Auditing: A Journal of Practice & Theory American Accounting Association

Vol. 31, No. 1 DOI: 10.2308/ajpt-10192

February 2012
pp. 17–38

Investors’ Perceptions of Audit Quality:

Effects of Regulatory Change
Jason L. Smith
SUMMARY: In an experiment using M.B.A. students as proxies for individual investors, I
examine the effects of two important regulatory changes on investors’ perceptions of
audit quality. The first change examined is from a bottom-up coverage-based approach
for auditing internal controls to a more top-down risk-based approach. The second
change is litigation reform further limiting auditor liability exposure following an alleged
audit failure. I find that investors perceive a reduction in audit quality following each of the
two regulatory changes. These observed effects are mediated by a perceived focus on
efficiency for the new auditing standard and by a perceived change in the auditor’s
economic incentives following the proposed litigation reform. I also find that investors
believe the perceived reduction in audit quality will lead management to invest fewer
resources in internal controls.
Keywords: audit quality; investor perceptions; internal control; auditor liability.

Data Availability: Contact the author.

Perceptions of audit quality vary amongst stakeholders depending on their level of direct
involvement in audits and on the lens through which they assess audit quality . . . This
implies that a broader and deeper understanding of the complexities and nuances of the
topic needs to be developed through studying audit quality more holistically. It also
implies that individual stakeholders should consider more carefully whether actions they
endorse might have detrimental effects on others’ perspectives of audit quality. Therefore,
understanding each other’s views and how one’s actions may impact on others’
perceptions of audit quality is critical to efforts to enhance audit quality.
(International Auditing and Assurance Standards Board [IAASB] 2011)

Jason L. Smith is an Assistant Professor at the University of Nevada, Las Vegas.

I express gratitude to my dissertation committee—William L. Felix, Jr. (chair), Jeffrey Schatzberg, and William S.
Waller—for their guidance and support. I am also grateful to Ken T. Trotman and two anonymous reviewers for their
helpful suggestions during the review process. I also appreciate helpful comments by Bill Messier, Kristian Mortenson, Lisa
Ordoñez, Lisa Sedor, Chad Simon, Nate Stephens, Rick Warne, David Wood, Arnie Wright, and workshop participants at
The University of Arizona, Georgia State University, University of Houston, University of Nevada, Las Vegas,
Northeastern University, The University of Texas at Arlington, and Virginia Polytechnic Institute and State University for
their valuable comments. I also thank the M.B.A. students who participated in the experiment described in the paper.
All errors are my own.
Editor’s note: Accepted by Ken Trotman.

Submitted: November 2010

Accepted: September 2011
Published Online: January 2012

17
18 Smith

INTRODUCTION

A
uditors provide a critical service to the world’s capital markets. Without high-quality
audits, managers would face a higher agency cost (Jensen and Meckling 1976), and
investors would be less confident in corporate disclosures (Libby 1979; Hodge 2001).
Because actual audit quality is an unobservable state, investors’ perceptions of audit quality
become their view of reality and likely affect their judgments and decisions (Smith and Minter
2005; SEC 2001). These perceptions of audit quality are developed, in part, by considering
auditors’ incentives (Lowe and Pany 1995). Regulators and other governance stakeholders have
long argued that investors’ perceptions of auditor independence and audit quality are vitally
important and, in some cases, tantamount to reality (e.g., IAASB 2011; SEC 2001; Chenok 1994).
As such, policy makers who enact changes affecting auditors’ incentives should also consider the
effects of those changes on investors’ perceptions of audit quality.
In an experiment using 101 Executive and Evening M.B.A. students as proxies for individual
investors, I examine the effects of two heavily debated regulatory changes on individual investors’
perceptions of audit quality. The first change I examine is from a bottom-up coverage-based
standard (i.e., Auditing Standard No. 2) to a top-down risk-based standard (i.e., Auditing Standard
No. 5) for conducting the audit of internal control pursuant to Section 404 of the Sarbanes-Oxley
Act (SOX; U.S. House of Representative 2002).1 The second change examined is the proposed
passage of legal reform that reduces auditors’ liability exposure following an alleged audit failure.
I examine individual investors because they represent an important set of financial reporting
users whose protection and advocacy is charged to oversight bodies like the Securities and
Exchange Commission (SEC) and because international regulators have recently called for research
to enhance our understanding of various stakeholders’—including investors’—perceptions of audit
quality (IAASB 2011).2
I study the effects of the two aforementioned regulatory issues because they both represent
significant changes in the auditing environment that affect auditors’ incentives and loss function for
an integrated audit engagement. The first regulatory change—replacing Auditing Standard No. 2
(AS2) with Auditing Standard No. 5 (AS5)—deals with the audit of internal control, which has
been shown in prior research to affect investors’ perceptions of firm value (Lopez et al. 2009). In
this study, I manipulate the regulatory environment in which that audit is performed. Although the
change in auditing standards has been enacted since this experiment was conducted, liability reform
reducing the auditor’s liability exposure continues to be discussed and debated in the United States
and abroad (U.S. Treasury 2008, C: 4–6; E.U. Commission 2008). Given the relatively fluid nature
of legislation and regulation surrounding the auditing environment, an examination of investors’
reactions to regulatory changes not only provides useful feedback on evolving standards, but it may
also provide important evidence for policy makers to consider when developing future regulatory
changes.
In a 2 3 2 experimental design where each of the two regulatory changes is (not) enacted by
policy makers, I find that both regulatory changes cause significant reductions in investors’
perceptions of audit quality. In the case of the change in auditing standards, I find evidence
suggesting that the perceived reduction in audit quality is driven by a perceived focus on efficiency
in the new standard. The results also suggest the perceived reduction in audit quality following the

1
Although AS2 (PCAOB 2004) does not contain language requiring a bottom-up approach, the PCAOB indicates
that ‘‘auditors approached the audit of internal control from the bottom up’’ in their Report on the Initial
Implementation of Auditing Standard No. 2 (PCAOB 2005). To ensure a different approach was taken pursuant
to AS5, the PCAOB explicitly prescribes a top-down approach in the replacement standard (PCAOB 2007).
2
According to the Investment Company Institute’s 2008 survey, more than 47 percent of U.S. households,
representing over 90 million Americans, own individual stocks or shares of mutual funds.

Auditing: A Journal of Practice & Theory

February 2012
Investors’ Perceptions of Audit Quality: Effects of Regulatory Change 19

auditor liability reform is driven by a perceived reduction in the auditor’s cost of an audit failure.
Interestingly, I find that the perceived reduction in audit quality following both regulatory changes
leads investors to expect a reduction in management’s investment in internal control. Finally, I find
that investors significantly reduce their equity investments following the liability reform, but the
effect of a change in auditing standards appears to be conditional on investors’ experience.
Specifically, investors who purchased individual stocks within the past three years appeared more
likely to increase their investment allocation following the change from AS2 to AS5 while
individuals who had not purchased individual stocks were more likely to reduce their equity
investment allocation.
Implications of these findings are threefold. First, contrary to regulators’ intent that AS5 would
improve efficiency without sacrificing effectiveness of the audit of internal controls, investors
appear to believe the new standard will achieve the improved efficiency by sacrificing audit
quality.3 Second, investors appear to believe that economic incentives to reduce audit quality
outweigh auditors’ reputation concerns following auditor liability reform. Third, individuals are
likely to reduce their equity investments if shareholder recourse is further limited by additional
liability reform.
The remainder of the paper is organized as follows. In the next section, I discuss the regulatory
background and prior literature in developing my hypotheses. The third and fourth sections
respectively describe the research method and results. The final section concludes with a summary
and discussion of limitations and future research implications.

