pecially given the SEC’s authority over the reporting practices of publicly traded entities. As stated earlier, this requirement creates a certain tension and pressure on the certifying officers to reduce earnings management practices. This is, this aspect of SOA is designed then to eliminate clear violation of GAAP (i.e., fraud) as well as earnings management through the application of GAAP, as described in Dechow and Skinner’s earnings management framework (Dechow and Skinner, 2000, p. 239).

Another important body that issued guidance on the application of GAAP is the Auditing Standards Board (ASB) of the AICPA. Although the Public Company Accounting Oversight Board (PCAOB) has since replaced the ASB as the promulgator of auditing standards, the ASB’s Statement on Auditing Standards (SASs) have been adopted as interim standards by the PCAOB and are in effect until modified or superceded. In 1992, the ASB issued Statement on Auditing Standards No. 69 (SAS 69), The Meaning of

Present Fairly In Conformity With Generally Accepted Accounting Principles

In The Independent Auditors’ Report (AICPA, 1992). In contrast to the SEC and the FASB, SAS 69 states that the implementation of GAAP ‘‘almost always’’ results in the fair presentation of the financial statements. However, it goes on to state that the ‘‘literal application’’ of GAAP might, in unusual

JOHN E. MCENROE142

circumstances, result in misleading financial statements. In this case the auditors should provide a qualified or adverse rather than an unqualified audit opinion (AICPA, 1992, par. 5).

The preceding discussion indicates that there is an official difference of opinion on the efficacy of GAAP to insure fair reporting. The SEC and FASB concede that there is the ability to ‘‘engineer’’ GAAP, while the ASB is more confident that the use of GAAP insures fair presentation.

It appears that the financial media sides with the SEC and FASB. Besides such derisive articles involving GAAP and earnings management, consider the following quote (Eichenwald, 2002).7

Can accounting that follows the stated rules still be unreliable? In other words, is there a

gap in GAAP? After a year of corporate scandals in which some of the most outrageous

financial reporting appears to have complied with generally accepted accounting prin-

ciples, or GAAP, the answer appears to be yes.

The author then mentioned the cases of Tyco and WorldCom in which both entities set up large reserves after an acquisition to account for anticipated costs. The reserves were later reversed, increasing future income. A review of the Wall Street Journal indicates that there continues to be concern about the use of GAAP to effect earnings management.8

Although some question the merits of academic research in identifying earnings management due to measurement issues (Healy & Wahlen, 1999; Dechow & Skinner, 2000), some recent survey research has been conducted, which overcomes these problems to some extent.

Nelson et al. (2003) surveyed audit partners from a Big Five audit firm as to the earnings management practices that they had encountered. The re- sults were that revenues and other gains represented 22 percent of the ex- amples, expenses and other losses totaled 42 percent, business combinations, 13 percent, and other approaches, 13 percent. As to the income effect in the current period, 53 percent of the practices increased income, 31 percent decreased income, and 16 percent had no clear effect.

Hodge (2003) surveyed investors regarding earnings quality, auditor in- dependence, and the relevance of audited financial information. The re- spondents believed that, on average, managers of publicly traded firms engaged in earnings management approximately 50 percent of the time. Hodge states that his findings support the validity of the SEC’s concerns involving a deterioration of earnings quality and auditor independence. Furthermore, investors perceived a decline in the reliability and the quality of earnings of publicly traded companies, as well as auditor independence.

Perceptions of the Effect of Sarbanes-Oxley 143

Cohen, Dey, and Lys (2005) studied earnings management practices over the pre-SOA and the post-SOA period. They further divided the pre-SOA period into a pre-scandal period and a scandal period. The pre-scandal period was from the first quarter of 1987 through the second quarter of 2001, and the scandal period from the third quarter of 2001 through the second quarter of 2002. This scandal period included the reporting of a number of corporate scandals, including Enron, Global-Crossing, and WorldCom, among others. The post-SOA period included the period from the third quarter of 2002 through the fourth quarter of 2003. They found that earnings management increased in a steady manner from the start of 1987 until the passage of SOA. They further discovered that it decreased after SOA was enacted.

