Gold Gronewold Pott IJA2012

Seediscussions,stats,andauthorprofilesforthispublicationat:http://www.researchgate.net/publication/228299297

TheISA700AuditorsReportandtheAuditExpectationGapDoExplanationsMatter?ARTICLEinINTERNATIONALJOURNALOFAUDITINGAPRIL2012DOI:10.2139/ssrn.1492082

CITATIONS9

READS638

3AUTHORS,INCLUDING:

AnnaGoldVUUniversityAmsterdam10PUBLICATIONS43CITATIONS

SEEPROFILE

Allin-textreferencesunderlinedinbluearelinkedtopublicationsonResearchGate,lettingyouaccessandreadthemimmediately.

Availablefrom:AnnaGoldRetrievedon:04November2015

The ISA 700 Auditors Report and theAudit Expectation Gap DoExplanations Matter?ija_452 286..307

Anna Gold1, Ulfert Gronewold2 and Christiane Pott31VU University Amsterdam2University of Potsdam3University of Mnster

In this paper we test the effectiveness of explanations asmandated by the revised ISA 700 auditors report in reducingthe audit expectation gap. German auditors and financialstatement users participated in an experiment where they reada summary of a firms financial statements and an auditorsreport, the latter of which we manipulated as being theauditors report including the explanations as mandated byISA 700 versus a mere audit opinion-only version. We elicitedparticipants perceptions about auditor versus managementresponsibilities and financial statement reliability. We findstrong evidence for a persistent expectation gap with respect tothe auditors responsibilities. Meanwhile, auditors and usersreach a reasonable belief consensus regarding managementsresponsibilities and financial statement reliability. Mostnotably, explanations of the ISA 700 auditors report do notresult in a smaller expectation gap. Our findings suggestthat the audit opinion alone may signal sufficient relevantinformation to users.

Key words: Audit expectation gap, auditors, financial statementusers, financial analysts, students, auditors report, responsibilityperceptions, reliability perceptions

SUMMARY

In response to repeated observations of an auditexpectation gap between financial statement usersand the audit profession, the InternationalAccounting and Assurance Standards Board

(IAASB) released a revision of the InternationalStandard on Auditing (ISA) 700, the standard onthe auditors report, which is effective for reportsdated on or after December 31, 2006. The revisionwas undertaken in order to improve usersunderstanding of an audit and to align usersexpectations with the actual responsibilities of theauditor and management as well as the reliabilityof audited financial statements (IFAC, 2008). Thisrevision of ISA 700 primarily featured the inclusion

Correspondence to: Anna Gold, Department of Accountancy,Faculty of Economics and Business Administration, VUUniversity Amsterdam, De Boelelaan 1105, NL-1081 HVAmsterdam, The Netherlands. Email: [email protected]

International Journal of Auditing doi:10.1111/j.1099-1123.2012.00452.xInt. J. Audit. 16: 286307 (2012)

ISSN 1090-6738 2012 Blackwell Publishing Ltd

in the report of explanations of auditor versusmanagement responsibilities and of the nature,scope, and procedures of the audit. This raises thequestion whether such explanations do in factresult in any smaller expectation gap than when nofurther explanations are offered in the report.Hence, the purpose of this paper is, firstly, todetermine the current state of the expectation gapunder the revised ISA 700 auditors report and,secondly, to test whether the presence versusabsence of explanations in the auditors report asmandated by the revised ISA 700 results in asmaller expectation gap.We report the results of an experiment in which

experienced German auditors and financialstatement users (i.e., financial analysts and businessstudents) read a brief company description, asummary of the firms financial statements, andan auditors report, the latter of which wemanipulated as being either the auditors reportincluding the explanations as mandated by therevised ISA 700 or a mere audit opinion-onlyversion. Participants then responded to questionsrelated to the perceived responsibility of theauditor versus management for the financialstatements and questions about the reliability ofthe audited financial statements.We find strong evidence for a persistent audit

expectation gap between auditors and financialstatement users under the revised ISA 700auditors report with respect to the auditorsresponsibilities. Meanwhile, auditors and usersreach a reasonable belief consensus regardingmanagements responsibilities and financialstatement reliability. Most notably, theexplanations of the ISA 700 auditors report ofauditor versus management responsibilities andof the nature, scope, and procedures of the auditdo not result in a smaller expectation gap. Ourfindings suggest that the audit opinion alone mayalready signal sufficient relevant information tofinancial statement users.Overall, while the fact that an audit expectation

gap still exists under the new ISA 700 auditorsreport is in line with our expectations, it isdisconcerting that the rather detailed explanationsof auditor versus management responsibilities donot favourably affect the gap. This observation mayindicate that the explanations would need to beformulated more explicitly and clearly, or even thatusers perceptions are simply not malleable byadditional information and explanations in theauditors report.

1. INTRODUCTION

The issuance of an unqualified audit opinionimplies that the auditor believes that the financialstatements give a true and fair view in accordancewith the applicable financial reporting framework(IFAC, 2008). Prior research demonstrates thatfinancial statement users (such as bankers,investors, and financial analysts) often associatean absolute level of assurance when they readsuch messages, potentially resulting in nave orunreasonable expectations (e.g., Epstein & Geiger,1994). However, in reality, the auditor merelyprovides a reasonable level of assurance (e.g., Hasanet al., 2005), as inherent audit limitations preventthe auditor from achieving absolute assurance (Gay,Schelluch & Baines, 1998). Furthermore, financialstatement users (henceforth called users) oftenassume audits to have a broader scope than theyactually have. For example, users may erroneouslyassociate audits with an approval of managementadequacy, a guarantee of the absence of fraud, anda recommendation to invest in the respective firm(Frank, Lowe & Smith, 2001). Prior studies (e.g.,Bailey, Bylinski & Shield, 1983; Nair & Rittenberg,1987; Anderson, Maletta & Wright, 1998) havealso found that users attribute disproportionateresponsibility toward auditors, whereas in reality,management (rather than the auditor) is primarilyresponsible for the adequacy of the financialstatements. These and other differences betweenwhat users expect from the auditor and what theauditor actually provides have become known asthe audit expectation gap.1

In response to this problem, the InternationalAccounting and Assurance Standards Board(IAASB) released a revision of the InternationalStandard onAuditing (ISA) 700, the standard on theauditors report, which is effective for reports datedonor afterDecember 31, 2006.With this revision, theIAASB mandates a new wording for the auditorsreport that includes explicit explanations of theresponsibilities of management and the auditorand of the nature, scope, and procedures of theaudit (IAASB, 2004). The first objective of our studyis to empirically assess the current state of theexpectation gap in Germany given the revised ISA700 auditors report. To this end, users (financialanalysts as sophisticated users and businessstudents as unsophisticated users) and experiencedauditors were asked to read an unqualified ISA 700auditors report in its revised form and to respondto questions about the responsibilities of the

The ISA 700 Auditors Report and the Audit Expectation Gap Do Explanations Matter? 287

Int. J. Audit. 16: 286307 (2012) 2012 Blackwell Publishing Ltd

auditor and management and the reliability of theaudited financial statements.The revised ISA 700 requires the auditors report

to provide explicit explanations of the auditorsversus managements responsibilities, along withan explanation of the nature, scope, and proceduresof the audit. The revision was undertaken in orderto improve users understanding of an audit andto align users expectations with the actualresponsibilities of the auditor and management aswell as the reliability of audited financial statements(IFAC, 2008).2 This recent revision of ISA 700 raisesthe question of whether financial statement userspay attention to such explanations and whetherthese explanations do in fact result in a smaller auditexpectation gap as compared to the absence of suchexplanations in the report. Prior research partlysuggests that specific wording in the auditorsreport may lead to a better understanding of thescope, nature, and significance of audit procedures(e.g., Bailey et al., 1983; Manson & Zaman, 2001;Chong& Pflugrath, 2008), but there is also evidencesuggesting only moderate, undesired, or even noeffect at all (e.g., Humphrey, Moizer & Turley, 1992;Brown, Hatherly & Innes, 1993; Monroe &Woodliff,1994; Kneer, Reckers & Jennings, 1996; Schelluch,1996; Chong & Pflugrath, 2008; Humphrey, Loft &Woods, 2009). Hence, the second purpose of thecurrent study is to provide empirical evidence onwhether the presence of explanations in theauditors report asmandated by the revised ISA 700results in a smaller expectation gap as compared tothe absence of such explanations in the report.We use the strongest experimental manipulationpossible, which compares perceptions formed onthe basis of the complete auditors report, whichincludes the explanations as mandated by therevised ISA 700, on the one hand, compared with aversion without any further explanations (i.e., anopinion-only version), on the other. A reduced gapin expectations on the basis of the completeauditors report would indicate a positive effect ofthe presence of explanations that accompany theaudit opinion in the revised ISA700auditors report.In contrast, an unchanged gap would indicate thatthe explanations beyond the audit opinion areineffective, possibly suggesting that the auditopinion alone might signal sufficient relevantinformation to users.The results of our experiment among