BACKGROUND AND HYPOTHESIS DEVELOPMENT

A Change in Auditing Standards

Following several years of SOX 404 compliance under AS2, some public registrants and other
interested parties (e.g., institutional shareholders) expressed concerns that the costs of compliance
with the stringent regulations outweighed the benefits of the internal control audit and
accompanying auditor opinion. Advocates for change claimed that reforms to ease the burden of
SOX 404 would reduce the cost of compliance for public companies and would free up time and
resources that could be used by management to add value to the organization (Scannell 2007).
Some argued that AS2 caused auditors to ‘‘spend hours questioning issues that [had] little relevance
to financial statements’’ (Johnson 2007a) and that a new standard would permit auditors to perform
more effective and efficient audits by utilizing a top-down risk-based approach to the internal
control audit that would primarily focus on high-risk areas (PCAOB 2007).
In contrast to those who argued for a new auditing standard, other parties (e.g., investor
advocate groups) pointed to the restored investor confidence and historic market performance
during the three years following the implementation of SOX 404 and AS2 as benefits of the
regulation (Scannell 2007). Others expressed concern that AS5 represented a ‘‘[backing]-off of the
objectives that [the PCAOB] set with AS2’’ that could lead to pre-Enron reporting quality (Johnson
2007b).
Ultimately, the PCAOB and the Securities and Exchange Commission (SEC) both approved
AS5 for integrated audits with year endings of November 15, 2007, or later. AS5 was introduced
‘‘to repeal [its] unduly expensive and inefficient’’ predecessor: AS2.4 The PCAOB claimed AS5

3
Although I observe a reduction in perceived audit quality following the change from AS2 to AS5, it is important
to note that both auditing standards require an independent audit of internal control over financial reporting. As
such, both standards represent significant efforts to enhance audit quality relative to the pre-SOX environment
(i.e., no audit of internal control).
4
Per SEC Chairman Chris Cox, as quoted in Johnson (2006). Also, see PCAOB (2006).

Auditing: A Journal of Practice & Theory

February 2012
20 Smith

was designed ‘‘to both increase the likelihood that material weaknesses in companies’ internal
control will be found before they cause material misstatement of the financial statements and steer
the auditor away from procedures that are not necessary to achieve the intended benefits’’ (PCAOB
2007).
DeZoort and Lee (1998) provide evidence relevant to the question of investors’ perceptions
following a change in auditing standards. In an experiment examining auditors’ perceptions of
newly passed SAS No. 82 (AICPA 1997), they show that the intent and perceived effect of a new
auditing standard may not be the same. Although regulators’ intent in enacting SAS No. 82 was
simply to clarify the auditor’s responsibility to detect fraud, participants consistently perceived an
increase in the auditor’s responsibility as a result of the new standard.5 Although the PCAOB’s
intent in passing AS5 was to promote efficiency without sacrificing effectiveness, the effect of the
change on investors’ perceptions is an empirical question.
If individual investors agree with the regulators’ intent and believe AS2 was ‘‘unduly
expensive and inefficient,’’ then a change to the revised auditing standard should not adversely
affect their perceptions of audit quality. Conversely, if individual investors disagree with regulators
and prefer the additional testing of the bottom-up coverage-based approach of AS2, then a change
to the new standard may result in a perceived reduction in audit quality.
Because individual investors generally hold under-diversified portfolios and are subject to
more idiosyncratic risk (Goetzman and Kumar 2008; Polkovnichenko 2005), I expect them to be
more concerned about audit effectiveness than efficiency. Because public discussion of the new
standard largely focused on improving audit efficiency, I predict that investors’ perceptions of audit
quality will be adversely affected as a result of the change.
H1: Investors’ perceptions of audit quality decrease with the change from AS2 to AS5.

Auditor Liability Reform

Because the public auditing industry has become increasingly concentrated over the past
twenty years and because the industry is crucial to the functioning of the world’s capital markets,
the risk of catastrophic damages stemming from an alleged audit failure poses a serious threat that
domestic and international regulators—led by the U.S. Treasury Department, the SEC, the U.S.
Chamber of Commerce, and the European Commission—have taken under serious consideration
(e.g., U.S. Treasury 2008; E.U. Commission 2008). Proposals to reduce auditor liability have
ranged from contractual (i.e., fee-based) liability caps, client-based (i.e., market value-based)
liability caps, statutory liability caps, proportionate liability laws, financial reporting insurance
policies, and securitization of financial reporting risk through bond issuances. In each case, the
objective of the proposed reform is to limit or reduce auditors’ liability following an alleged audit
failure. Although some countries have recently enacted auditor liability reform (e.g., United
Kingdom, Germany, Austria), the debate is currently ongoing in the United States and the European
Union.
While some interested parties (e.g., institutional investors, investor advocate groups) argue that
further limiting auditor liability may introduce economic incentives for auditors to accept higher
levels of audit risk and to reduce the quality of public audits (Reilly 2006; E.U. Commission 2006,
155), auditors, legislators, and corporations have indicated that the audit firms’ reputation concerns

5
Although DeZoort and Lee (1998) use auditors for participants, their findings that users’ perceptions are not
always consistent with regulators’ intent are pertinent to a setting where investors interpret the effect of a new
standard.

Auditing: A Journal of Practice & Theory

February 2012
Investors’ Perceptions of Audit Quality: Effects of Regulatory Change 21

would outweigh any economic incentives introduced by possible litigation reform (E.U.
Commission 2006, 154–155).6
Schwartz (1997), Dopuch et al. (1994), and Gramling et al. (1998) all provide evidence
consistent with auditors exerting less effort and providing lower quality audits in settings where
auditor liability is reduced. In addition, Lee and Mande (2003) and Geiger et al. (2006) provide
evidence that audit quality may have diminished following the passage of the Private Securities
Litigation Reform Act (PSLRA) of 1995. Burton et al. (2009) find that the size, distribution, and
probability of penalties for an audit failure affect auditor effort and audit quality. Each of these
studies focuses on a measure of actual audit quality and on auditors’ judgments and decisions as a
result of changes in liability reform. In contrast, my study attempts to expand our knowledge with
respect to liability reform by examining its effect on investors’ perceptions of audit quality and on
their related investment allocation decisions.
Given that the prior research in this area has not focused on investors’ perceptions, their
reaction to a reduction in auditor liability exposure is an empirical unknown. Although prior
findings suggest that actual audit quality is likely to be reduced following such a reform, individual
investors may not expect a reduction if they believe auditors’ incentives to maintain a strong
reputation or to adhere to professional norms are paramount to other economic incentives to reduce
audit quality (see Barton 2005; Mayhew 2001; Craswell et al. 1995).
Consistent with prior research on actual audit quality following a reduction in auditor liability, I
propose the following hypothesis with respect to perceived audit quality.
H2: Investors’ perceptions of audit quality decrease with the passage of litigation reform
limiting auditor liability exposure.

Implications for Investments in Internal Control

A survey of corporate executives in 2004 revealed that 79 percent of respondents reported
having stronger internal controls as a result of complying with SOX 404 under the AS2 regime
(Oversight Systems 2004). The same report, however, finds that only 13 percent of respondents felt
the benefits of SOX compliance outweighed the costs of compliance. These results—along with
experimental findings by Wu and Tuttle (2011)—suggest that the audit of internal controls
encourages management to focus time and resources on maintaining controls in order to avoid an
adverse or disclaimer opinion. Given the opportunity, however, the survey also suggests that
managers might welcome the chance to reduce their investment in internal controls. Consistent with
prior research examining the strategic interactions of auditors and managers (Bloomfield 1995,
1997; Zimbelman and Waller 1999; Bowlin 2010), managers may respond to reductions in
auditors’ testing of controls by reducing their investments in maintaining effective internal
controls.7 This perceived reduction in testing may result from either a change in auditing standards
(i.e., AS2 to AS5) or from a reduction in the cost of an audit failure (i.e., liability reform). Because I
assume the auditor and manager are both viewed by investors as economically rational agents
attempting to minimize the costs associated with SOX 404 compliance, I predict investors will
expect a reduction in management’s investment in internal control given changes in either the
auditing standard or the auditor liability laws.

6
In their final report, the U.S. Treasury’s Advisory Committee on the Auditing Profession stated: ‘‘The effect of
private litigation on auditing firms has been contentious for decades and it is not surprising that it continued to
defy a consensus solution, but the Committee’s dialogue nonetheless has laid the groundwork for continued and
constructive effort in the future’’ (U.S. Treasury Department 2008).
7
Management may choose to reduce investment in controls for nefarious purposes (i.e., to commit fraud) or to
divert funds to activities viewed as more likely to add value to the firm.

Auditing: A Journal of Practice & Theory

February 2012
22 Smith

H3a: Investors perceive that management will invest less in internal controls with the change
in auditing standard.
H3b: Investors perceive that management will invest less in internal controls with the passage
of litigation reform limiting auditor liability exposure.