Lobo and Zhou (2005) investigated whether the SOA officer certification requirement for SEC financial statements resulted in increased conservatism in financial reporting. They found less income increasing earnings in the initial year of certification by the CEOs and CFOs than in the previous, non- certification year and faster incorporation of losses than gains in the cer- tification years. They interpreted their findings as evidence that the SEC officer certification requirement has increased the quality of earnings through increased conservatism. Although the preceding two studies have similar findings, Cohen et al. (2005, p. 10) state that beyond these, studies on the effects of SOA on earnings management have produced inconsistent results. They state that this might be a result of the rather limited time period subsequent to the passage of SOA.

Research Question

As explained earlier, it is expected that the SOA expanded definition of ‘‘fair presentation’’ without ‘‘conformity to generally accepted accounting princi- ples’’ as a safe harbor will create a certain atmosphere of tension and result in fewer incidents of earnings management. Joseph Floyd of Huron Consulting Group LLC, a firm that specializes in forensic accounting, cited SOA in part for the 13 percent increase in error driven restatements of annual reports in 2003 over 2002. Floyd asserted that SOA engendered heightened scrutiny of annual reports by various parties, including management, audit committees, auditors and regulators (Bryan-Law, 2004). Floyd again listed SOA as the reason for the subsequent 28 percent restatement increase in 2004, caused by ‘‘an unprecedented level of scrutiny’’ (Countryman, 2005, p. 1).

Norris (2004) observed that many corporate executives used to perceive that there was a gray area between flexible accounting and fraud. If SOA is

JOHN E. MCENROE144

effective, it should concomitantly reduce the amount of aggressive accounting practices in audited financial statements, besides increasing such restatements due to errors. This, in turn, should increase the perceptions of individuals involved in financial reporting that earnings management has decreased. I use the perceptions of those involved in the financial reporting process to assess the impact of SOA on earnings management. This expectation of a higher reporting standard leads to the following research question:

Do independent auditors and CFOs perceive that the Sarbanes-Oxley Act has been

effective in reducing earnings management practices in audited financial statements?

For the purpose of this research I adopt Healy and Wahlen’s (1999, p. 368) definition of earnings management:

Earnings management occurs when managers use judgment in financial reporting and in

structuring transactions to alter financial reports to either mislead some stakeholders

about the underlying economic performance of the company or to influence contractual

outcomes that depend on reported accounting numbers.

In accordance with the scope of Nelson et al. (2003, p. 17) this definition includes earnings management practices that are in accordance with GAAP, those that are difficult to distinguish from GAAP, and clear violations of GAAP.

Research Method

The research instrument contains 15 statements related to earnings man- agement practices and are listed in Table 1. Some of the transactions are in conformity with GAAP, while others are not. I developed the earnings management practices from a review of academic and popular business literature, including Levitt (1998), Brown (2002), Nelson (2003), Kieso, Weygandt, and Warfield (2004) and numerous other sources, especially the Wall Street Journal and Business Week. I also had several individuals review the research instrument prior to the mailing, including a former member of the Auditing Standards Board.

I mailed the research instrument in June, 2004 to the CFOs of the Fortune 500 firms and 500 audit partners from the 33 largest audit firms in the U.S. by revenue. The audit partners represent a random sample obtained from a list provided by the American Institute of Certified Public Accountants (AICPA). These individuals also belong to the SEC interest section of the AICPA and therefore should be very familiar with the officer certification requirements of SOA. I included the SEC interest criterion to ensure that the

Table 1. Earnings Management Practices.

Please mark the extent to which you agree or disagree with the following statements. Your responses should be in reference to material amounts (with the

exception of Statement No. 6) in audited financial statements

The Sarbanes-Oxley Act has

been Effective in Significantly

Reducing the Following

Events in Audited Financial

Statements

n Mean

Score

Significance

of t-Test

Very

Strongly

Disagree

Strongly

Disagree

Disagree Neither

Agree

nor

Disagree

Agree Strongly

Agree

Very

Strongly

Agree

(1) (2) (3) (4) (5) (6) (7)

1. Earnings management

practices that are

accomplished through

the employment of

GAAP

CFO 88 3.56 % 5.7 15.9 33 15.9 22.7 5.7 1.1

(1.36) NS

CPA 104 3.63 % 4.8 10.6 29.8 29.8 21.2 3.8 0

(1.18)

2. Earnings management

practices that are

accomplished through

the violation of GAAP

CFO 88 4.32 % 3.4 9.1 18.2 12.5 40.9 10.2 5.7

(1.45) NS

CPA 104 4.42 % 3.8 6.7 12.6 18.3 39.4 17.3 1.9

(1.35)