experienced German auditors and users withvarying levels of sophistication provide strongevidence for a persistent audit expectation gap

under the revised ISA 700 auditors report in someareas (auditors responsibilities), while auditorsand users reach a reasonable belief consensuswith respect to other areas (managementsresponsibilities and financial statement reliability).Notably, the explanations of the ISA 700 auditorsreport do not have an effect on the gap betweenauditors and users expectations and perceptions.Our findings are important for a number of

reasons. First, to our knowledge, this is the firststudy to investigate the expectation gap under therevised ISA 700 in a European country. Priorexpectation gap research has focused largely onAnglo-Saxon countries such as the US, UK, NewZealand, and Australia (e.g., Gay, Schelluch &Reid, 1997; Gay et al., 1998; Dewing & Russell, 2002;Kirk, 2006; Chong & Pflugrath, 2008) and Asiancountries such as Singapore, China, Bangladesh,and Lebanon (e.g., Best, Bucky & Tan, 2001; Lin &Chen, 2004; Sidani, 2007; Siddiqui, Nasreen &Choudhury-Lena, 2009).3 Second, we inquiredamong a substantial range of different groupsdealing with financial statements and auditopinions, including experienced auditors, financialanalysts as sophisticated users, and businessstudents as unsophisticated or nave users. Thisapproach allowed us to capture the impact ofdifferent levels of user expertise and, thus, toinvestigate the expectation gap in a differentiatedway. Third, in light of rather mixed prior evidence,our findings provide insight into whetherfinancial statement users pay attention to explicitexplanations of auditor versus managementresponsibilities and of the nature, scope, andprocedures of the audit included in the report, andwhether the presence (versus absence) of suchexplanations results in a smaller expectation gap. Toour knowledge, we are the first to investigate suchan effect based on the revised ISA 700.This paper is organized as follows. The next

section reviews the literature and develops ourhypotheses. Section 3 describes the researchmethod, the results are presented in Section 4, andSection 5 provides a summary and discussion ofthe results.

2. DEVELOPMENT OF HYPOTHESESAND RESEARCH QUESTIONS

2.1 Persistence of the audit expectationgap (H1)

One of the first studies on the expectation gap wasconducted by Libby (1979), who investigated

288 A. Gold, U. Gronewold and C. Pott


bankers and auditors perceptions of the messagecommunicated by the auditors report. Libby foundthat the perceptions of both groups were relativelysimilar to the intended communication. In contrast,Nair and Rittenberg (1987) observed that bankersplaced greater responsibility for the completenessand accuracy of the financial statements on auditorsand less responsibility on management than didcertified public accountants (CPAs). Houghton(1987) compared shareholders and accountantsperceptions of the phrase true and fair view andfound significant differences between the twogroups.Lowe (1994) was the first researcher to compare

perception differences between auditors andjudges. Lowe found a large divergence inperceptions regarding their expectations of theauditing profession. Similarly, Frank et al. (2001)compared auditors, accounting students, andjurors attitudes toward the accounting profession.A large difference in perceptions existed betweenauditors and jurors, whereas accounting studentresponses were very similar to auditor responses.Despite efforts by the accounting professionto explicitly differentiate managementsresponsibilities for the financial statements fromthe auditors role in expressing an opinion,jurors ascribed greater responsibility to auditors.Studying perceptions related to variousdimensions of the attest function, McEnroe andMartens (2001) found that investors had higherexpectations for various facets and assurances ofthe audit, compared to auditors. Reckers et al. (2007)examined the attitudes of judges, law students,MBA students, and auditors toward the publicaccounting profession over time and found thatauditors attitudes were more favorable than thoseof the other three groups. More recently, Butler,Ward and Zimbelman (2010) investigated auditorsand investors perceptions of the intendedmeanings of key terms used to define auditorsresponsibilities. They found differences betweenauditors and investors, indicating an expectationgap.Evidence from outside the US also suggests

the existence of an expectation gap. For example,Best et al. (2001) found evidence of a wide auditexpectation gap among auditors, bankers, andinvestors in Singapore concerning several areasof auditor responsibility. Comparing perceptionsof accountants, corporate finance directors,investments analysts, bank lending officers, andfinancial journalists from the UK, Humphrey,

Moizer and Turley (1993) demonstrated theexistence of an expectation gap in a variety ofaspects, in the nature of the audit function, and theperceived performance of auditors. Research fromAustralia and New Zealand also supports thesefindings (Porter, 1993; Gay et al., 1997, 1998). Asomewhat different research approach was used byhman et al. (2006) in a study of Swedish auditorspractices. They found that auditors devoted moreresources to areas that were easily verified andfewer resources to areas perceived to be ofimportance to primary stakeholders.In conclusion, the expectation gap appears to be

a persistent phenomenon across time and nationalborders, despite several institutional changesimplemented over the last several decades. Giventhese findings, we expect that the likelihood of therevised ISA 700 auditors report narrowing theexpectation gap will be rather limited. Therefore,we hypothesize that users will still ascriberelatively more responsibility to the auditor ascompared to auditors themselves, even when theyare exposed to the revised version of anunqualified ISA 700 report. While auditors areknowledgeable about their actual role of merelyproviding assurance on financial statements forwhich management is primarily responsible, priorresearch confirms that users may have unrealisticexpectations about the auditors task andresponsibilities relative to the responsibilities ofmanagement. Therefore, we expect that usersresponsibility ascriptions toward management willbe relatively lower compared to the responsibilityascriptions of auditors. Furthermore, becauseauditors as compared to users are better aware ofthe inherent limitations of an audit and anunqualified auditors report, we expect that theyascribe less reliability to audited financialstatements than do users.

H1a: Users (i.e., students and financial analysts)of financial statements with an unqualified ISA700 auditors report ascribe relatively moreresponsibility for the financial statements to theauditor than auditors do.

H1b: Users (i.e., students and financial analysts)of financial statements with an unqualified ISA700 auditors report ascribe relatively lessresponsibility for the financial statements tomanagement than auditors do.

H1c: Users (i.e., students and financial analysts)of financial statements with an unqualified ISA



700 auditors report ascribe relatively morereliability to the underlying financial statementsthan auditors do.

2.2 Effect of users sophistication on theaudit expectation gap (H2)

Prior literature on the auditors report has involvedparticipants with different sophistication (i.e.,experience and knowledge levels) in the area offinancial reporting (e.g., bankers, jurors, andstudents). Overall, research has found that moreknowledgeable users place less responsibility onauditors than less knowledgeable users. Theseresults indicate differences in the size of theexpectation gap, depending on the experience andknowledge of potential users (e.g., Bailey et al.,1983; Humphrey et al., 1993; Manson & Zaman,2001). Similarly, Monroe and Woodliff (1993)studied the influence of education on theexpectation gap. They surveyed auditors andundergraduate students, who were eitheruneducated or educated in auditing. There werefewer differences between the auditors and theeducated students in comparison to theuneducated students, suggesting an educationeffect on the expectation gap. Similarly, Gay et al.(1997) reported that users with considerablebusiness experience had more moderateexpectations of auditors responsibilities andperceptions that were closer to those of the auditingprofession when compared to less experiencedusers. In contrast, Gramling, Schatzberg andWallace (1996) found no evidence of an educationeffect (completion of an undergraduate auditingcourse) on the audit expectation gap betweenstudents and auditors.Another set of studies measured the effect of

knowledge or education on the audit expectationgap by means of experimental manipulation, e.g.,by offering extra educational materials to asubsample of users. Such reading material helpededucate Malaysian users and corrected somemisconceptions related to the audit expectationgap (Fadzly & Ahmad, 2004). Similarly, evidencefrom Bangladesh suggests that audit educationsignificantly reduces the audit expectation gap as itrelates to financial statement reliability (Siddiquiet al., 2009).Because most prior research finds experience

and knowledge effects on the size of theexpectation gap, we divided our sample of financialstatement users into financial analysts and

students, in order to control for this effect anddetermine whether it holds for the revised ISA 700auditors report. Financial analysts are specializedin financial statement analysis and companydata interpretations. Given their comprehensivepractical experience with audited financialstatements, they should at least have a moderatelysophisticated knowledge of the implications of anauditors report. Students were chosen to representthe individual unsophisticated investor, who hasless practical experience with financial statementsand auditors reports. We selected students frombusiness and economics programs, who have notspecialized in the area of auditing, because they aresomewhat familiar with financial statements but donot have detailed formal knowledge about auditingand auditors reports (as should be the case withtypical unsophisticated investors in real-worldmarkets). Our second hypothesis is stated in threeparts as follows:

H2a: The difference in perceptions of theunqualified ISA 700 auditors report betweenunsophisticated users (students) and auditors isrelatively greater compared to the differencebetween sophisticated users (financial analysts)and auditors regarding the extent to whichresponsibility for the financial statements isascribed to the auditor.