Implications for Investing Decisions

Perceived reductions in audit quality (H1 and H2) and in management’s internal control
investment (H3) are likely to lead to a perceived increase in uncertainty regarding the reliability of
future financial reporting. As such, I expect investors to allocate fewer resources to equity investments
as a result of these regulatory changes and the predicted increase in financial statement risk.
H4a: Investors will allocate fewer resources to equity investments under AS5 than under AS2.
H4b: Investors will allocate fewer resources to equity investments under a lower auditor
liability setting.

EXPERIMENTAL METHOD
Design
I implement a 2 3 2 between-subjects repeated measures design. A repeated measures design is
used to create an environment analogous to a natural setting wherein participants all begin in the
same regulatory environment and are then faced with a change (or no change) to one or two
regulatory issues. The independent variables I manipulate are the applicable auditing standard (AS2/
AS5) and the relevant auditor liability law (higher liability/lower liability).8

Participants
The participants in this study are 101 Executive and evening M.B.A. students from a large,
public university.9 Participants have an average of 11.57 years of full-time work experience. Most
of them have previously invested in mutual funds (85 percent), have purchased individual stocks
within the past three years (57.4 percent), and plan to invest in individual stocks within the next two
years (71.3 percent). On a post-experimental questionnaire with a scale from 1 (Not at all Familiar)
to 9 (Very Familiar), participants self-assessed their familiarity with financial statements (5.9),
investing options (5.33), and Section 404 of the Sarbanes-Oxley Act (3.71). Participants were
randomly assigned to one of the four treatment conditions.10

Procedures
At the beginning of the experiment, participants were asked to assume the role of an investment
advisor charged with investing $5,000 on behalf of an established client.11 The client asked that the

8
All participants begin in a setting similar to the regulatory environment contemporary to the date of the
experiment. After making their assessments and investment decision in the contemporary setting, the participants
viewed the independent variable manipulations and were asked to revise their assessments.
9
One hundred sixteen M.B.A. students participated in the experiment. Fifteen participants who failed one or both
of the manipulation checks were not included in the analyses. Including these participants does not change any
reported inferences.
10
Experimental conditions do not vary significantly along any of the demographic factors (i.e., p-values . 0.10).
11
Participants were asked to serve as an investment advisor instead of being asked to invest their own money in
order to mitigate differences in individual risk preferences due to individual wealth differences. Although
random assignment to conditions should theoretically control for these differences, this design choice was used
as an additional control to mitigate any potential systematic differences between conditions.

Auditing: A Journal of Practice & Theory

February 2012
Investors’ Perceptions of Audit Quality: Effects of Regulatory Change 23

participant allocate the funds with a 12-month horizon between two investment options: (1) shares
of a publicly traded corrugated container manufacturing firm and (2) a 12-month FDIC-insured
certificate of deposit account (CD) at a local bank.12 The participant was told that the client would
close all positions at the end of one year and would realize any gains or losses at that time.
Participants were instructed that their objective was to maximize their client’s wealth at the end of
the one-year period.13
Participants first viewed background and financial information for both investment options. To
increase the level of mundane realism in the experimental setting, all information was taken from
actual companies’ web sites and SEC filings; the names of the bank and manufacturing company
were both changed to avoid recognition.14 In addition to investment-specific information, all
participants also read a section that described a current auditing standard governing audits of
internal control (i.e., AS2) and a description of the applicable laws governing auditor litigation
exposure.
After reviewing the information about the investments and the regulatory environment,
participants were asked to provide their assessment of the company’s performance, future earnings
potential, perceptions of the amount of testing performed by the auditor, three perceived levels of
audit quality, the perceived level of management ICFR investment, and the perceived costs of an
audit failure. Participants then predicted the stock price in 12 months and made an investment
allocation decision to invest the $5,000 between the CD account and the individual stock.15
The three measures of perceived audit quality mentioned above are measures of the respective
likelihoods that (1) a material weakness would go undetected by the auditor, (2) an intentional
material misstatement would be present in future financial statements, and (3) an unintentional
material misstatement would be present in future financial statements. DeAngelo (1981) defines
audit quality as the likelihood that an auditor will both discover and report a breach in the client’s
accounting system. Each of these three events represents a failure to identify or report a material
breach in the client’s system.
In the next stage of the experiment, participants viewed an e-mail from their client prior to
locking in their investment allocation decision. In the e-mail, the client asked the participant to
review two articles that might be relevant to an investment decision; the articles discussed the
possible changes to the auditing standard and to the auditor liability laws. The news articles were
compilations of actual articles and press releases that appeared on regulator web sites, in the Wall
Street Journal and in other popular business news outlets.16 The article discussing AS5 summarized
the PCAOB’s actual proposed changes (PCAOB 2007), whereas the article discussing litigation

12
The CD account is included to provide participants with a productive benchmark alternative investment. This use
of two possible investments mitigates the possibility of investing all the money in the stock due to the lack of a
reasonable alternative. The annual percentage yield (APY) on the CD represents the national average for 12-
month CDs during the time the experiment was conducted per BankRate.com.
13
The experimental sessions were conducted in October 2007. The Chicago Board Options Exchange Volatility
Index (VIX) averaged $19.11 during the month, with a low of $16.12 and a high of $22.96. The VIX index
average from January 1, 2000, to December 31, 2008, was $21.08, with a low of $9.89 and a high of $80.86.
Thus, market volatility during the time of the experiment was likely not a significant factor for participants.
14
The public company had an assets-to-debt ratio of 1.56, a current ratio of 1.79, and a gross profit margin of 20
percent. Net sales showed monotonic growth over a three-year period. Five-year stock performance data were
provided and showed a non-monotonic price increase of 22.37 percent (from $19.94 to $24.40). The current P/E
ratio was 81.33.
15
Participants were permitted to allocate their investments between the two investment options in any manner they
chose. The mean investments in the stock and CD were $1,785.50 and $3,214.50, respectively. Thirteen percent
of participants chose to invest all $5,000 in the stock, and 32 percent of participants chose to invest all of their
money in the CD. The decision to invest all or none of the money in the stock is not significantly associated with
either of the independent variable manipulations (p-values . 0.10).
16
News outlets included: WSJ.com, BusinessWeek.com, ComplianceWeek.com, and CFO.com.

Auditing: A Journal of Practice & Theory

February 2012
24 Smith

reform presented a hypothetical change in the legal environment that reduces auditor liability
exposure.17
To provide assurance that the compilation articles provided unbiased information in a clear
manner representative of what might be found in a popular news publication, 47 Master’s of
Accountancy students enrolled in advanced auditing courses at two large U.S. universities
previewed the auditing standard and auditor liability compilation articles. Master’s of Accountancy
students were specifically chosen to preview the articles because of their exposure to and
understanding of the details of the existing and impending auditing standard, their understanding of
auditor litigation issues, and their relative freedom from bias compared to other potential reviewers
(e.g., audit partners, regulators). These students rated the articles in terms of clarity, freedom from
journalistic bias, length, and likelihood of appearing in a popular business publication. Reviewers
also indicated their level of exposure to and knowledge of the topics discussed in the articles. As
shown in Table 1, the reviewers were knowledgeable about the regulatory changes, and they
indicated that the articles were clearly written, free from journalistic bias, and were generally
representative of articles commonly found in popular business news publications.
After participants in the experiment read the news articles discussing the two possible changes,
they viewed two short press releases announcing that each of the two changes discussed in the
articles had (or had not) been approved and implemented.
In the section following the news articles, participants were asked to revisit their preliminary
judgments, stock price prediction, and investment allocation decision. They were asked to answer
the same questions while considering the information provided in the news articles. After making
these revised judgments, the participants completed a post-experimental questionnaire that included
independent variable manipulation checks and appropriate demographic questions.18

RESULTS
Descriptive Statistics
Because this study focuses on the effects of regulatory changes on investors’ perceptions, I use
the changes between pre- and post-manipulation measures as my dependent variables. Descriptive
statistics—including pre- and post-manipulation values—and correlations of the change variables
are shown in Panel A and Panel B of Table 2, respectively.19

Tests of H1 and H2: Perceptions of Audit Quality

My first two hypotheses predict that investors will perceive a reduction in audit quality with
each of the two regulatory changes included in this study. Each participant provided a judgment of
the likelihoods of three undesirable events: (1) the likelihood of a material weakness going
undetected by the auditor, (2) the likelihood of an intentional material misstatement (i.e., fraud)
being present in future financial statements, and (3) the likelihood of an unintentional material
misstatement (i.e., error) being present in future financial statements. Participants were asked to

17
The auditor liability manipulation introduces an abstract representation of legal reform that would limit the
auditor’s liability exposure by removing the possibility of punitive damages. Although the manipulation is an
abstraction, it was designed to be easily understood and to have the same directional effect as the actual reform
alternatives being considered in practice.
18
The two manipulation checks presented in the post-experimental questionnaire asked participants whether the
proposed auditing standard and proposed liability reform law were accepted and implemented. Fifteen
participants failed one or both of the manipulation checks and were dropped from the analysis. Including their
responses does not change any reported inferences.
19
Results from alternative analyses using the post-manipulation measure as the dependent variable and the pre-
manipulation measure as a covariate are qualitatively the same for all reported hypothesis tests (i.e., no
inferences change due to variations in statistical significance).