3. The overstatement of one

time ‘‘Big Bath’’

restructuring charges

in order to provide a

reserve to increase

future earnings

CFO 88 3.84 % 3.4 14.8 20.5 26.1 27.3 6.8 1.1

(1.32) NS

CPA 104 3.84 % 6.7 4.8 27.9 27.9 25 6.7 1

(1.29)

4. The establishment of

‘‘cookie jar reserves’’

through the

overstatement of such

items as sales returns,

loans, losses, or

warranty costs, etc. in

order to increase future

earnings

CFO 88 3.95 % 5.7 10.2 21.6 20.5 30.7 10.2 1.1

(1.41) NS

CPA 104 3.74 % 5.8 7.7 30.8 25 24 6.7 0

(1.26)

P ercep

tio n s o f th e E ffect

o f S a rb a n es-O

x ley

1 4 5

Table 1. (Continued )

Please mark the extent to which you agree or disagree with the following statements. Your responses should be in reference to material amounts (with the

exception of Statement No. 6) in audited financial statements

The Sarbanes-Oxley Act has

been Effective in Significantly

Reducing the Following

Events in Audited Financial

Statements

n Mean

Score

Significance

of t-Test

Very

Strongly

Disagree

Strongly

Disagree

Disagree Neither

Agree

nor

Disagree

Agree Strongly

Agree

Very

Strongly

Agree

(1) (2) (3) (4) (5) (6) (7)

5. The premature

recognition of revenue,

before a sale is

complete, or when the

customer has options

to terminate, void or

delay the sale

CFO 88 4.03 % 3.4 10.2 20.5 22.7 34.1 6.8 2.3

(1.33) NS

CPA 104 4.17 % 4.8 6.7 16.3 22.2 39.4 9.6 1

(1.31)

6. The violation of GAAP if

the belief is that the

independent auditors

will not find the effects

to the material

CFO 88 4.14 % 3.4 8 18.2 20.4 42 8 0

(1.24) NS

CPA 104 4.17 % 2.9 6.7 22.1 19.2 37.5 10.6 1

(1.27)

7. The parking of equity

securities in the

‘‘available for sale’’

category in order to

‘‘cherry pick’’ gains

and increase future

earnings when the

stock is sold

CFO 87 3.59 % 3.4 9.2 26.4 48.3 11.5 1.2 0

(0.971) NS

CPA 104 3.52 % 7.7 6.7 24 50 10.6 1 0

(1.06)

8. The retirement of old

debt with low interest

at a deep discount in

order to record gain,

even though the

CFO 87 3.24 % 6.9 12.6 36.8 37.9 4.6 1.2 0

(1.01) NS

CPA 104 3.53 % 6.7 7.7 23.1 52.9 7.7 1.9 0

(1.04)

JO H N

E . M

C E N R O E

1 4 6

transaction is

accomplished by

issuing new debt with

larger after tax interest

9. The classification as an

otherwise capital lease

as an operating lease

by having a third-party

guarantee the residual

value

CFO 87 3.74 % 2.3 11.5 25.3 36.8 21.8 0 2.3

(1.13) NS

CPA 104 3.61 % 5.8 2.9 34.6 43.2 8.7 4.8 0

(1.05)

10. The overstatement of the

amount of the

purchase price of an

acquired company

allocated to ‘‘in

process research and

development’’ in order

to reduce the amount

of recorded goodwill

CFO 87 3.64 % 2.3 15 24.1 37.9 18.4 0 2.3

(1.15) NS

CPA 104 3.65 % 6.7 3.8 30.8 35.6 22.1 1 0

(1.10)

11. The overstatement of the

assumed rates used to

discount pension and

post-retirement benefit

obligations in order to

improve the

appearance of the

Company’s funded

status of its plans

CFO 86 3.48 % 5.8 15.1 30.2 29.1 14 5.8 0

(1.24) NS

CPA 104 3.65 % 6.7 6.7 24 41.3 19.2 1.9 0

(1.13)

12. The deferral of expenses

in order to improve

earnings

CFO 86 3.98 % 2.3 12.8 27.9 10.5 37.2 7 2.3

(1.37) NS

CPA 104 3.92 % 7.7 4.8 27.9 20.2 26.9 11.5 1

(1.41)