H2b: The difference in perceptions of theunqualified ISA 700 auditors report betweenunsophisticated users (students) and auditors isrelatively greater compared to the differencebetween sophisticated users (financial analysts)and auditors regarding the extent to whichresponsibility for the financial statements isascribed to management.

H2c: The difference in perceptions of theunqualified ISA 700 auditors report betweenunsophisticated users (students) and auditors isrelatively greater compared to the differencebetween sophisticated users (financial analysts)and auditors regarding the extent to whichreliability is ascribed to audited financialstatements.

2.3 Effect of explanations on the auditexpectation gap (RQ1)

Prior research has examined the effect of wordingdifferences in the auditors report on theexpectation gap. While one stream of research



explores the alterations in readers perceptionsafter a revision of the relevant auditors reportstandard (e.g., Bailey et al., 1983), a second streamaddresses the effect of different types of auditorsreports on readers perceptions (e.g., Holt &Moizer, 1990). A third stream investigates theimpact of additional information provided to thereader, such as comparing longer versus shorterauditors reports (e.g., Miller, Reed & Strawser,1993; Humphrey et al., 1992, 2009).Early research by Bailey et al. (1983) showed

that the American Institute of Certified PublicAccountants (AICPA) proposed wording changesshifted readers perceptions of the responsibilityfor financial statements from the auditor towardmanagement in the desired way. Similarly, thenew Statement on Auditing Standards (SAS)No. 58 auditors report led to an increase inunderstandability regarding the purposes of theaudit and the responsibility of management forthe financial statements (Kelly & Mohrweis, 1989).Miller et al. (1993) compared bank loan officersperceptions of a new auditors report based on SASNo. 58 and found that loan officers who read thenew auditors report were better able to identifythe responsibilities assumed for the financialstatements by both management and the auditorscompared to loan officers who read the oldauditors report.In the early 1990s, the Australian Accounting

Research Foundation (AARF) proposed a newwording for their standard auditors report withthe issuance of the Statement of Auditing Practice(AUP) No. 3, which followed the US exampleof SAS No. 58. Monroe and Woodliff (1994)investigated the effect of the proposed wordingchanges by surveying auditors and various types ofusers. After establishing the existence of the gapunder the old form of the auditors report, theresearchers found that the new version eliminatedsome of the differences in perceptions (e.g.,regarding the auditors responsibilities), but alsocreated new differences, especially in areas notmentioned in the auditors report (e.g., fraudprevention). Kneer et al. (1996) confirmed that thelanguage of the auditors report can influenceusers perceptions of auditors responsibilities andthat the improved language in SAS No. 58achieved modest success in this regard (p. 25).In the UK, Holt and Moizer (1990) studied the

auditors report by investigating the extent towhich auditors and sophisticated usersdistinguished between various reports. Their

results indicated disagreement between the twogroups concerning the unqualified auditorsreport. This disagreement was related to boththe meaning of the unqualified auditors reportand the interpretation of the qualificationsused by auditors. In particular, users perceivedgoing-concern qualifications to be more seriousthan did auditors. Miller et al. (1993) found thatbankers considered a longer report to be moreuseful and understandable than a short-formreport. Similarly, the purpose of the study byHatherly, Innes and Brown (1991) was todisentangle whether different meanings wereattached to a complete versus short-formunqualified auditors report (using a UK derivativeof the SAS No. 58 standard auditors report). Onlya complete auditors report changed readersperceptions. This is consistent with Schelluch(1996), who found that the expectation gapregarding auditor responsibilities was diminishedwith the introduction of a complete auditorsreport. However, a complete report did not reducethe expectation gap regarding perceptions offinancial statement reliability. Innes, Brown andHatherly (1997) concluded that an expandedSAS 600 report would allow the audit professionto enhance its status without any change inactual audit activities or auditor accountability.Similarly, Koh and Woo (1998) observed abetter understanding of the scope, nature, andsignificance of the audit with a complete report.In contrast, recent Australian evidence showed

that more detailed descriptions of the respectiveresponsibilities did not result in a reduction ofthe expectation gap, whereas placing the auditopinion at the beginning of the report appearedto have some beneficial consequences on usersperceptions (Chong & Pflugrath, 2008). Similarly,Brown et al. (1993) found that an expansion of theauditors report actually increased the number ofdifferences perceived by users.Humphrey et al. (1992) suggest that auditors

report changes are a way to ensure the auditexpectation gap debate is framed in terms ofimproving the understanding of users rather than aconsideration of what the role of the audit shouldbe, or to provide more information on the natureand quality of audit performance. Instead ofdetailing considerations and findings regarding theenterprise as specific outcomes of the particularaudit, standard setters had moved towardlonger-form audit reports, where the dominantemphasis was on providing information on



generalized audit responsibilities. Similarly,Humphrey et al. (2009) conclude that the auditorsreport contains general, standardized statementson the role and limitations of the audit, but littleinformation about the specific work undertakenand the findings obtained by auditors. Thesefindings call into question whether including suchgeneralized statements or explanations in auditorsreports in addition to the audit opinionmay conveyany additional value or relevant information forusers.In conclusion, several prior studies suggest that

the presence of explanations in an auditors reportmay lead to a better understanding of theresponsibilities of management and auditors, aswell as the nature, scope, and procedures of anaudit among users. However, there are alsocontrasting findings, which suggest that the effectof such explanations may be rather small,dysfunctional, or entirely absent. These findingsraise some general doubts about the possibility toaffect users perceptions by providing explanationsin the auditors report. Given that previousfindings in this respect are mixed, we posit aresearch question rather than a directionalhypothesis to examine whether the presence of theexplanations in the ISA 700 auditors report resultsin a smaller expectation gap (which could be thecase if users pay attention to these explanations andif this leads to a shift of their perceptions towardthose of auditors) as compared to the absence ofsuch explanations, i.e. when a short-formopinion-only report is used. The research questionis stated in the following three parts:

RQ1a: Does the presence versus absence of theexplanations in the unqualified ISA 700 auditorsreport result in a smaller difference inperceptions between auditors and financialstatement users regarding the extent to whichresponsibility for the financial statements isascribed to the auditor?

RQ1b: Does the presence versus absence of theexplanations in the unqualified ISA 700 auditorsreport result in a smaller difference inperceptions between auditors and financialstatement users regarding the extent to whichresponsibility for the financial statements isascribed to management?

RQ1c: Does the presence versus absence ofthe explanations in the unqualified ISA 700auditors report result in a smaller difference in

perceptions between auditors and financialstatement users regarding the extent to whichreliability is ascribed to financial statements?

3. RESEARCH METHOD

3.1 Research design and participants

We conducted a full-factorial two (completeunqualified ISA 700 auditors report and anunqualified opinion-only version of this report) bythree (auditors, financial analysts, and students)between-subjects experiment with participantsfrom Germany. Participating auditors came fromone German Big 4 audit firm. Part of the contactdetails of financial analysts was received from theBloomberg database. Additional financial analystswere approached through the German Society ofInvestment Professionals (DVFA). Finally, studentsat the Ruhr University Bochum and the Universityof Mnster participated in the study.

3.2 Measurement of the auditexpectation gap

The primary dependent variables measuring theoverall audit expectation gap are multiple-itemconstructs adopted from instruments used in priorresearch (Kelly & Mohrweis, 1989; Holt & Moizer,1990; Hatherly et al., 1991; Miller et al., 1993;Monroe &Woodliff, 1993; Gay et al., 1997; Best et al.,2001; Frank et al., 2001), which assess: (1) the extentto which participants ascribe responsibility for thefinancial statements toward the auditor (auditorresponsibility), (2) the extent to which participantsascribe responsibility for the financial statementstoward management (management responsibility),and (3) the extent to which participants ascribereliability to the audited financial statements(reliability). Table 1 illustrates our expectation gapbelief scales and items.

3.3 Experimental procedure andmanipulations

The experiment was conducted as a web-basedsurvey. Participants received an invitation emailthat provided a link to the survey. The survey wasdesigned with the online survey softwareGlobalpark. This software allows the randomdistribution of two experimental treatments(complete unqualified ISA 700 auditors reportwith explanations versus unqualified audit



opinion-only version of this report withoutexplanations) between participants. The survey wasadministered in German. An English translation isprovided in the Appendix.All participants were asked to read a short

description of a (fictitious) stock-listed company,followed by summarized financial statementinformation from two consecutive years. Followingthis firm-specific information, all participantswere shown the auditors report on the financialstatements. In the opinion-only condition,participants only read the auditors opinion,whereas the complete auditors report conditiondisclosed the complete text of the ISA 700 auditorsreport (including the opinion). Hence, thedifference between the two conditions wasthe presence versus absence of the first partof the text of the ISA 700 auditors report, whichcontains explanations of the auditors (vis--vismanagements) responsibilities and the nature,scope, and procedures of the audit. We use thisrelatively strong manipulation in order to addressthe doubts regarding the effectiveness of including

explanations in the auditors report, which we baseon the mixed findings in prior research. If the gapbetween auditors and users remains unchangedusing a complete ISA 700 auditors report ascompared to a mere opinion-only version, thiswould be evidence of the ineffectiveness of ISA700s explanations beyond the audit opinion.4

Following the case description, participantsresponded to several sets of questions relating tothe three expectation gap constructs (see Table 1).Within each set, the order of questions wasrandomized, and the questions were followed bymanipulation checks and demographic questions.5

4. RESULTS

4.1 Sample demographics

Auditors

We contacted approximately 1,450 Big Fourauditors, and a total of 163 auditors participated inthe experiment (11.24% response rate). The averageauditor was 42.1 years old and had 14.8 years of

Table 1: Expectation gap belief scales

Panel A: Auditor responsibility factor (Cronbachs alpha = 0.851)(all item scales range from 1 = strongly disagree to 7 = strongly agree)According to my impression . . .. . . the auditor is responsible for detecting all fraud.. . . the auditor is responsible for the soundness of the internal control structure of the entity.. . . the auditor is responsible for maintaining accounting records.. . . the auditor is responsible for producing the financial statements.. . . the auditor is responsible for preventing fraud.