Auditing: A Journal of Practice & Theory

February 2012
 TABLE 1
Characteristics of Compilation News Articles
Descriptive Statistics (Mean, (Standard Deviation), n ¼ Sample Size)

February 2012
Familiarity with
Clarity of Writing Length of Article Tone of Article Likely to Appear in WSJ . . . Subject Matter
1 ¼ Very Unclear 1 ¼ Too Short 1 ¼ Negative Bias 1 ¼ Very Unlikely 1 ¼ Not at all Familiar
3 ¼ Not Labeled 3 ¼ About Right 3 ¼ No Bias 3 ¼ Equally Likely/Unlikely 3 ¼ Somewhat Familiar
5 ¼ Very Clear 5 ¼ Too Long 5 ¼ Positive Bias 5 ¼ Very Likely 5 ¼ Very Familiar
Article Discussing 4.31 3.13 2.88 3.06 4.25
Auditing Standard (0.704) (0.341) (0.680) (1.237) (0.856)
n ¼ 16 n ¼ 16 n ¼ 24 n ¼ 16 n ¼ 16

Auditing: A Journal of Practice & Theory

Article Discussing 4.20 3.00 3.27 3.47 4.60
Auditor Liability Law (1.373) (0.000) (0.703) (0.916) (0.507)
n ¼ 15 n ¼ 15 n ¼ 22 n ¼ 15 n ¼ 15
The table provides descriptive statistics for independent reviewers’ perceptions of the two compilation articles summarizing publicly available information about proposed changes
in auditing standards and auditor liability settings.
Forty-two Master’s of Accountancy students from two large U.S. universities previewed the news articles used to introduce the independent variable manipulations in the
experimental case. At one university, students were only asked to rate the tone of the articles to determine the articles’ freedom from bias. At the other university, students were
asked to answer all five questions concerning the two articles. Four students at the first university viewed both articles. The remaining 38 students viewed one of the articles.
The articles’ ‘‘tone’’ measures whether or not the article presents information regarding the proposed change in a balanced and unbiased manner. The question posed to those asked
to rate the ‘‘tone’’ is as follows: ‘‘Please indicate below the degree of bias with regards to the proposed change, if any, present in the article.’’ Respondents indicated their response
Investors’ Perceptions of Audit Quality: Effects of Regulatory Change

on a Likert scale where 1 indicated Negative Bias and 5 indicated Positive Bias; the mid-point (3) was set as No Bias.
A t-test around the No Bias response for the article introducing the auditing standard (t-stat ¼ 0.9010, p-value ¼ 0.3769) provides evidence that the information is free from
perceived bias. A similar t-test for the article discussing the auditor liability law (t-stat ¼ 1.8209, p-value ¼ 0.0829) is not significant at the 0.05 level, but is significant at the 0.10
level. The direction, however, suggests a positive bias, which biases against finding predicted results.
25
 26
TABLE 2
Descriptive Statistics and Correlations
Panel A: Descriptive Statistics—Mean (Std. Dev.)
Auditing Standard
Existing Standard (i.e., AS2) Proposed Standard (i.e., AS5)
Auditor Liability n Pre-Manipulation Post-Manipulation Change n Pre-Manipulation Post-Manipulation Change
Existing Law (Liability Reform)
Amount of Testing 27 5.037 5.670 0.630 27 5.185 4.220 0.963
(1.721) (1.144) (1.245) (1.360) (1.476) (1.427)
Likelihood of MW 27 5.000 4.300 0.704 27 4.480 5.590 1.111
(2.000) (1.540) (1.857) (2.064) (1.927) (3.178)
Likelihood of Fraud 27 4.333 4.190 0.148 27 4.074 4.780 0.704
(1.776) (1.594) (1.231) (1.616) (2.154) (1.996)
Likelihood of Error 27 5.148 4.850 0.296 27 4.889 4.780 0.111
(1.657) (1.610) (1.514) (1.502) (1.948) (2.207)
ICFR Investment 27 6.480 5.810 0.667 27 6.300 4.630 1.667
(1.477) (1.241) (1.664) (1.325) (1.713) (1.797)
Amount Invested in Stock 24 1552.08 1787.04 250.00 27 1638.89 1925.93 287.04
(1618.47) (1816.99) (1063.22) (1599.78) (1737.34) (1001.69)
Proposed Law (Liability Reform)
Amount of Testing 21 5.119 4.640 0.476 26 5.423 4.040 1.385
(0.999) (1.131) (1.289) (1.102) (1.435) (1.423)
Likelihood of MW 21 4.790 4.880 0.095 26 4.620 5.600 0.981
(1.554) (1.564) (1.513) (1.768) (1.613) (1.770)
Likelihood of Fraud 20 4.600 4.790 0.100 26 4.173 5.480 1.308
(1.536) (1.779) (1.294) (1.679) (1.446) (1.970)
Likelihood of Error 20 4.500 4.980 0.350 26 4.808 5.870 1.058
(0.827) (1.289) (1.040) (1.625) (1.480) (2.001)
ICFR Investment 21 6.260 4.400 1.857 26 6.400 4.250 2.154
(1.114) (0.800) (1.315) (1.342) (1.478) (1.719)
Amount Invested in Stock 18 1888.89 1809.52 361.11 25 1492.00 1612.00 120.00
(1753.61) (1661.90) (1257.98) (1628.37) (1741.96) (1218.61)

February 2012
Auditing: A Journal of Practice & Theory
Smith

(continued on next page)

 TABLE 2 (continued)

Panel B: Spearman and Pearson (above/below the diagonal, respectively)

Variable 1 2 3 4 5 6

February 2012
DAmount of Testing 0.480 0.289 0.388 0.270 0.183
DLikelihood of MW 0.485 0.353 0.366 0.259 0.060
DLikelihood of Fraud 0.306 0.346 0.373 0.326 0.064
DLikelihood of Error 0.430 0.395 0.398 0.236 0.209
DICFR Investment 0.269 0.283 0.349 0.263 0.019
DAmount Invested in Stock 0.108 0.142 0.040 0.162 0.085
Italicized data are significant at the 0.05 level (two-tailed p-values).
Bolded data are significant at the 0.01 level (two-tailed p-values).