13. The use of special

purpose entities (SPEs)

in order to secure off

balance sheet financing

CFO 86 4.48 % 2.3 9.3 18.6 12.8 29.1 20.9 7

(1.52) NS

CPA 104 4.42 % 5.8 2.9 18.2 21.2 28.8 14.4 8.7

(1.51)

P ercep

tio n s o f th e E ffect

o f S a rb a n es-O

x ley

1 4 7

Table 1. (Continued )

Please mark the extent to which you agree or disagree with the following statements. Your responses should be in reference to material amounts (with the

exception of Statement No. 6) in audited financial statements

The Sarbanes-Oxley Act has

been Effective in Significantly

Reducing the Following

Events in Audited Financial

Statements

n Mean

Score

Significance

of t-Test

Very

Strongly

Disagree

Strongly

Disagree

Disagree Neither

Agree

nor

Disagree

Agree Strongly

Agree

Very

Strongly

Agree

(1) (2) (3) (4) (5) (6) (7)

14. The exchange of assets

for equity securities at

a gain and recording

the transaction as a

monetary exchange in

order to record the

gain later and increase

future earnings when

the stock is sold

CFO 85 3.72 % 2.4 10.6 22.3 47.1 14.1 2.4 1.2

(1.06) NS

CPA 104 3.66 % 5.8 2.9 21.1 60.6 8.6 1 0

(.93)

15. The premature

capitalization of

software development

costs to be sold, leased

or marketed to third

parties by arbitrarily

completing a detailed

program design and

declaring that

technological feasiblity

has been established

CFO 86 3.63 % 3.5 9.3 31.4 36 17.4 1.2 1.2

(1.09) NS

CPA 104 3.68 % 7.7 2.8 26 43.3 18.2 1 1

(1.13)

All percentages total 100. The standard deviations are in parenthesis below the mean score. NS ¼ non-significant difference.

JO H N

E . M

C E N R O E

1 4 8

Perceptions of the Effect of Sarbanes-Oxley 149

individuals were familiar with the SEC requirements associated with SOA. I used the 33 largest firms as a cutoff point because a former ASB member stated that this revenue level was the point at which the firms would have a substantial number of clients subject to SOA. I obtained the CFO sample from public records. I mailed a second request to the nonrespondents in September, 2004. The response rates are depicted below.

CFOs

CPAs

First mailing

60

32

Second mailing

27

72

Total

87

104

Response rate (%)

17.4

20.8

These response rates compare favorably with that of Nelson et al. (2003) of 16 percent in their earnings management study of audit partners and managers. Given the sensitive nature of earnings management in conjunction with the time demands on the respondents, the response rate is considered adequate.

I tested for non-response bias via the ‘‘wave technique,’’ which treats the two mailings as separate ‘‘waves’’ of responses (Kanuk & Berenson, 1975; Hawkins, 1975). With this technique, I compared the CFO’s group mean scores from the first mailing to those responses from the second mailing for each item in Table 1. I repeated this process for the auditor group. A two- tailed t-test found none of the responses of the 15 statements to be signifi- cantly different for either group. Thus, nonresponse bias does not seriously affect the results. I calculated the statistic ‘‘coefficient alpha’’ to test the reliability of the instrument, and this exercise yielded a score of 95 percent. Nunnally (1978, p. 245) states that for basic research an alpha score of 0.80 is adequate; the score of 0.95 far exceeds that criterion.

The respondents were asked to mark the extent to which they agreed or disagreed that SOA has significantly reduced the following events in audited financial statements, with the values ranging from 1 (very strongly disagree) to 7 (very strongly agree).

DISCUSSION OF RESULTS

The number of responses by each group to each question, the mean scores, level of significance, and percentage response to each item are reported in

JOHN E. MCENROE150

Table 1. The two groups had very similar responses to most of the items and the t-test for a difference in the means between the two groups revealed no significant differences.9 While most of the mean scores fall in the ‘‘disagree’’ range, the standard deviations for all the means are rather large, and thus a review of the response percentages is appropriate. Table A1 presents graphs representing the percentages of the CFO and CPA responses agreeing that SOA has reduced the particular earning management practice.

The items that were perceived by the respondents to be the least affected (reduced) by SOA were Items 1 (general GAAP employment), Item 8 (debt retirement), Item 9 (capital lease misclassification), Item 10 (overstatement of ‘‘in process’’ R&D), Item 11 (overstatement of pension discount rates), and Item 15 (premature software development cost capitalization). In each of these cases less than 30 percent of both the CFOs and CPAs perceived that SOA reduced the practices, and in each case a much larger percentage of each group disagreed.