Panel B: Management responsibility factor (Cronbachs alpha = 0.753)(all item scales range from 1 = strongly disagree to 7 = strongly agree)According to my impression . . .. . . management is responsible for detecting all fraud.. . . management is responsible for the soundness of the internal control structure of the entity.. . . management is responsible for maintaining accounting records.. . . management is responsible for producing the financial statements.. . . management is responsible for preventing fraud.

Panel C: Financial statement reliability factor (Cronbachs alpha = 0.860)(all item scales range from 1 = strongly disagree to 7 = strongly agree)Users can have absolute assurance that the financial statements contain no material misstatements.The audited financial statements give a true and fair view of the financial position of the entity.The entity is free from fraud.The audited financial statements comply with accepted accounting practice.The audited financial statements contain no deliberate distortions.The audited financial statements contain no accidental errors.The audited financial statements have no significant omissions.The amounts and disclosures contained in the audited financial statements are credible.The company has kept proper accounting records during the year.



public accounting experience. Of the sample ofauditors, 127 (77.9%) were male. There were 29partners, three directors, 37 senior managers, 85managers, and nine senior staff auditors. We askedall participants from the different groups to ratewhich reputation level they believe the auditprofession holds. On a scale ranging from 1 (low) to7 (high), auditors perceived the audit profession tohold a fairly high reputation level (m = 5.45).

Financial analysts

A total of 868 financial analysts were contacteddirectly, and 105 financial analysts participated inthe study (7.14% response rate among the directlycontacted analysts).6 The average financial analystwas 38.2 years old and had 13 years of workexperience. Of the financial analysts, 96 (91.4%)were male. There were 32 security analysts, 17portfolio managers, three fixed-income securityanalysts, four fixed-income portfolio managers, sixdirectors of research, four chief investment officers,and 37 analysts with other functions. On a scaleranging from 1 (low) to 7 (high), financial analystsheld relatively high levels of experience withfinancial reports (m = 5.11) and knowledge aboutfinancial reporting (m = 4.76), while theirknowledge about auditing was rather moderate(m = 3.32). Finally, analysts perceived the auditprofession to hold a reputation level slightly belowthe mid-point of the scale (m = 3.68).

Students

Out of 706 students from two German universities,202 responded to our experimental survey andparticipated in the study (28.61% response rate).The average student was 24.1 years old and had 0.6years of general work experience. Of the students,143 (70.8%) were male. On a scale ranging from 1(low) to 7 (high), students had moderate overallexperience with financial reports (m = 3.56) andknowledge of financial reporting (m = 4.24).Knowledge about auditing was moderate as well(m = 3.37). Finally, students perceived the auditprofession to hold a relatively high reputation level(m = 4.87).

4.2 Manipulation checks

To verify the effectiveness of the manipulation ofthe presence versus absence of the explanations inthe auditors report (i.e., complete auditors report

versus opinion-only), we asked participantstwo questions after they had completed theexperimental survey. First, participants were askedabout the extent to which they agreed with thestatement The auditors report provided in thecase materials explicitly described the respectiveresponsibilities of management and auditors(scale from 1 = strongly disagree to 7 = stronglyagree). The overall mean response was 4.81 forthe complete report treatment and 2.20 for theopinion-only treatment. The means aresignificantly different (p < 0.01), indicatingsuccessful manipulation. The manipulation checkwas also successful when conducted separately foreach individual group (i.e., auditors, financialanalysts, and students) (largest p = 0.000).Second, participants were asked about the extent

to which they agreed with the statement Theauditors report provided in the materials explicitlydescribed the scope and principles of the auditorswork (scale from 1 = strongly disagree to7 = strongly agree). The overall mean responsewas 4.52 for the complete report treatment and2.77 for the opinion-only treatment. The meansare significantly different (p < 0.01), indicatingsuccessful manipulation. The manipulation checkwas also successful when conducted separately foreach individual group (largest p = 0.01).

4.3 Preliminary testing

As described previously, we measured threedimensions of the expectation gap, two of whichassess the perceived responsibility for financialreporting (i.e., the management responsibility andauditor responsibility dimensions), while the othergauges the perceived reliability of the auditedfinancial statements. An exploratory factor analysisusing all raw items as input variables wasperformed to verify whether our three underlyingtheoretical constructs were empirically reproducedas separate factors. This analysis indeed supportsthe existence of three stable constructs:7 (1) theresponsibility of the auditor for the financialreporting (auditor responsibility), (2) theresponsibility of management for the financialreporting (management responsibility), and (3) thereliability of the audited financial statements(reliability). Reliability analyses reveal highCronbachs alphas (0.85 for auditor responsibility,0.78 for management responsibility, and 0.86 forfinancial statement reliability), indicating that allfactors measure the underlying construct with a



high degree of consistency. Consequently, wetested our hypotheses on the basis of averageindices of each of the three constructs, rather thanusing the individual items.Furthermore, to reduce the overall experimental

error, we included the following covariates in ourdata analyses (i.e., ANCOVAs and post-hoc meancomparisons): age, gender, self-reported readingintensity of the financial statements, self-reportedreading intensity of the auditors report, andperceived audit profession reputation.8,9,10 Eventhough not all measured variables are associatedwith each of the three dependent constructs (seenote 9) we include them as covariates consistentlyacross all analyses to control as much as possiblefor the potential demographic variability in ourresponses. Notes to the upcoming analyses indicatewhether exclusion of the covariates modifies theresults. As shown, when this is the case, the mainreported results are more conservative than theresults in the notes.

4.4 Hypotheses H1 and H2

H1 predicts that the perceptions of auditors andfinancial statementusers about the ISA700 auditorsreport will differ significantly in terms of threedimensions: (1) the responsibility ascribed toauditors (H1a), (2) the responsibility ascribed tomanagement (H1b), and (3) the reliability ascribedto the audited financial statements (H1c). H2suggests that all three differences will be relativelygreater between students and auditors as comparedto financial analysts and auditors. Tables 24provide adjusted means (Panel A), ANCOVAs(Panel B), and post-hocmean comparisons (Panel C)for tests of H1ac and H2ac.For tests of H1 and H2, we used only those

observations that were provided by respondents inthe treatment condition of the complete auditorsreport, thus omitting the opinion-only data.11 Thereasoning behind this is that H1 and H2 relate todifferences in perceptions regarding the report asprescribed by ISA 700 (i.e., the complete reportincluding the explanations). Upcoming tests of RQ1(see Tables 57) will incorporate all the responses,i.e. including the data from participants in thetreatment condition of the opinion-only version ofthe report.

Auditor responsibility (H1a and H2a)

We conducted an ANCOVA with group (auditor,financial analyst, and student) as the independent

variable and the auditor responsibility index asthe dependent variable (see Table 2, Panel B). Thereis a significant difference between user groupsregarding the responsibility ascribed to auditors(p < 0.001). Post-hoc mean comparisons (seeTable 2, Panel C) reveal that the auditors meanresponsibility rating of 1.58 is significantly lowerthan the mean responsibility ratings of bothfinancial analysts (4.13) and students (3.99)(p < 0.001).12,13 These findings clearly support H1a,such that users ascribe greater responsibility forthe financial statements to auditors, in comparisonto auditors themselves. Further, the difference inascribed auditor responsibility between studentsand auditors (D = 2.41) is statistically equivalent tothe difference between financial analysts andauditors (D = 2.55) (difference-in-difference notsignificant; t-statistic = 0.50; p = 0.31; one-tailed).14As such, the results are not consistent with H2a,which predicted that students ratings would differmore from auditors ratings than those of financialanalysts.