Auditing: A Journal of Practice & Theory

Variable Definitions:
Amount of Testing ¼ amount of testing to be performed by the auditor in conducting audit of internal controls (1 ¼ Too Little Testing; 5 ¼ Optimal Amount; 9 ¼ Too Much Testing);
Likelihood of MW ¼ likelihood that a material weakness will go undetected by the auditor (1 ¼ Very Unlikely; 5 ¼ Equally Likely and Unlikely; 9 ¼ Very Likely);
Likelihood of Fraud ¼ likelihood that an intentional misstatement (i.e., fraud) will be present in future financial statements (1 ¼ Very Unlikely; 5 ¼ Equally Likely and Unlikely; 9 ¼
Very Likely);
Likelihood of Error ¼ likelihood that an unintentional misstatement (i.e., error) will be present in future financial statements (1 ¼ Very Unlikely; 5 ¼ Equally Likely and Unlikely; 9
¼ Very Likely);
ICFR Investment ¼ management’s investment in internal controls over financial reporting (1 ¼ Less Investment; 5 ¼ No Effect; 9 ¼ More Investment); and
Amount Invested in Stock ¼ amount of money to be invested in company’s stock price (money not invested in the stock price is invested in a 12-month certificate of deposit).
Investors’ Perceptions of Audit Quality: Effects of Regulatory Change
27
28 Smith

provide these judgments both before and after viewing the independent variable manipulations.
Because the likelihood assessments were for three undesirable outcomes, I take the additive inverse
(i.e., x) of the difference between the before and after measures (which measure the likelihood of
an audit failure) as three individual measures of audit quality.
Because these three measures relate to the same theoretical construct, I perform a common
factor analysis and find that all three measures load on a single factor. I extract the common factor
using principal component analysis. The factor has an Eigenvalue of 1.76, and the three variables
have factor loadings of 0.75, 0.76, and 0.79, respectively. The Cronbach alpha score for internal
reliability of the three measures is 0.64.20 I use this extracted factor—the Change in Perceived
Audit Quality—as the dependent variable when testing H1 and H2.
To test these first two hypotheses, I conduct an ANOVA with the audit quality factor as the
dependent variable. Table 3 provides descriptive statistics and ANOVA results. Both the change in
auditing standard (F ¼ 10.638; p ¼ 0.001, one-tailed) and the change in auditor liability exposure (F
¼ 3.963; p ¼ 0.025, one-tailed) significantly reduce investors’ perceptions of audit quality.21 These
results provide strong support for both H1 and H2.
In addition to providing likelihood assessments for each of the three dependent variables
mentioned above, participants also provided judgments of the amount of testing performed by the
auditor prior to and following the independent variable manipulations. On a nine-point Likert scale
where 1 equals ‘‘Too Little Testing,’’ 5 equals ‘‘Optimal Amount,’’ and 9 equals ‘‘Too Much
Testing,’’ participants judged auditor testing to be near optimal (5.193, p ¼ 0.148, two-tailed) prior
to the independent variable manipulations. As shown in Table 4, ANOVA results show that both
the change from AS2 to AS5 (F ¼ 21.386; p , 0.001, one-tailed) and the liability reform (F ¼
7.977; p ¼ 0.003, one-tailed) negatively affected participants’ perceptions of the amount of testing
being performed. The post-manipulation mean assessment values for participants who were in the
auditing standard change conditions (4.132, p , 0.001, two-tailed) and the auditor liability reform
conditions (4.309, p , 0.001, two-tailed) were both significantly lower than the ‘‘Optimal Amount’’
mid-point on the scale. These results provide additional support for H1 and H2 and suggest that,
although the PCAOB intended for AS5 to improve efficiency without sacrificing effectiveness,
investors’ perceive that the reduction in testing is one from an optimal amount to one that involves
too little testing.

Test of H3: Perceptions of Investment in Internal Control

H3 predicts that investors will perceive a reduction by management in its investment in internal
control as a result of the regulatory changes. Panel A of Table 5 presents descriptive statistics and
the mean comparisons for investors’ perceptions of investment in internal control. Panel B of Table
5 presents results of an ANOVA using the change in investors’ perceptions of management’s
investment in internal control as the dependent variable. Consistent with H3a and H3b, I find that
both the change in auditing standard (F ¼ 3.851; p ¼ 0.027, one-tailed) and the reduction in auditor
liability exposure (F ¼ 6.446; p ¼ 0.007, one-tailed) cause investors to believe management will
reduce future investment in internal control.

20
Although the Cronbach alpha score of 0.64 is slightly below conventional levels (a . 0.70), the measure
inherently underestimates the reliability of factors when few measures are used (Novick and Lewis 1967). In this
case, only three variables are used in the factor analysis.
21
ANOVA, using the average of the three audit quality measures in place of the audit quality factor, yields one-
tailed p-values of , 0.001 and 0.032 for the auditing standard and liability variables, respectively. MANOVA
with each of the three change variables yields one-tailed p-values for the auditing standard and liability of 0.002
and 0.048, respectively.

Auditing: A Journal of Practice & Theory

February 2012
Investors’ Perceptions of Audit Quality: Effects of Regulatory Change 29

TABLE 3
Change in Perception of Audit Quality (H1 and H2)

Panel A: Descriptive Statistics (Mean, (Standard Deviation), n ¼ Sample Size)

Existing Standard Proposed Standard
(i.e., AS2) (i.e., AS5) Totals
Existing Law (No Liability Reform) 0.488 0.101 0.193
(0.767) (1.238) (1.062)
n ¼ 27 n ¼ 27 n ¼ 54
Proposed Law (Liability Reform) 0.140 0.509 0.227
(0.505) (1.004) (0.879)
n ¼ 20 n ¼ 26 n ¼ 46
Totals 0.340 0.301
(0.684) (1.137) n ¼ 100
n ¼ 47 n ¼ 53

Panel B: ANOVA Results

Variable Degree of Freedom F-statistic p-valuea
Auditing Standard (AS) 1 10.638 0.001
Auditor Liability (AL) 1 3.963 0.025
AS  AL 1 0.022 0.438

a
All reported p-values are one-tailed.
The Perceived Audit Quality measure results from a factor analysis of the three dependent variables defined below,
which load on a single factor with an Eigenvalue of 1.76 and a Cronbach alpha score of 0.64. The factor loading score for
each variable is included after its name.
Panel A provides descriptive statistics, by experimental condition, for the Perceived Audit Quality measure.
Panel B provides ANOVA results examining the effects of changes in the auditing standards and auditor liability settings
on a factor representing participants’ perceptions of audit quality. Results indicate a perceived reduction in audit quality
following both the change in auditing standards and the change in auditor liability.

Variable Definitions:
Likelihood of MW (0.754) ¼ likelihood that a material weakness will go undetected by the auditor (1 ¼ Very Unlikely; 5
¼ Equally Likely and Unlikely; 9 ¼ Very Likely);
Likelihood of Fraud (0.756) ¼ likelihood that an intentional misstatement (i.e., fraud) will be present in future financial
statements (1 ¼ Very Unlikely; 5 ¼ Equally Likely and Unlikely; 9 ¼ Very Likely); and
Likelihood of Error (0.788) ¼ likelihood that an unintentional misstatement (i.e., error) will be present in future financial
statements (1 ¼ Very Unlikely; 5 ¼ Equally Likely and Unlikely; 9 ¼ Very Likely).

Test of H4: Investment Allocation Decisions

My final hypothesis predicts that investors will adjust their investment allocations away from
the public company as a result of the regulatory changes. I expect this behavior because of my
predictions of reductions in perceived audit quality (H1 and H2) and perceived management
investment in internal control (H3) and the increased uncertainty that would accompany such
perceptions. Panel A of Table 6 includes descriptive statistics for H4 data. Panel B of Table 6
displays ANOVA results using the change in participants’ equity investment allocations as the
dependent variable. ANOVA results indicate that investors do not significantly change their equity
allocations following the change in auditing standard (F ¼ 1.210; p ¼ 0.137, one-tailed), but they do
reduce their equity investment allocations following the auditor liability reform (F ¼ 2.728; p ¼
0.051, one-tailed). Thus, H4b is supported while H4a is not.