The accounting practices that the respondents perceived to be reduced the most by SOA are Items 2 (general GAAP violations), Item 5 (premature revenue recognition), Item 6 (immaterial GAAP violations), and Item 13 (the use of SPEs for off-balance sheet financing).10 In each of these cases both groups agreed at a 50 percent rate or greater (with the exception of the CPAs for Item 6; 49.1%).

The responses by both groups were mixed for Items 3 (‘‘big bath’’ re- structuring), Item 4 (‘‘cookie jar’’ reserves), and Item 12 (expense deferral). In these cases, the responses in the ‘‘agree’’ range were about the same as those in the ‘‘disagree.’’ Last, 47 percent or greater of both groups neither agreed or disagreed to Items 7 (parking equity gains) and Item 14 (parking nonmonetary gains), indicating uncertainty as to whether SOA reduced these practices (although in each case a much larger percentage disagreed than agreed that SOA reduced such transactions).

I performed another analysis, based on a grouping of relatively clearer GAAP violations (Items 2, 5, 6, and 12) versus earnings management within GAAP (all the other questions). Given that there had been no significant differences between the CFOs and CPAs, I combined them into one group for the test. The mean for the GAAP violation group was 4.16 versus 3.71 for the within GAAP set, with a t-test, two tailed score of 9.7, significant at a probabilityo0.000.11 Since Item 6 is potentially influenced by the SEC Staff Acounting Bulletin on materiality: No. 99-Materiality (SEC, 1999), and Item 13 might be viewed as a violation of GAAP due to the SEC emphasis on SPEs (i.e., SEC, 2003, 2005), I ran the same tests dropping these with results again significant at a po0.000, t-test ¼ 10.46. This demonstrates that

Perceptions of the Effect of Sarbanes-Oxley 151

the respondents feel that SOA has had more of an impact on GAAP vi- olations than earnings management within GAAP.

While these findings, in general, do not agree with the works of Cohen et al. (2005) and Lobo and Zhou (2005), reviewed previously, this is not surprising. First, this study’s research method employs a different approach than their studies, which utilize statistical models and focus on such vari- ables of interest as discretionary accruals. Second, as Cohen et al. (2005) indicate, there are mixed results involving the impact of SOA on earning management even within that research approach.

SUMMARY AND CONCLUSIONS

The findings indicate that CFOs and audit partners of firms affected by SOA do not perceive, for the most part, that the legislation has significantly reduced earnings management in audited financial statements. They feel, however, that it has been more effective in reducing clear GAAP violations than earnings management within GAAP.

The respondents further perceived that SOA reduced earnings manage- ment in only 4 of the areas listed of the 15. In the other 11 areas they either disagreed (six areas), had mixed responses (three areas), or were uncertain (two areas). Although, as stated previously, in the latter two instances a large percentage of both groups disagreed that SOA significantly reduced these earnings management practices.

It should be acknowledged that this research studies what CFOs and CPAs perceive has occurred in the area of earnings management practices as a result of SOA. This is a limitation in that actual changes (i.e., increases or decreases) in earnings management are not measured by some objective measure. Nonetheless, as Dechow and Skinner (2000) state, practitioners observe earnings management on a regular basis. CFOs, concomitantly, would either observe it at their own firm, if it is practiced there, or have a thorough exposure to it through media coverage and other professional outlets or interactions. Given that these two groups are key stake holders in financial reporting practices, their perspectives are important indicators of the impact of SOA on earnings quality. Accordingly, the results contribute to the survey analysis literature involving earnings management. Further- more, the findings associated with the clear GAAP violations and the SPEs serve as an area of future research. The results suggest that comprehensive accounting standards in a given area (i.e., SPEs) may constrain earnings management. Future research could be undertaken to test this proposition.

JOHN E. MCENROE152

Since the findings indicate that these respondents feel that the earnings management climate has not, for the most part, been improved by SOA, there are actions that might be considered on the part of accounting reg- ulators, notably the FASB, SEC, and PCAOB. A combined course of action involving ‘‘bright line’’ rules and a principles-based judgment approach by the auditor might be employed. The FASB and the SEC could issue pro- nouncements that would help to reduce such practices.