Management responsibility (H1b and H2b)

Next, we conducted an ANCOVA with group asthe independent variable and the managementresponsibility index as the dependent variable (seeTable 3, Panel B). There is a significant differencebetween user groups regarding the responsibilityascribed to management (p < 0.01). Post-hoc meancomparisons reveal that the auditors meanresponsibility rating of 6.76 is not significantlydifferent from the mean responsibility rating offinancial analysts (6.68; p = 0.516) but issignificantly higher than students mean rating(6.29; p < 0.01). The difference between the meanresponsibility rating of financial analysts andstudents is also significant (p < 0.01).15 Thesefindings partially support H1b in that students (butnot financial analysts) ascribe lower responsibilityto management, in comparison to auditors. Thedifference in ascribed management responsibilitybetween students and auditors (D = 0.47) issignificantly greater than the difference betweenfinancial analysts and auditors (D = 0.08;t-statistic = 3.01; p = 0.00; one-tailed), supportingH2b.16

Financial statement reliability (H1c and H2c)

To test Hypotheses H1c and H2c, we conductedan ANCOVA with group as the independent



variable and the financial statement reliabilityindex as the dependent variable (see Table 4, PanelB). There is a marginally significant effect of usergroup on financial statement reliability (p < 0.10).Post-hoc mean comparisons (see Panel C) showthat the auditors mean financial statementreliability rating of 3.89 is marginally lower thanthe mean reliability rating provided by students(4.51; p < 0.10) but is not significantly differentfrom financial analysts mean reliability rating(3.92; p = 0.922). These findings partially supportH1c, such that students (but not financial analysts)ascribe greater reliability to audited financialstatements than do auditors. Again, the resultsalso support H2c, because the difference inperceived financial statement reliability betweenstudents and auditors (D = 0.62) is significantlygreater than the difference between financialanalysts and auditors (D = 0.03; t-statistic = 2.10;p = 0.02; one-tailed).17

4.5 Research question 1: Effectof explanations

RQ1 examines whether the presence (versusabsence) of explanations in the unqualified ISA 700

auditors report reduces the difference inperceptions between auditors and users in terms ofresponsibility ascribed to the auditor (RQ1a),responsibility ascribed to management (RQ1b),and reliability ascribed to the audited financialstatements (RQ1c). Again, we conducted differentANCOVAs, one for each dimension of theexpectation gap. However, this time, we added theopinion-only data and included the auditorsreport type (complete report [i.e. withexplanations] versus opinion-only report [i.e.without explanations]) as a second factor in eachmodel. Tables 57 present means (adjusted forcovariates) and ANCOVAs for tests of RQ1.

Auditor responsibility

We conducted an ANCOVA with group (auditorversus financial analyst versus student) andauditors report type (complete report versusopinion-only) as the independent variables andthe auditor responsibility index as the dependentvariable. Table 5 presents the adjusted means pertreatment condition (Panel A) and the ANCOVAresults (Panel B). As shown in Panel B, neither themain effect of the auditors report type (p = 0.765)

Table 2: Test of H1a and H2a (auditor responsibility index)

Panel A: Adjusted means for auditor responsibility index

Group Mean S.E. N

Auditor 1.58 0.19 71Financial analyst 4.13 0.19 47Student 3.99 0.17 95

Panel B: ANCOVA results for auditor responsibility index

Source Type III sumof squares

df Meansquare

F-value p-value

Corrected model 323.61 7 46.23 33.92 0.000Group 149.02 2 74.51 54.66 0.000Covariates:Age 0.00 1 0.00 0.00 0.989Gender 0.54 1 0.54 0.40 0.528Self-reported reading intensity of financial statements 0.79 1 0.79 0.58 0.446Self-reported reading intensity of auditors report 5.65 1 5.65 4.15 0.043Perceived audit profession reputation 7.02 1 7.02 5.15 0.024

Error 279.44 205 1.36

Panel C: Post-hoc mean comparisons (least significant difference)

S.E. p-value

Auditor vs. Financial analyst 0.25 0.000Auditor vs. Student 0.31 0.000Financial analyst vs. Student 0.28 0.626



nor the interaction effect between group andauditors report type (p = 0.446) on auditorresponsibility ratings are significant. However,non-tabulated simple mean comparisons of thesignificant group effect (p = 0.000) reveal thatauditors (adjusted) mean auditor responsibilityrating (1.74) remains significantly lower than bothfinancial analysts (4.10) and students meanrating (3.93), also when including the opinion-only treatment data in the study sample (bothp-values < 0.001).18 In conclusion, our resultssuggest that the presence (versus absence) ofexplanations in the ISA 700 auditors report doesnot affect the gap in perceptions between auditorsand users regarding the auditors responsibility.

Management responsibility

Next, we conducted an ANCOVAwith group andauditors report type as the independent variablesand the management responsibility index as thedependent variable. Table 6 presents the adjustedmeans per treatment condition (Panel A) and

the ANCOVA results (Panel B). Again, the onlysignificant effect is the main effect of group(p < 0.001), such that auditors rank the level ofmanagements responsibility marginally higher(6.87) than financial analysts (6.75; p < 0.10) andsignificantly higher than students (6.26; p < 0.001)do. The main effect of the report type and theinteraction effect are both insignificant.19 Inconclusion, our results suggest that the presence(versus absence) of explanations in the ISA 700auditors report does not affect the gap inperceptions between auditors and users regardingthe managements responsibility.

Financial statement reliability

To examine RQ1c, we conducted an ANCOVAwith group and auditors report type as theindependent variables and the financial statementreliability index as the dependent variable. Table 7reports the adjusted means per treatment condition(Panel A) and the ANCOVA results (Panel B).Neither the two main effects nor the interaction

Table 3: Test of H1b and H2b (management responsibility index)

Panel A: Adjusted means for management responsibility index

Group Mean S.E. N


Panel B: ANCOVA results for management responsibility index


df Meansquare

F-value p-value


Error


S.E. p-value




effect are significant (smallest p = 0.346). Hence,the presence (versus absence) of the explanationsin the ISA 700 auditors report does not bringusers reliability ratings closer to the auditorsratings.20

At least to the extent that auditor perceptionscan be used as a benchmark for normativeperceptions of management and auditorresponsibilities and financial statement reliability,our results do not suggest that the presence ofthe ISA 700 explanations of managements andauditors respective responsibilities and of thenature, scope, and procedures of the audit in thecomplete auditors report affects the expectationgap. In contrast, we find a similar gap inperceptions when the complete report with itsdetailed explanations is provided as when a reportwith the mere audit opinion is provided.

5. DISCUSSION AND CONCLUSIONS

Over 70 jurisdictions around the world haveadopted the ISAs, and several additionalimplementations are pending, most notably in theEU and China. These developments reinforce

the need for further empirical evidence oninternational users perceptions of ISA 700, whichis one of the most important standards with respectto enhancing investors confidence in financialreporting. Given permanent revisions of currentstandards to address changing investor needs,empirical evidence on user perceptions is a crucialmeans of providing insight into possible needsand directions for standard revisions. Becauseindividual jurisdictions plan to adopt and build onthe existing ISA 700, such as the EU or nationalstandard-setters, high-quality standards arean important condition for consistency andstandard-setting transparency.Prior research has repeatedly indicated the

existence of an expectation gap between auditors(who are knowledgeable about the purpose andimplications of an audit) and users regarding thereliability of audited financial statements and therespective responsibilities that managers andauditors assume. This phenomenon has beendemonstrated across time and national borders,and even despite various institutional changes tonarrow the gap. Indeed, the revised ISA 700auditors report represents one such attempt.

Table 4: Test of H1c and H2c (financial statement reliability index)

Panel A: Adjusted means for financial statement reliability index

Group Mean S.E. N


Panel B: ANCOVA results for financial statement reliability index


df Meansquare

F-value p-value


Error 300.70 205 1.47


S.E. p-value




Hence, this study was conducted to empiricallyassess the current state of the expectation gapunder the revised ISA 700 by investigating theperceptions of a sample of German auditors andfinancial statement users.Consistent with our prediction, we found that

users ascribed far greater responsibility for thefinancial statements to auditors than did auditors,who ascribed almost no responsibility to theirown profession. Although we expected a greatergap between auditors and students whencompared to the gap between auditors andanalysts, the results are not significant in thisregard. These findings are in line with recent USliterature on the audit expectation gap, whereReckers et al. (2007) found that auditors attitudestoward the public accounting profession weremore favorable than those of judges, law students,and MBA students, and where Butler et al. (2010)identified important differences between auditorsand investors perceptions of the intended

meanings of key terms used to define auditorsresponsibilities.Considering the assessment of managements

responsibility, we observed a gap between auditorsand users, which was driven by a somewhat lowerrating by the group of less experienced financialstatement users (students). However, it isimportant to note that all three groups assigned arelatively high level of responsibility (all meansgreater than 6 on a 7-point scale) to management,indicating a clear narrowing of the expectation gapin this regard as compared to prior research.Finally, with respect to perceived financialstatement reliability, all three groups provided amoderate rating (around 4 on a 7-point scale),although students rated reliability somewhathigher than auditors and analysts. Again, weconclude that the gap in this dimension is relativelyminor.These findings suggest that the audit expectation

gap persists but appears to be largely driven by

Table 5: Test of RQ1a (auditor responsibility index)