Auditing: A Journal of Practice & Theory

February 2012
30 Smith

TABLE 4
Change in Perceived Amount of Auditor Testing (H1 and H2)

Panel A: Descriptive Statistics (Mean, (Standard Deviation), n ¼ Sample Size)

Existing Standard Proposed Standard
(i.e., AS2) (i.e., AS5) Totals
Existing Law (No Liability Reform) 0.630 0.963 0.167
(1.245) (1.427) (1.551)
n ¼ 27 n ¼ 27 n ¼ 54
Proposed Law (Liability Reform) 0.476 1.385 0.979
(1.289) (1.423) (1.426)
n ¼ 21 n ¼ 26 n ¼ 47
Totals 0.146 1.170
(1.368) (1.428) n ¼ 101
n ¼ 48 n ¼ 53

Panel B: ANOVA Results

Variable Degree of Freedom F-statistic p-valuea
Auditing Standard (AS) 1 21.386 , 0.001
Auditor Liability (AL) 1 7.977 0.003
AS  AL 1 1.600 0.105

a
All reported p-values are one-tailed.
The Change in Perceived Amount of Auditor Testing variable represents a change between the pre- and post-
manipulation responses to a question asking participants to judge the amount of testing to be performed by the auditor
when conducting audits of internal controls (1 ¼ Too Little Testing; 5 ¼ Optimal Amount; 9 ¼ Too Much Testing). See
Table 2 for descriptive statistics.
The overall mean assessment for the amount of testing (5.193) was not significantly different from the mid-point of
‘‘Optimal Amount’’ prior to the independent variable manipulation. The post-manipulation mean assessment scores for
participants in both the change in auditing standards conditions (4.132) and the change in auditor liability conditions
(4.309) were significantly lower than the scale’s mid-point (two-tailed p-values ¼ 0.0001, 0.0009).
Panel A provides the descriptive statistics, by experimental condition, of participants’ perceptions of the change in
auditor testing following the independent variable manipulations.
Panel B provides ANOVA results examining the effects of changes in the auditing standards and the auditor liability
settings on participants’ perceptions of the amount of auditor testing for audits of internal controls over financial
reporting (ICFR). Results indicate a perceived reduction in auditor testing following both the change in auditing
standards and the change in auditor liability.

Panel C of Table 6 displays results of an ANOVA where an indicator variable denoting whether
or not participants had purchased individual stocks within the past three years is included as an
additional fixed factor. This result is displayed because it is the only analysis where the inclusion of a
demographic variable yielded significant results. Of interest is the significant interaction (F ¼ 9.738,
p ¼ 0.002, one-tailed) observed between the change in auditing standard and investing experience.22
While more experienced investors increased the amount of money invested in the stock following the
change from AS2 to AS5, less experienced investors reduced their investment allocation. Although I

22
This particular interaction is the only observed interaction that is statistically significant (i.e., p-value , 0.10) when
examining the various control variables as possible covariates or fixed factors in the various reported tests. Neither
an indicator variable asking if individuals planned on purchasing individual stocks within the next two years nor a
nine-point Likert scale question indicating a level of familiarity with investing options yielded similar results.

Auditing: A Journal of Practice & Theory

February 2012
Investors’ Perceptions of Audit Quality: Effects of Regulatory Change 31

TABLE 5
Change in Perceived ICFR Investment (H3)

Panel A: Descriptive Statistics (Mean, (Standard Deviation), n ¼ Sample Size)

Existing Standard Proposed Standard
(i.e., AS2) (i.e., AS5) Totals
Existing Law (No Liability Reform) 0.667 1.667 1.167
(1.664) (1.797) (1.788)
n ¼ 27 n ¼ 27 n ¼ 54
Proposed Law (Liability Reform) 1.857 2.154 2.021
(1.315) (1.719) (1.543)
n ¼ 21 n ¼ 26 n ¼ 47
Totals 1.188 1.910
(1.620) (1.760) n ¼ 101
n ¼ 48 n ¼ 53

Panel B: ANOVA Results

Variable Degree of Freedom F-statistic p-valuea
Auditing Standard (AS) 1 3.851 0.027
Auditor Liability (AL) 1 6.446 0.007
AS  AL 1 1.133 0.145

a
All reported p-values are one-tailed.
The Change in Perceived ICFR Investment variable represents a change between the pre- and post-manipulation
responses to a question asking participants to judge the current regulatory settings’ effect on the company’s investment in
internal controls (1 ¼ Less Investment, 5 ¼ No Effect, 9 ¼ More Investment). See Table 2 for descriptive statistics.
Panel A provides descriptive statistics, by experimental condition, of participants’ perceptions of management’s change
in internal control investments following the independent variable manipulations.
Panel B provides ANOVA results examining the effects of changes in the auditing standards and auditor liability settings
on participants’ perceptions of the management’s investment in internal controls over financial reporting (ICFR). Results
indicate a perceived reduction in ICFR investment following both the change in auditing standards and the change in
auditor liability.

am unable to determine empirically the cause for this divergence in behavior, it is possible that the
difference is due to experienced investors being more diversified, less risk averse, and more aware of
base rates for material misstatements than their less experienced counterparts.

Mediation Analysis
Given the ANOVA results providing support of H1, H2, H3a, H3b, and H4b, I conduct a series
of mediation analyses consistent with Baron and Kenny (1986) to better understand the
mechanisms by which the manipulated independent variables impact the dependent variables. To
perform this mediation analysis, I estimate three regression models as defined below:
MEDIATORi ¼ b0 þ b1 ðIND: VARIABLEi Þ: ð1Þ

DEP: VARIABLEi ¼ b0 þ b1 ðIND: VARIABLEi Þ: ð2Þ

DEP: VARIABLEi ¼ b0 þ b1 ðIND: VARIABLEi Þ þ b2 ðMEDIATOR VARIABLEi Þ: ð3Þ

Auditing: A Journal of Practice & Theory

February 2012
32 Smith

TABLE 6
Change in Equity Investment Allocation (H4)
Panel A: Descriptive Statistics (Mean, (Standard Deviation), n ¼ Sample Size)
Existing Standard Proposed Standard
(i.e., AS2) (i.e., AS5) Totals
Existing Law (Higher Liability) 250.00 287.04 269.608
(1063.22) (1001.69) (1020.837)
n ¼ 24 n ¼ 27 n ¼ 51
Proposed Law (Lower Liability) 361.11 120.00 81.395
(1257.98) (1218.61) (1243.700)
n ¼ 18 n ¼ 25 n ¼ 43
Totals 11.905 206.731
(1176.436) (1103.383) n ¼ 94
n ¼ 42 n ¼ 52

Panel B: ANOVA Results

Variable Degree of Freedom F-statistic p-valuea
Auditing Standard (AS) 1 1.210 0.274
Auditor Liability (AL) 1 2.730 0.102
AS  AL 1 0.888 0.348

Panel C: Post hoc ANOVA Results with Investing Experience

Variable Degree of Freedom F-statistic p-valuea
Auditing Standard (AS) 1 0.593 0.443
Auditor Liability (AL) 1 3.168 0.079
Investing Experience (IE) 1 0.049 0.825
AS  AL 1 1.277 0.262
AS  IE 1 9.738 0.002
AL  IE 1 0.297 0.587
AS  AL  IE 1 0.465 0.497

a
All reported p-values are two-tailed.
Panel A provides descriptive statistics, by experimental condition, for participants’ change in equity investment
allocations following the independent variable manipulations.
Panel B provides ANOVA results examining the effects of changes in the auditing standards and auditor liability settings
on participants’ equity investment allocations. Results indicate no significant reduction in investment following the
change in auditing standards and a marginal reduction following the change in auditor liability.
Panel C provides ANOVA results of post hoc analysis where investing experience is included as an additional fixed
factor. Fifty-four of 94 participants providing equity investment allocation data indicated they had purchased individual
stocks within the past three years (Investing Experience ¼ 1). Results indicate a significant interaction effect between
investing experience and the change in auditing standards. Experienced participants increased their equity investment
allocations following a change in the standard while less experienced participants reduced their investment.