For example, in the area of operating leases, the FASB could modify the current standard and include the present value of the residual value in the 90 percent test, even if the residual value is guaranteed by a third party. An- other example is the refunding of old debt for new. The current GAAP is to calculate the gain or loss by comparing the carrying value of the old debt versus the reacquisition price. A more accurate measure would be to include the interest costs of the old versus the new debt in the calculation of the gain or loss. The SEC can also issue pronouncements aimed at reducing earnings management transactions. Its statement listed earlier in the paper articu- lating its view of what constitutes a ‘‘fair presentation’’ is a step in that direction and a principles-based reporting standard.

Some of the other practices can be reduced by the independent auditor’s refusal to issue an unqualified audit opinion if the client is engaging in such a transaction (i.e., overstatement of ‘‘in process’’ R&D, violation of immaterial GAAP). For example, the PCAOB might consider changing the standard audit report language from ‘‘present fairly in conformity with generally ac- cepted accounting principles’’ to ‘‘give a true and fair view.’’ Such language might, as in the United Kingdom and several other countries, attenuate the perception of GAAP as a safe harbor, despite the SOA certification language, and encourage auditors to employ a ‘‘principles-based’’ judgment approach, thereby mitigating the use of GAAP to manage earnings.

NOTES

1. Arthur Levitt Jr., former Securities and Exchange (SEC) Chairman, Paul Volker, former Chairman of the Federal Reserve (Volker & Levitt, 2004), Harvey Goldschmid, then Commissioner of the SEC (Goldschmid, 2003). 2. The SEC later made the language more specific to include ‘‘the financial

condition, results of operations and cash flows of the issuer’’ (SEC, 2002, p. 4). 3. For example, an entire issue of the Accounting Review (2002, Vol. 77) and

Accounting Horizons (2003) was devoted to the issue of quality of earnings. The topic also receives much attention in popular business media publications such as The Wall Street Journal and Business Week and will be cited in a later section.

Perceptions of the Effect of Sarbanes-Oxley 153

4. The FASB does acknowledge, however, that a principles-based approach does have limitations, and that even good faith professional judgments can result in different interpretations involving similar transactions, impairing comparability. The FASB further states that without adequate guidance on the part of the FASB, knowledgeable individuals who would provide interpretations of the transactions would become ‘‘de facto standard-setting bodies’’ without the due process system that the FASB possesses (FASB, 2002, p. 9). 5. The Committee substituted the term ‘‘concept-based’’ for ‘‘principles-based.’’ 6. Attorneys for HealthSouth Corp. founder Richard Scrushy, who was charged

with falsely certifying HealthSouth’s financial statements and violating SOA, claimed that the SOA language ‘‘fairly presents’’ and ‘‘in all material respects’’ is uncon- stitutional because the terms are ‘‘too vague’’ (Wall Street Journal, 2004a, p. B9). 7. For example, see Byrnes, Melcher, and Sparks (1998), Henry (2001), and

Brown (2002). 8. The GAAP and/or accounting practices involved such areas as business ac-

quisitions (Brooks, 2004), misclassification of securities (Wall Street Journal, 2004b), expense deferral (Heinz, 2004; Hagerty, McKinnon, & Kopecki, 2004), misapplica- tion of GAAP (Hagerty et al., 2004), off-balance sheet financing (Weil, 2004; Weil & Starkman, 2004), pension discount rates (Aeppel, 2004; Schultz, 2004), reserves (Wall Street Journal, 2004b), revenue recognition (Martinez, 2004; Solomon & Ber- man, 2004), and involved such companies as Bristol Myers, CVS, Enron, Fannie Mae, Krispy Kreme, Shurgard, UAL Corp., Union Pacific, U.S. Airways Group Inc., Walgreens, Winn Dixie, among others. 9. As an additional exercise I performed a test of independence between the two

groups for each of the 15 statements. I collapsed the responses into three categories: ‘‘agree,’’ ‘‘uncertain,’’ and ‘‘disagree.’’ The number of responses in the ‘‘agree’’ group were the total of the three ‘‘agree’’ categories, while the ‘‘disagree’’ group was compiled in the same manner. Only one statement, Item 8, indicated a significant difference between the two groups (w2 ¼ 13.16, df ¼ 2, po0.005). This results from the CPAs having a greater percentage response in the ‘‘uncertain’’ category (52%) than the CFOs (37%). 10. Section 401(a)(j) directed the SEC to issue rules regarding material off-balance

sheet transactions arrangements, obligations, and other relationships of the issuer with unconsolidated entities or persons. Section 401(c)(1) directed the SEC to study the extent of off-balance sheet transactions by issuers and whether GAAP resulted in off-balance sheet transactions to be disclosed in a transparent manner. 11. This grouping is in accordance with the Dechow and Skinner (2000) earnings

management framework.