Panel A: Adjusted means for auditor responsibility index

Group Report type Mean S.E. N

Auditor Opinion-only 1.75 0.14 86Complete 1.72 0.16 71Overall 1.74 0.12 157

Financial analyst Opinion-only 4.03 0.16 56Complete 4.17 0.17 47Overall 4.10 0.13 103

Student Opinion-only 4.03 0.14 105Complete 3.83 0.14 95Overall 3.93 0.11 200

Overall Opinion-only 3.27 0.08 247Complete 3.24 0.08 213Overall 3.26 0.06 460

Panel B: ANCOVA results for auditor responsibility index


df Meansquare

F-value p-value

Corrected model 719.89 10 71.99 58.17 0.000Group 294.67 2 147.34 119.05 0.000Report type 0.11 1 0.11 0.09 0.765Group Report type 2.00 2 1.00 0.81 0.446Covariates:Age 3.64 1 3.64 2.94 0.087Gender 0.60 1 0.60 0.48 0.487Self-reported reading intensity of financial statements 0.24 1 0.24 0.19 0.661Self-reported reading intensity of auditors report 3.19 1 3.19 2.58 0.109Perceived audit profession reputation 17.25 1 17.25 13.94 0.000

Error 555.69 449 1.24



discrepant expectations between auditors and usersregarding the formal responsibilities and abilitiesof the auditor. However, the gap is smaller withregard to management responsibilities and theperceived reliability of audited financialstatements.We further investigated the effect of the

explanations provided in the ISA 700 auditorsreport by comparing user perceptions based on thecomplete report that contained these explanationsversus an opinion-only version that did not. Thisapproach allowed us to test whether or not thepresence of the explanations mandated by therevised ISA 700 would lead to an alignment ofusers perceptions toward those of auditors. Ourresults suggest that, despite the underlyingobjective of the revised ISA 700 auditors report,the presence of the (quite extensive) ISA 700explanations does not affect the expectation gap.This study has some limitations that should be

acknowledged. First, a maintained assumption is

that auditors perceptions can be considered abenchmark for the normative responsibilities andfinancial statement reliability. Although auditorsare likely to have self-interest and may alsorespond strategically to questions regarding theirown responsibilities, we suggest that it is plausibleto assume that auditors are the best professionavailable to use as a benchmark in this regard, dueto their in-depth knowledge of the audit process.Indeed, this is a maintained assumption in mostprior empirical research on the audit expectationgap. Second, our study is limited to comparing theperceptions of auditors, financial analysts, andstudents, and we acknowledge that there may beother stakeholders whose perceptions are alsoimportant to consider.21 Finally, to judge whetherthere has been a true change in perceptions, somemight argue that the pre-ISA 700 auditors reportwould be amore suitable alternative condition thanour opinion-only version of the ISA 700 auditorsreport that excludes all explanations. However, as

Table 6: Test of RQ1b (management responsibility index)

Panel A: Adjusted means for management responsibility index






Panel B: ANCOVA results for management responsibility index


df Meansquare

F-value p-value


Error 117.50 449 0.26



discussed extensively in note 4, this alternativedesign would have involved severe problems andlost the particular benefits of our design choice.Overall, while the fact that an audit expectation

gap still exists under the new ISA 700 auditorsreport is in line with our expectations, it isdisconcerting that the (quite detailed) explanationsof auditor versus management responsibilities andof the nature, scope, and procedures of the audit donot favorably affect the gap. This observation mayindicate that the explanations would need to beformulated more explicitly and clearly, or even thatusers perceptions are not malleable by additionalexplanations in the auditors report. The latterconcern is also supported by a recent qualitativestudy by Gray et al. (2011), which similarly suggeststhat financial statement users only consider theactual audit opinion (as long as it is unqualified)as the relevant signal, but that they disregard thelong text of the report and the explanationsincluded beyond the opinion itself. This may

suggest that wording changes are not the solutionto overcoming the expectation gap, possiblybecause users expectations are determined byrather stable preconceptions.22 At this stage,however, any specific recommendations arespeculative and require further research forvalidation.

ACKNOWLEDGMENTS

This project is part of the Auditing Standards Boardand International Auditing and AssuranceStandards Board sponsored research onUnqualified Auditors Report Communications.We are grateful for the opportunity to be part of thisresearch initiative as well as the funding andresearch support provided. We appreciate valuablecomments provided by Douglas Prawitt, MarkTaylor, Stuart Turley, participants at the 5thEARNet Symposium 2009 in Valencia (where anearlier version of this paper won the Best Paper

Table 7: Test of RQ1c (Financial statement reliability index)

Panel A: Adjusted means for financial statement reliability index






Panel B: ANCOVA results for financial statement reliability index


df Meansquare

F-value p-value


Error 676.00 449 1.51



Award), members of theAuditing Standards Board,members of the International Auditing andAssurance Standards Board, and the editor andanonymous reviewer of the International Journal ofAuditing. We thank the Institute of GermanAuditors (IDW), two Big 4 audit firms, the Societyof Investment Professionals in Germany (DVFA)for their collaboration, and the London School ofEconomics for research support. Finally, we thankEvelien Smits for her programming assistance andtranslation work.

NOTES

1. For conceptual inquiry into the expectationgap, see Dennis (2010b); for a critical debate onthe effectiveness of the audit function given thepresence of an expectation gap, see Humphreyand Owen (2000).

2. All European Union (EU) member states arerequired to replace their existing nationalauditors reports with the new format as soonas the European Commission adopts aninternational auditing standard that covers thesame subject matter as a national standard(European Parliament and Council, 2006). InGermany, the implementation of the ISA 700into IDW PS 400 was in the process of beingundertaken at the time of the study. For thepurposes of this study, the unpublished draftversion of the German translation of the ISA700 was used, which was kindly provided bythe Institute of German Auditors (IDW).

3. Spanish evidence on the usefulness of theauditors report to make investment andlending decisions is provided by DurndezGmez-Guillamn (2003). Instead ofinvestigating the persistence of the auditexpectation gap after an unqualified auditorsreport, he compares the effect of different auditopinions (i.e., clean, qualified, adverse ordisclaimer).

4. We considered a variety of alternative designchoices, including the possibility of using thepre-ISA 700 report instead of the short-form,opinion-only version. We concluded that thispossibility was less appropriate for twoimportant reasons. First, with such an approachit would have been impossible to accuratelyattribute any effects to the specific underlyingdrivers, since we would have concurrentlyvaried many different factors (because therevision of ISA 700 contained numerous

wording changes within the explanationssection of the auditors report), but not thepresence/absence of explanations as such thatwe are interested in (and which is a clear-cutfactor). Second, as outlined in the literaturereview, prior research evidence is mixed,suggesting that the effects of explanations maybe rather small, dysfunctional, or entirelyabsent. We wished to address this possibility inour research design. If the pre-ISA 700 reporthad been used as the control condition, findingno effect of the new report would not beinformative, because this could be attributed tolack of statistical power. Should, however, noeffect be observed with our current design(which also features control for participantsreading intensity), we believe this would bestrong evidence for a lack of effectiveness of theexplanations, because the same (differences in)expectations would be achieved without anyadditional information beyond the mere auditopinion. Hence, our design choice allows formore informative findings in the case ofobserving no effects.

5. After participants moved from one set ofquestions to a subsequent set, they could not goback and change their previous answers. Thiswas important to ensure that answers to thedependent variables could not subsequently bechanged after having read the manipulationcheck questions. In addition, participants couldnot go back to the previous page of the auditorsreport after they were presented with thequestions. In this way, we ensured thatparticipants reported their true perceptionsafter having been exposed to the auditorsreport. Otherwise, they would have been ableto match their answers to the detailedexplanations contained in the auditors report,which they might have re-read after knowingthe questions.

6. The number of financial analysts contactedvia the German Society of InvestmentProfessionals (DVFA) is unknown to us.

7. This outcome is robust using either principalcomponents analysis with varimax rotation,which assumes uncorrelated factors, ormaximum-likelihood analysis with promaxrotation, which allows the factors to becorrelated.

8. We did not include experience-related variablesas potential covariates because we largelycontrol for these attributes by splitting the



sample into auditors, financial analysts, andstudents.

9. First, age is positively correlatedwith perceivedmanagement responsibility and negativelyassociated with perceived auditorresponsibility and financial statement reliability(all p-values < 0.01). Second, gender issignificantly associated with perceivedmanagement responsibility (p < 0.01), such thatmens mean perception rating is higher (6.63)than womens mean rating (6.41). Third, weasked participants how intensively they readthe financial statements (1 = did not read them/skipped them; 2 = scanned them/read themdiagonally; 3 = read them fairly thoroughly(word by word/number by number); and4 = read them very thoroughly (e.g., severaltimes/tried to memorize)). This measure is notcorrelated with any of the three dependentvariables. Fourth, using the same scale, weasked participants how intensively they readthe auditors report. This measure is positivelycorrelated with the dependent measure ofperceived management responsibility andnegatively correlated with perceived auditorresponsibility (both p-values < 0.05). Finally,we asked participants about their perceivedaudit profession reputation on a 7-point scale(ranging from 1 = low to 7 = high). Thismeasure is negatively correlatedwith perceivedauditor responsibility (p < 0.01), positivelycorrelated with perceived managementresponsibility (p < 0.01), and negativelycorrelated with perceived financial statementreliability (p < 0.05). Hence we control for theseeffects in all upcoming analyses.