Mediation exists if each of the following conditions is met: (A) the coefficient on IND. VARIABLE
in Equation (1) is significant, (B) the coefficient on IND. VARIABLE in Equation (2) is significant,
and (C) the coefficient on MEDIATOR VARIABLE is significant in Equation (3). Full mediation
exists if these three conditions are met and if (C 0 ) the coefficient on the IND. VARIABLE in

Auditing: A Journal of Practice & Theory

February 2012
Investors’ Perceptions of Audit Quality: Effects of Regulatory Change 33

Equation (3) is insignificant. Full mediation indicates that a single mediating variable is the cause of
the observed effect of the independent variable on the dependent variable. Partial mediation, which
is more likely, indicates that the mediating variable is one of multiple possible mediators that are
driving the observed effect.
As discussed in the development of H1, I proposed that the perceived reduction in audit quality
would be driven by a perceived reduction in auditor testing of controls. Although results in Table 4
already provide some evidence to this effect, mediation analysis provides a more precise estimate of
the impact of the mediator variable on the observed relationship. As shown in Panel A of Figure 1,
the relationship between the auditing standard and perceived audit quality is fully mediated by the
perceived change in internal control testing. This suggests that investors’ perceptions of audit
quality are primarily being determined by the reduction of internal control testing as perceived by
the change from AS2 to AS5. The Sobel (1982) test statistic for the mediated relationship (Baron
and Kenny 1986, 1177) is significant (Z ¼ 3.45, p , 0.001, one-tailed).23
When developing H2, I predicted that a perceived reduction in the auditor’s cost of an audit
failure (i.e., economic incentive) would cause investors to perceive a reduction in audit quality
following the passage of liability reform. As shown in Panel B of Figure 1, the observed reduction
in perceived audit quality following the passage of the auditor liability reform appears to be fully
mediated by the perceived change in the cost of an audit failure (Z ¼1.58, p ¼ 0.057, one-tailed).
In H3a and H3b, I predicted that the perceived reduction in audit quality would lead investors
to expect management to reduce their investment in internal control. Panel C of Figure 1 displays
the results of a pair of tests examining this proposed mediation. The perceived change in audit
quality effectively mediates the relationship between each of the two independent variables and
investors’ perceptions of management investment in internal controls. This relationship is fully
mediated for the auditing standard, but not for the auditor liability law. Both mediating effects are
statistically significant (p , 0.05, one-tailed).
My final pair of hypotheses predicts that investors will reduce their equity investment
allocation as a result of the perceived reduction in audit quality (H1 and H2) and the perceived
reduction in internal control investment (H3a and H3b). ANOVA results provide support for H4b—
a reduction in equity investment allocation following the auditor liability reform—but only partial
support for H4a. Panel D displays results of a pair of tests examining the effects of both proposed
mediating variables on the observed relationship. The results are not consistent with a mediating
effect of either the change in perceived audit quality (Z ¼ 0.353, p ¼ 0.721, two-tailed) or the
change in perceived internal control investment (Z ¼ 1.066, p ¼ 0.287, two-tailed).
In summary, the results of the mediation analysis provide several interesting insights. First,
more evidence is presented suggesting that investors’ perceptions of reduced audit quality following
the change from AS2 to AS5 are driven by a perceived reduction in testing. Second, investors’
perceptions of reduced audit quality following liability reform appear to be driven by the expected
reduction in the cost of an audit failure, which suggests that investors believe auditors respond more
to economic incentives than to reputation concerns or professional norms. Third, mediation results
suggest that the perceived reduction in audit quality observed in H1 and H2 is the driving
mechanism behind investors’ expectations that management would reduce investment in internal
control as a result of the regulatory changes. Finally, the inability to identify a significant mediating
variable for H4a and H4b suggests that the ultimate investment allocation decision is complex and
that it may not be driven by perceived changes in audit quality or internal control investment.

23
The Sobel (1982) test statistic is an additional measure to determine whether an observed mediating effect is
statistically significant. Readers are referred to Sobel (1982) and Baron and Kenny (1986) for more information.

Auditing: A Journal of Practice & Theory

February 2012
34 Smith

FIGURE 1
Tests for Mediation Testsa

Panel A: Test of H1

Z-test statistic: 3.45 (p , 0.001)

Panel B: Test of H2

Z-test statistic: 1.58 (p ¼ 0.057)

Panel C: Tests of H3a and H3b

Z-test statistic (1): 2.43 (p ¼ 0.008)

Z-test statistic (2): 1.85 (p ¼ 0.032)

(continued on next page)

Auditing: A Journal of Practice & Theory

February 2012
Investors’ Perceptions of Audit Quality: Effects of Regulatory Change 35

FIGURE 1 (continued)

Panel D: Test of H4a and H4b

Z-test statistic (1): 0.353 (p ¼ 0.362)

Z-test statistic (2): 1.066 (p ¼ 0.144)
Reported p-values are one-tailed.
a
See Baron and Kenny (1986) for more information on tests for mediation effects.

CONCLUSIONS, LIMITATIONS, AND FUTURE RESEARCH

Audit quality is a critical factor in fostering the ‘‘combination of transparency and trust’’ that
enables financial markets to function properly (U.S. Treasury 2008, II:1). Given that actual audit
quality is difficult to observe—particularly for outside investors—regulators have emphasized that
research expanding our understanding of ‘‘how one’s actions may impact on others’ perceptions of
audit quality is critical to efforts to enhance audit quality’’ (IAASB 2011). In this study, I examine
the effects of two important regulatory changes on individual investors’ perceptions of audit
quality.
Although the PCAOB’s intent in replacing AS2 with AS5 was to improve audit efficiency
without undermining audit quality, results from this study provide evidence that investors’
perceptions of audit quality and of management investment in internal controls are both negatively
affected by the change. This negative impact on investors’ perceptions of audit quality is derived
from investors’ perceptions that reductions in internal control testing will sacrifice effectiveness in
order to gain efficiency. Interestingly, it is possible that the effects of this change in perceived audit
quality may affect experienced and inexperienced investors differently as it relates to investment
allocation decisions.
The results from this study should also be of interest to regulators and legislators around the
world who continue to debate the potential costs and benefits of auditor liability reform. Although
auditors have an empirically documented incentive to maintain reputations by maintaining
high-quality audits, findings from this study suggest that auditor liability reform causes investors to
perceive a reduction in audit quality as a result of a perceived reduction in the auditor’s cost of an
audit failure. Additionally, investors expect management to reduce investment in internal control as
a result of liability reform, and investors’ equity investment allocations may be reduced as a result
of liability reform that appears to reduce the auditor’s cost of an audit failure.
This study contributes to the literature and responds to calls from regulators by being the first to
examine individual investors’ perceptions of changes in audit quality that result from regulatory
changes in the auditing environment. Although the results provide consistent evidence of perceived

Auditing: A Journal of Practice & Theory

February 2012
36 Smith

reduction in audit quality, readers are cautioned to consider the results of this study in relation to the
benefits of the discussed regulatory changes. Consistent with admonitions from the IAASB (2011)
and findings from DeZoort and Lee (1998), the findings from this study suggest that regulators
should explicitly consider both the intended and perceived effects of proposed regulatory changes
on investors’ perceptions of audit quality.
The results of this study are subject to several limitations. First, the experiment employed a
repeated measures design. Although the clarity and neutrality of the materials presented to
participants were confirmed through statistical tests, a repeated measures design may increase the
possibility of demand effects. Second, although all of the information in the experimental
instrument is available to individual investors, it is possible that the average individual investor
would be less informed about internal control standards or liability laws than those who participated
in the study. As such, the observed reduction in perceived audit quality may not generalize to
less-informed investors. Third, the investors in this study were asked to assume the role of an
investment advisor, were subject to a one-year investment horizon, and were only presented with
two investment choices. Although these factors are held constant across all participants, their
investment decisions may not reflect the same motives and incentives of investors with more
alternatives and different investment horizons. Future research might further examine the effects of
regulatory change on investors’ decisions when more investment alternatives are available and
when longer- and shorter-term horizons are implicit. Fourth, the experimental instrument focused
on the extent of testing between AS2 and AS5, but perceptions of changes in the nature of testing
and the types of controls to be tested were not addressed. Future research may be needed to
understand more fully the effects of regulatory changes that affect the nature, extent, and target of
control testing. Fifth, I observe diverging investment allocation decisions from more and less
experienced investors following the change from AS2 to AS5, but the current study is unable to
determine the cause of this behavior. Future researchers may wish to examine how regulatory
changes affect investors of varied experience levels. Sixth, the litigation reform presented in this
experiment was an abstract representation of one possible reform aimed at reducing auditor liability
exposure. Researchers may wish to examine the relative effects of various alternatives to this and
other proposed methods of litigation reform. Finally, this study examines two regulatory changes
and their effects on investors’ perceptions of audit quality. Consistent with the IAASB’s (2011) call
for more research in this area, more studies are needed to improve our understanding of various
stakeholders’ perceptions as a determinant of overall audit quality.