REFERENCES

Accounting Horizons (2003). Accounting Horizons, 17, 1–128.

Accounting Review (2002). Quality of earnings conference. Accounting Review, 77, 1–202.

Aeppel, T. (2004). Goodyear and the butterfly effect. Wall Street Journal, June 4, p. C3.

American Accounting Association Financial Accounting Standards Committee (AAA). (2003).

Evaluating concepts-based vs. rules-based approaches to standard setting. Accounting

Horizons, 17, 73–89.

JOHN E. MCENROE154

American Institute of Certified Public Accountants (AICPA). (1992). Statement on auditing

standards no. 69: The meaning of present fairly in conformity with generally accepted

accounting principles in the independent auditors’ report. New York: AICPA.

Brooks, R. (2004). SEC looks into Krispy Kreme practices. Wall Street Journal, July 30, p. A3.

Brown, K. (2002). Creative accounting: How to buff a company. Wall Street Journal, February

21, pp. C1, C18.

Bryan-Law, C. (2004). Error-driven restatements rose again in 2003, but pace slowed. Wall

Street Journal, January 13, p. C3.

Buckmaster, D. (2001). Development of the income smoothing literature 1893–1998. London:

JAI.

Byrnes, N., Melcher, R., & Sparks, D. (1998). Earnings hocus-pocus. Business Week, October 5,

pp. 134–142.

Cohen, D., Dey, A., & Lys, T. (2005). Trends in earnings management and informativeness of

earnings announcements in the pre- and post-Sarbanes Oxley periods. Working paper.

Social Science Research Network.

Countryman, A. (2005). Restatements for ‘04 up 28% as law takes hold. Chicago Tribune,

January 20, Sections 3, 1, 8.

Dechow, P., & Skinner, D. (2000). Earnings management: Reconciling the views of accounting

academics, practitioners, and regulators. Accounting Horizons, 14, 235–250.

Eichenwald, K. (2002). Pushing accounting rules to the edge of the envelope. The New York

Times, December 31, pp. C1–C2.

Financial Accounting Standards Board (FASB). (2002). Proposal. Principles-based approach to

U.S. standard setting. Norwich, CT: FASB.

Goldschmid, H. (2003). The A.A. Sommer, Jr. annual lecture on corporate securities and

financial law: Post Enron America: An SEC perspective. Fordham Journal of Corporate

& Financial Law, 8, 335–353.

Hagerty, J., McKinnon, J., & Kopecki, D. (2004). Regulator details a wide range of accounting

problems at Fannie. Wall Street Journal, September 23, p. A1.

Hawkins, D. (1975). Estimation of non response bias. Sociological Methods and Research, 3,

461–481.

Healy, P., & Wahlen, J. (1999). A review of the earnings management literature and its im-

plications for standard setting. Accounting Horizons, 13, 365–383.

Heinz, M. (2004). Nortel may face more restatements. Wall Street Journal, June 30, p. A3.

Henry, D. (2001). The numbers game. Business Week, May 14, pp. 100–110.

Hodge, F. (2003). Investors’ perception of earnings quality, auditor independence, and the

usefulness of audited financial information. Accounting Horizons, 17, 37–48.

Kanuk, L., & Berenson, C. (1975). Mail surveys and response rates: A literature review. Journal

of Marketing Research, 12, 440–453.

Kieso, D., Weygandt, J., & Warfield, T. (2004). Intermediate accounting (11th ed.). Hoboken,

NJ: Wiley.

Levitt, A. (1998). The ‘‘Numbers game’’. Available at http://www.sec.gov/news/speecharchive/

1998/spch220.txt.

Lobo, G., & Zhou, J. (2005). Did conservatism in financial reporting increase after the Sarbanes-

Oxley Act and CEO/CFO certification of financial statements? Working paper, Univer-

sity of Houston and SUNY-Binghamton.

Martinez, B. (2004). Two major drug companies settle suits. Wall Street Journal, August 2,

p. B2.

Perceptions of the Effect of Sarbanes-Oxley 155

Nelson, M. (2003). Behavioral evidence on the effects of principles and rules-based standards.