10. With respect to the variable measuringself-reported reading intensity of the auditorsreport, it is possible that responses fromparticipants reporting that they did not readthe auditors report at all (n = 12) would distortthe validity of the experiment. After omittingthese observations, we re-ran all statistical testsof the hypotheses and the research question.All results remain equivalent; hence, we do notreport these sensitivity analyses in furtherdetail.

11. As a result, the number of observations fortesting H1 and H2 is 124 for auditors, 39 forfinancial analysts, and 57 for students.

12. There is no significant difference between themean ratings provided by financial analystsand students (p = 0.63).

13. Results are equivalent when omittingcovariates from the ANCOVA model andpost-hoc analyses.

14. When omitting covariates from thedifference-in-difference analysis (hence,employing raw means rather than adjustedmeans), the difference between auditors andstudents (D = 2.41) is significantly smaller thanthe difference between auditors and financialanalysts (D = 2.74; t-statistic = 1.83; p = 0.03;one-tailed), but since the difference is rathertrivial and is only observed in the ANOVAmodel without the control variables, we do notdiscuss it any further.

15. Results are equivalent when omittingcovariates from the ANCOVA model andpost-hoc analyses, with the exception that thedifference between auditors and financialanalysts is then also marginally significant(p < 0.10).

16. Results are equivalent when omittingcovariates from the difference-in-differenceanalysis.

17. Results are equivalent when omittingcovariates from the ANCOVA model,the post-hoc comparisons, and thedifference-in-difference analysis, with theexception that the significance levels forthe group main effect (p < 0.01) and thedifference between auditor and student ratings(p < 0.05) are greater.

18. Results are equivalent when omittingcovariates from the ANCOVA model.

19. Results are equivalent when omittingcovariates from the ANCOVA model, exceptthat the difference between auditors andfinancial analysts is greater (p < 0.05).

20. When omitting covariates from theANCOVA model, the group main effect turnssignificant (p = 0.001), and non-tabulated meancomparisons reveal that students unadjustedmean (4.31) is significantly greater than bothauditors unadjusted mean (4.02; p < 0.05) andfinancial analysts unadjusted mean (3.77;p < 0.001).

21. See, e.g., Ruhnke, Schmiele and Schwind(2010) for survey-based evidence on generalexpectations regarding the auditors work ofsome additional stakeholder groups inGermany, including university professors,business press journalists, or bank officers.

22. For a general discussion on the clarity of auditstandards, see Dennis (2010a).



REFERENCES

Anderson, B., Maletta M. & Wright, A. (1998),Perceptions of auditor responsibility: Views of thejudiciary and the profession, International Journal ofAuditing, Vol. 2, No. 3, pp. 21532.

Bailey, K. E., Bylinski, J. H. & Shield, M. D. (1983),Effects of auditors report wording changes on theperceived message, Journal of Accounting Research,Vol. 21, No. 2, pp. 35570.

Best, P., Bucky, S. & Tan, C. (2001), Evidence of theaudit expectation gap in Singapore, ManagerialAuditing Journal, Vol. 16, No. 3, pp. 13444.

Brown, T., Hatherly, D. & Innes, J. (1993), The reviewreport: An empirical investigation, Accounting andBusiness Research, Vol. 24, No. 93, pp. 1118.

Butler, S. A., Ward, B. & Zimbelman, M. F. (2010), Anexploration of differences in auditors and usersperceptions of key terms used to define auditorsresponsibilities, International Journal of Accounting,Auditing and Performance Evaluation, Vol. 6, No. 1,pp. 80107.

Chong, K.-M. & Pflugrath, G. (2008), Do differentauditors report formats affect shareholders andauditors perceptions? International Journal ofAuditing, Vol. 12, No. 3, pp. 22141.

Dennis, I. (2010a), What do you expect? Areconfiguration of the audit expectation gap,International Journal of Auditing, Vol. 14, No. 2, pp.13046.

Dennis, I. (2010b), Clarity begins at home: Anexamination of the conceptual underpinnings of theIAASBs Clarity Project, International Journal ofAuditing, Vol. 14, No. 3, pp. 294319.

Dewing, I. P. & Russell, P. O. (2002), UK fundmanagers, audit regulation and the newaccountancy foundation: Towards a narrowing ofthe audit expectation gap, Managerial AuditingJournal, Vol. 17, No. 9, pp. 53745.

Durndez Gmez-Guillamn, A. D. (2003), Theusefulness of the auditors report in investment andfinancing decisions, Managerial Auditing Journal,Vol. 18, No. 6/7, pp. 54959.

Epstein, M. J., & Geiger, M. A. (1994), Investor viewsof audit assurance: Recent evidence of theexpectation gap, Journal of Accountancy, Vol. 177,No. 1, pp. 6066.

European Parliament and Council (2006), Directive2006/43/EC of the European Parliament and of theCouncil of 17 May 2006 on statutory audits ofannual accounts and consolidated accounts,amending Council Directive 78/660/EEC and83/349/EEC and repealing Council Directive84/253/EEC, available at: http://ec.europa.eu/internal_market/auditing/directives/index_en.htm (accessed 26 April 2011).

Fadzly, M. N. & Ahmad, Z. (2004), Audit expectationgap: The case of Malaysia, Managerial AuditingJournal, Vol. 19, No. 7, pp. 897915.

Frank, K. E., Lowe, D. J. & Smith, J. K. (2001), Theexpectation gap: Perceptual differences between

auditors, jurors and students, Managerial AuditingJournal, Vol. 16, No. 3, pp. 14549.

Gay, G., Schelluch, P. & Reid, I. (1997), Usersperceptions of the auditing responsibilities for theprevention, detection and reporting of fraud, otherillegal acts and errors, Australian AccountingReview, Vol. 7, No. 13, pp. 5161.

Gay, G., Schelluch, P. & Baines, A. (1998), Perceptionsof messages conveyed by review and auditorsreports, Accounting, Auditing & AccountabilityJournal, Vol. 11, No. 4, pp. 47294.

Gramling, A. A., Schatzberg, J. W. & Wallace, W. A.(1996), The role of undergraduate auditingcoursework in reducing the expectation gap, Issuesin Accounting Education, Vol. 11, No. 1, pp. 13160.

Gray, G. L., Turner, J. L., Coram, P. J. & Mock, T. J.(2011), Perceptions and misperceptions regardingthe unqualified auditors report by financialstatement preparers, users, and auditors,Accounting Horizons, Vol. 25, No. 4, pp. 65984.

Hasan, M., Maijoor, S., Mock, T. J., Roebuck, P.,Simnett, R. & Vanstraelen, A. (2005), The differenttypes of assurance services and levels of assuranceprovided, International Journal of Auditing, Vol. 9,No. 2, pp. 91102.

Hatherly, D., Innes, J. & Brown, T. (1991), Theexpanded auditors report: An empiricalinvestigation, Accounting and Business Research, Vol.21, No. 84, pp. 31119.

Holt, G. &Moizer, P. (1990), The meaning of auditorsreports, Accounting and Business Research, Vol. 20,No. 78, pp. 11121.

Houghton, K. A. (1987), True and fair view: Anempirical study of connotative meaning,Accounting, Organizations and Society, Vol. 12, No. 2,pp. 14352.

Humphrey, C. & Owen, D. (2000), Debating thepower of audit, International Journal of Auditing,Vol. 4, No. 1, pp. 2950.

Humphrey, C., Moizer, P. & Turley, S. (1992), Theaudit expectation gap: Plus a change, plus cest lamme chose? Critical Perspectives in Accounting,Vol. 3, No. 2, pp. 13761.

Humphrey, C., Moizer, P. & Turley, S. (1993), Theaudit expectation gap in Britain: An empiricalinvestigation, Accounting and Business Research, Vol.23, No. 91A, pp. 395411.

Humphrey, C., Loft, A. & Woods, M. (2009), Theglobal audit profession and the internationalfinancial architecture: Understanding regulatoryrelationships at a time of financial crisis,Accounting, Organizations and Society, Vol. 34, No.67, pp. 81025.

Innes, J., Brown, T. & Hatherly, D. (1997), Theexpanded auditors report A research studywithin the development of SAS 600, Accounting,Auditing and Accountability Journal, Vol. 10, No. 5,pp. 70217.

International Auditing and Assurance StandardsBoard (IAASB) (2004), Project History: TheIndependent Auditors Report on a Complete Set of



General Purpose Financial Statements Completed. Available at: http://www.ifac.org/IAASB/ProjectHistory.php?ProjID=0013 (accessed10 September 2010).