REFERENCES
American Institute of Certified Public Accountants (AICPA). 1997. Consideration of Fraud in a Financial
Statement Audit. Statement on Auditing Standards No. 82. New York, NY: AICPA.
Baron, R. M., and D. A. Kenny. 1986. The moderator-mediator variable distinction in social psychological
research: Conceptual, strategic and statistical considerations. Journal of Personality and Social
Psychology 51: 1173–1182.
Barton, J. 2005. Who cares about auditor reputation? Contemporary Accounting Research 22 (3): 549–586.
Bloomfield, R. 1995. Strategic dependence and inherent risk assessments. The Accounting Review 70 (1):
71–90.
Bloomfield, R. 1997. Strategic dependence and the assessment of fraud risk: A laboratory study. The
Accounting Review 72 (4): 517–538.
Bowlin, K. 2010. Can Strategic Reasoning Prompts Improve Auditors’ Sensitivity to the Strategic Risk of
Fraud? Working paper, University of Mississippi.
Burton, F. G., T. J. Wilks, and M. F. Zimbelman. 2009. Audit Penalty Regimes and Audit Effort to Detect
Fraud. Working paper, Brigham Young University.

Auditing: A Journal of Practice & Theory

February 2012
Investors’ Perceptions of Audit Quality: Effects of Regulatory Change 37

Chenok, P. B. 1994. Perception vs. reality. Journal of Accountancy 177 (1): 47–51.
Craswell, A. T., J. R. Francis, and S. L. Taylor. 1995. Auditor brand name reputations and industry
specializations. Journal of Accounting and Economics 20 (3): 297–322.
DeAngelo, L. E. 1981. Auditor size and audit quality. Journal of Accounting and Economic 3 (3): 183–199.
DeZoort, F. T., and T. A. Lee. 1998. The impact of SAS No. 82 on perceptions of external auditor
responsibility for fraud detection. International Journal of Auditing 2 (1): 167–182.
Dopuch, N., R. R. King, and J. W. Schatzberg. 1994. An experimental investigation of alternative damage-
sharing liability regimes with an auditing perspective. Journal of Accounting Research 32
(Supplement):103–130.
European Commission (E.U. Commission). 2006. Study on the Economic Impact of Auditor’s Liability
Regimes. London, U.K.: London Economics.
European Commission (E.U. Commission). 2008. Commission recommendation concerning the limitation
of the civil liability of statutory auditors and audit firms. Official Journal of the European Union
21 (6).
Geiger, M. A., K. Raghunandan, and D. V. Rama. 2006. Auditor decision-making in different litigation
environments: The private securities litigation reform act, audit report and audit firm size. Journal of
Accounting and Public Policy 25 (3): 332–253.
Goetzman, W. N., and A. Kumar. 2008. Equity portfolio diversification. Review of Finance 12: 433–463.
Gramling, A. A., J. W. Schatzberg, A. D. Bailey, Jr., and H. Zhang. 1998. The impact of legal liability
regimes and differential client risk on client acceptance, audit pricing, and audit effort decisions.
Journal of Accounting, Auditing, and Finance 13 (Fall): 437–460.
Hodge, F. 2001. Hyperlinking unaudited information to audited financial statements: Effects on investor
judgments. The Accounting Review 76 (4): 675–691.
International Auditing and Assurance Standards Board (IAASB). 2011. Enhancing Audit Quality: An
IAASB Perspective. New York, NY. Available at: http://www.IASB.org
Investment Company Institute (ICI) and Securities Industry and Financial Markets Association (SIFMA).
2008. Equity and Bond Ownership in America, 2008. Washington, D.C.: ICI and SIFMA.
Jensen, M. C., and W. H. Meckling. 1976. Theory of the firm: Managerial behavior, agency costs, and
ownership structure. Journal of Financial Economics 3 (4): 305–360.
Johnson, S. 2006. PCAOB proposes AS2 ‘‘repeal.’’ CFO.com (December 19). Available at: http://www.cfo.
com/article.cfm/8465177
Johnson, S. 2007a. AS5: Clarification or confusion? CFO.com (January 12). Available at: http://www.cfo.
com/article.cfm/8512083
Johnson, S. 2007b. AS5: More flexible, less effective? CFO.com (February 23). Available at: http://www.
cfo.com/article.cfm/8756854
Lee, H. Y., and V. Mande. 2003. The effect of the Private Securities Litigation Reform Act of 1995 on
accounting discretion of client managers of Big 6 and non-Big 6 auditors. Auditing: A Journal of
Practice & Theory 22 (1): 93–108.
Libby, R. 1979. The impact of uncertainty reporting on the loan decision. Journal of Accounting Research
17: 35–57.
Lopez, T. J., S. D. Vandervelde, and Y. Wu. 2009. Investor perceptions of an auditor’s adverse internal
control opinion. Journal of Accounting and Public Policy 28 (1): 231–250.
Lowe, D. J., and K. Pany. 1995. CPA performance of consulting engagements with audit clients: Effects on
financial statement users’ perceptions and decision. Auditing: A Journal of Practice & Theory 14 (2):
35–53.
Mayhew, B. 2001. Auditor reputation building. Journal of Accounting Research 39 (3): 599–617.
Novick, M. R., and C. Lewis. 1967. Coefficient alpha and the reliability of composite measurements.
Psychometrika 32: 1–13.
Oversight Systems, Inc. 2004. The 2004 Oversight Systems Financial Executive Report on Sarbanes-Oxley.
Atlanta, GA. Available at: http://www.OversightSystems.com
Polkovnichenko, V. 2005. Household portfolio diversification: A case for rank-dependent preferences. The
Review of Financial Studies 18 (4): 1467–1502.

Auditing: A Journal of Practice & Theory

February 2012
38 Smith

Public Company Accounting Oversight Board (PCAOB). 2004. An Audit of Internal Control over Financial
Reporting Performed in Conjunction with an Audit of Financial Statements. Auditing Standard No. 2
(AS 2). PCAOB Release No. 2004-001. March 9. Washington, D.C.: PCAOB.
Public Company Accounting Oversight Board (PCAOB). 2005. Report on the Initial Implementation of
Auditing Standard No. 2—An Audit of Internal Control over Financial Reporting Performed in
Conjunction with an Audit of Financial Statements. PCAOB Release No. 2005-023. November 30.
Washington, D.C.: PCAOB.
Public Company Accounting Oversight Board (PCAOB). 2006. Proposed Auditing Standard—An Audit of
Internal Control over Financial Reporting That Is Integrated with an Audit of Financial Statements.
PCAOB Release No. 2006-007. December 19. Washington, D.C.: PCAOB.
Public Company Accounting Oversight Board (PCAOB). 2007. An Audit of Internal Control over Financial
Reporting That Is Integrated with an Audit of Financial Statements. Auditing Standard No. 5 (AS 5).
PCAOB Release No. 2007-005A. June 12. Washington, D.C.: PCAOB.
Reilly, D. 2006. Booming audit firms seek shield from suits. Wall Street Journal (November 1).
Scannell, K. 2007. Softening a Sarbanes-Oxley thorn: SEC makes progress on tweaking the rules for
unpopular provision. Wall Street Journal (April 7): C2.
Schwartz, R. 1997. Legal regimes, audit quality and investment. The Accounting Review 72 (3): 385–406.
Securities and Exchange Commission (SEC). 2001. Final Rule: Revision of the Commission’s Auditor
Independence Requirements. Release No. 33-7919. February 5. Washington, D.C.: SEC.
Smith, D. L., and F. C. Minter. 2005. Independence: Perception or reality? Strategic Finance (December):
46–51.
Sobel, M. E. 1982. Asymptotic confidence intervals for indirect effects in structural equation models. In
Sociological Methodology, edited by S. Leinhardt, 290–312. Washington, D.C.: American
Sociological Association.
United States House of Representatives. 2002. The Sarbanes-Oxley Act of 2002. Public Law No.107-204
[H.R. 3763]. Washington, D.C.: Government Printing Office.
United States Treasury Department. 2008. The Treasury Department’s Advisory Committee on the Auditing
Profession: Final Report. October 6. Washington, D.C.: Government Printing Office.
Wu, Y., and B. Tuttle. 2011. Investor Confidence in Earnings: The Interactive Effects of Internal Control
Audits and Manager Legal Liability. Working paper, Case Western Reserve University.
Zimbelman, M. F., and W. S. Waller. 1999. An experimental investigation of auditor-auditee interaction
under ambiguity. Journal of Accounting Research 37 (Supplement): 135–155.

Auditing: A Journal of Practice & Theory

February 2012