Accounting Horizons, 17, 91–104.

Nelson, M., Elliot, J., & Tarply, R. (2003). How are earnings managed? Examples from au-

ditors. Accounting Horizons, 17, 17–35.

Norris, F. (2004). If Ebbers masterminded the fraud, why didn’t he sell more stock? The New

York Times, October 4, p. C1.

Nunnally, J. (1978). Psychometric theory. New York, NY: McGraw-Hill.

P.L. No. 107–204. (2002). Sarbanes-Oxley Act of 2002.

Securities and Exchange Commission (SEC). (1999). SEC Staff Accounting Bulletin. No. 99-

Materiality. Available at http://www.sec.gov/interps/account/sab99.htm.

Securities and Exchange Commission (SEC). (2002). Final rule: Certification of disclosure in

companies quarterly and annual reports. Available at http://www.sec.gov/rules/final/

33-8124.htrr.

Securities and Exchange Commission (SEC). (2003). Final rule: Disclosure in management’s

discussion and analysis about off-balance sheet arrangements and aggregate contractual

obligations. Available at http://www.sec.gov/rules/final/33-8182.htm.

Securities and Exchange Commission (SEC). (2005). Report and recommendations pursuant

to Section 401(c) of the Sarbanes-Oxley Act of 2002 on arrangements with off-balance

sheet implications: Special purpose entities, and transparency of filings by issuers.

Washington: SEC.

Schipper, K. (2003). Principles-based accounting standards. Accounting Horizons, 17, 61–72.

Schultz, E. (2004). SEC investigates six companies on pension benefit accounting. Wall Street

Journal, October 18, p. A4.

Solomon, D., & Berman, D. (2004). SEC gets tough with settlement in Lucent case. Wall Street

Journal, May 17, p. A1.

The Wall Street Journal (2004a). Scrushy attacks Sarbanes-Oxley. Wall Street Journal, June 17,

p. B9.

The Wall Street Journal (2004b). A Fannie Mae Enron. Wall Street Journal, October 4, p. A16.

Volker, P., & Levitt, A. (2004). In defense of Sarbanes-Oxley. Wall Street Journal, June 14, p.

A16.

Weil, J. (2004). How leases play a shadowy role in accounting. Wall Street Journal, September

22, p. A1.

Weil, J., & Starkman, D. (2004). Shurgard keeps lid on weak venture. Wall Street Journal,

January 4, p. C3.

APPENDIX

Table A1. Graph of Agreement Responses.

0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0%

1. Earnings management that are accomplished through the employment of GAAP.

2. Earnings management practices that are accomplished through the violation of GAAP.

3. The overstatement of one time “Big Bath" restructuring charges in order to provide a reserve to increase future earnings.

4. The establishment of "cookie jar reserves" through the overstatement of such items as sales returns, loans, losses, or warranty costs, etc. in order to increase future earnings.

5. The premature recognition of revenue, before a sale is complete, or when the customer has options to terminate, void or delay the sale.

6. The violation of GAAP if the belief is that the independent auditors will not find the effects to the material.

7. The parking of equity securities in the "available for sale" category in order to "cherry pick" gains and increase future earnings when the stock is sold.

CFO

CPA

CFO

CPA

JO H N

E . M

C E N R O E

1 5 6

Table A1. (Continued )

0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0%

8. The retirement of old debt with low interest at a deep discount in order to record gain, even though the transaction is accomplished by issuing new debt with larger after tax interest.

9. The classification as an otherwise capital lease as an operating lease by having a third party guarantee the residual value.

10. The overstatement of the amount of the purchase price of an acquired company allocated to "in process research and development" in order to reduce the amount of recorded goodwill.

11. The overstatement of the assumed rates used to discount pension and post retirement benefit obligations in order to improve the appearance of the Company's funded status of its plans.

12. The deferral of expenses in order to improve earnings.

13. The use of special purpose entities (SPEs) in order to secure off balance sheet financing.

14. The exchange of assets for equity securities at a gain and recording the transaction as a monetary exchange in order to record the gain later and increase future earnings when the stock is sold.

15. The premature capitalization of software development costs to be sold, leased or marketed to third parties by arbitrarily completing a detailed program design and declaring that technological feasiblity has been established

CFO

CPA

P ercep

tio n s o f th e E ffect

o f S a rb a n es-O

x ley

1 5 7