International Federation of Accountants (IFAC) (2008),International Standard on Auditing (ISA) 700: Theindependent auditors report on a complete set ofgeneral purpose financial statements, Handbookof International Auditing, Assurance, and EthicsPronouncements, Part I, New York: IFAC, pp. 56179.

Kelly, A. S. & Mohrweis, L. C. (1989), Bankers andinvestors perceptions of the auditors role infinancial statement reporting: The impact of SASNo. 58, Auditing: A Journal of Practice & Theory, Vol.9, No. 1, pp. 8797.

Kirk, N. E. (2006), Perceptions of the true and fairview concept: An empirical investigation,ABACUS, Vol. 42, No. 2, pp. 20535.

Kneer, D. C., Reckers, P. M. J. & Jennings, M. M.(1996), An empirical examination of the influenceof the new US auditors report and fraudred-flags on perceptions of auditor culpability,Managerial Auditing Journal, Vol. 11, No. 6, pp. 1830.

Koh, H. C. & Woo, E. (1998), The expectation gap inauditing,Managerial Auditing Journal, Vol. 13, No. 3,pp. 14754.

Libby, R. (1979), Bankers and auditors perceptionsof the message communicated by the auditorsreport, Journal of Accounting Research, Vol. 17, No. 1,pp. 99122.

Lin, Z. J. & Chen, F. (2004), An empirical study of theaudit expectation gap in the Peoples Republic ofChina, International Journal of Auditing, Vol. 8, No. 2,pp. 93115.

Lowe, D. (1994), The expectation gap in the legalsystem: perception differences between auditorsand judges, Journal of Applied Business Research, Vol.10, pp. 3944.

Manson, S. & Zaman, M. (2001), Auditorcommunication in an evolving environment: Goingbeyond SAS 600 auditors report on financialstatements, British Accounting Review, Vol. 33, No. 2,pp. 11336.

McEnroe, J. & Martens, S. C. (2001), Auditors andinvestors perceptions of the expectation gap ,Accounting Horizons, Vol. 15, No. 4, pp. 34558.

Miller, J. R., Reed, S. A. & Strawser, R. H. (1993),Bank loan officers perceptions of the newauditors report, Accounting Horizons, Vol. 7, No. 1,pp. 3952.

Monroe, G. S. & Woodliff, D. R. (1993), The effect ofeducation on the audit expectation gap, Accountingand Finance, Vol. 33, No. 1, pp. 6178.

Monroe, G. S. & Woodliff, D. R. (1994), An empiricalinvestigation of the audit expectation gap:Australian evidence, Accounting and Finance, Vol.34, No. 1, pp. 4774.

Nair, R. D. & Rittenberg, L. E. (1987), Messagesperceived from audit, review, and compilationreports: Extension to more diverse groups,Auditing: A Journal of Practice, Vol. 7, No. 1, pp. 1538.

hman, P., Hckner, E., Jansson, A. M. & Tschudi, F.(2006), Swedish auditors view of auditing: Doingthings right versus doing the right things, EuropeanAccounting Review, Vol. 15, No. 1, pp. 89114.

Porter, B. (1993), An empirical study of the auditexpectation-performance gap, Accounting andBusiness Research, Vol. 24, No. 93, pp. 4968.

Reckers, P. M. J., Jennings, M., Lowe, D. J. & Pany, K.(2007), Judges attitudes toward the publicaccounting profession, European Accounting Review,Vol. 16, No. 3, pp. 62545.

Ruhnke, K., Schmiele, C. & Schwind, J. (2010), DieErwartungslcke als permanentes Phnomen derAbschlussprfung: Definitionsansatz, empirischeUntersuchung und Schlussfolgerungen, Zeitschriftfr betriebswirtschaftliche Forschung, Vol. 62, No. 4,pp. 394421.

Schelluch, P. (1996), Long form audit-reportmessages: Further implications of the auditexpectation gap, Accounting Research Journal, Vol. 9,No. 1, pp. 4855.

Sidani, Y. M. (2007), The audit expectation gap:Evidence from Lebanon, Managerial AuditingJournal, Vol. 22, No. 3, pp. 288302.

Siddiqui, J., Nasreen, T. & Choudhury-Lena, A. (2009),The audit expectations gap and the role of auditeducation: the case of an emerging economy,Managerial Auditing Journal, Vol. 24, No. 6,pp. 56483.

AUTHOR PROFILES

Anna Gold is an associate professor at VUUniversity Amsterdam and teaches auditingand accounting information systems at the BSc andMSc levels. Her main research interest lies inbehavioral accounting research, with a focus onjudgment and decision making in auditing. She hasconducted studies in various areas, such as biasesin auditor-client inquiry, fraud brainstorming,fraud consultation, the audit expectation gap, theimpact of organizational factors on auditorsbehavior, and auditor rotation.Ulfert Gronewold is a full professor at the

University of Potsdam and teaches accounting andauditing in the private and public sector at theBSc and MSc levels. His research comprisespsychology-based behavioral accounting andauditing studies and covers various topics such asauditors judgment and decision making, theimpact of organizational factors on auditorsbehavior, the audit expectation gap, and issues inaudit methodology.Christiane Pott is an assistant professor at the

Mnster Institute of Accounting and Taxation atthe University of Mnster, School of Business



Administration and Economics. Her experimentaland archival research is related to accountingregulation and corporate governance. Her researchcovers areas such as transparency reporting,sustainability reporting, auditor compensation, theaudit expectation gap, and auditor rotation.

APPENDIX

Experimental case with auditors reportmanipulation (opinion-only versuscomplete report)*

In the following, you will obtain information aboutBerens Electronics AG and the auditors report ofits financial statements auditor. Upon reading thecase, you will be asked a set of questions.

Berens Electronics AG

Berens Electronics is a large publicly tradedcompany that manufactures and distributes audio,video, and other multimedia equipment to retailersthroughout Europe. Berens Electronics hascompleted the fiscal year 2007 and published theIFRS consolidated financial statements, which areoutlined in the following:


Consolidated Balance Sheet

12/31/2007in million

12/31/2006in million

Non-current assets 600 596Current assets 490 479Thereof cash and cash

equivalents156 150

Thereof accountsreceivables

191 189

Thereof inventory 143 140Total assets 1,090 1,075Equity 640 638Liabilities 450 437Total equity andliabilities

1,090 1,075


Consolidated Income Statement

2007 inmillion

2006 inmillion

Sales 1,300 1,181Cost and expenses 1,201 1,084Operating profit 99 97Financial income andexpenses

2 1

Profit before tax 101 98

Berens Electronics financial statements have beenaudited by the audit firm G&B for the precedingthree years. G&B is a large audit firm with a goodreputation.In the following, this years auditors report of

G&B relating to the audit of this years IFRSconsolidated financial statements according toInternational Standards on Auditing (ISA) isreproduced:[Experimental Manipulation][The following text was shown in the complete

report auditors report condition:]

Independent auditors report

To Berens Electronics AG

We have audited the accompanying consolidatedfinancial statements for the year from January 1,2007 to December 31, 2007.

Managements responsibility for thefinancial statements

The management of the company is responsiblefor the preparation and fair presentation ofthese financial statements in accordance withInternational Financial Reporting Standards asadopted by the European Union. Thisresponsibility includes: designing, implementing and maintaining

internal controls relevant to the preparation andfair presentation of the financial statements,which are to be free frommaterial misstatement,whether due to fraud or error;

selecting and applying appropriate accountingpolicies; and

making accounting estimates that arereasonable in the circumstances.

* The original versions used in this study were in German. Here,we provide an English translation.



Auditors responsibility

Our responsibility is to express an opinion onthe financial statements based on our audit. Weconducted our audit in accordance withInternational Standards on Auditing. Thosestandards require that we comply with ethicalrequirements and plan and perform the audit toobtain reasonable assurance that the financialstatements are free from material misstatement.An audit involves performing procedures to

obtain audit evidence about the amounts anddisclosures in the financial statements. Theprocedures selected depend on the auditorsjudgment, including the assessment of the risks ofmaterial misstatement of the financial statements,whether due to fraud or error. In making these riskassessments, the auditor considers internal controlrelevant to the entitys preparation and fairpresentation of the financial statements to designaudit procedures that are appropriate in thecircumstances but not for the purpose ofexpressing an opinion on the effectiveness of theentitys internal control. An audit also includesevaluating the appropriateness of accountingpolicies used and the reasonableness of accountingestimates made by management as well asevaluating the overall presentation of the financialstatements.We believe that the audit evidence we have

obtained is sufficient and appropriate to provide abasis for our audit opinion.

Opinion

In our opinion, the financial statements give a trueand fair view of the financial position of Berens

Electronics AG, as of December 31, 2007, and of itsfinancial performance and cash flows for the yearfrom January 1, 2007 to December 31, 2007, inaccordance with International Financial ReportingStandards.

G&B(Public Accounting Firm)

[]

[The following

Documents

Gold Gronewold Pott IJA2012