The impact of professionaThe role of availability an

Darius Fatemi a, John HasseldineaHaile/US Bank College of Business, Northern Kentuckb Paul College of Business and Economics, University ocKelley School of Business, Indiana University, United

a r t i c l e i n f o

Article history:Available online 3 March 2014

Keywords:AvailabilityCode of conduct

. In contrast, whend with an uve responsendard. We c

that two AICPA standards (as currently worded) are best understood when they aregated. Whether this nding holds for professionals is an empirical question for

http://dx.doi.org/10.1016/j.racreg.2014.02.0031052-0457/ 2014 Elsevier Ltd. All rights reserved.

Corresponding author. Tel.: +1 603 862 3342; fax: +1 603 862 3383.E-mail addresses: Fatemid1@nku.edu (D. Fatemi), john.hasseldine@

unh.edu (J. Hasseldine), hitep@indiana.edu (P. Hite).

Research in Accounting Regulation 26 (2014) 2639

Contents lists available at ScienceDirect

Research in Accounting Regulation

journal homepage: www.elsevier .com/ locate/ racregAICPA Rule 102-6 allowing client advocacy, they report conservativelythe prime is a joint presentation of the standards, participants respondecision, which differs signicantly from the consistently conservaticontrol group as well as by the participants primed with an isolated stanbiasedby the

oncludeaggre-futureHence, a joint presentation of countervailing standards should result in a more balancedjudgment, reecting neither a conservative nor pro-client tendency.The psychology literature suggests that heuristics, such as availability priming and com-

parative analysis, are more likely to affect novice decision makers (e.g., jurors, clients, newhires, students) than experts whose work experiences could drive the results. This studyexamines the responses of upper-level accounting majors, and the results show that theparticipants are inclined toward conservative decision making. Participants exposed to aseparately-stated standard for integrity respond conservatively, just as they do in a controlgroup without explicit access to the professional standard. Similarly, even when exposed toClient advocacyIntegrityProfessional standardsConservatisml standards on accounting judgments:d comparative informationb,, Peggy Hite c

y University, United Statesf New Hampshire, United StatesStates

a b s t r a c t

Given scant research on the inuence of the AICPAs Code of Conduct, this study examinesthe effects of professional standards for advocacy and integrity on a nancial reportingdecision. Based on the availability and priming literature, we test whether the currentwording of two AICPA professional standards inuence nancial reporting decisions. Prioraccounting research has documented cases where professionals were inclined toward aconservative or skeptical bias (Francis & Krishnan, 1999; Jenkins & Lowe, 1999) while otherstudies have documented an inclination toward a client-conrming bias (Hackenbrack &Nelson, 1996; Roberts, 2010). Our study examines whether using AICPA ethical standardsas primes results in a neutral, unbiased nancial reporting decision in a context in whichthere is substantial, yet inconclusive, evidence. Roberts (2010) documents the tendency forprofessionals to view integrity and advocacy as segregated objectives: one for promotingunbiased reporting, associated frequently with accounting-related decisions, and the othercondoning client advocacy, typically associated with tax-related judgments. Hence, we testfor availability effects based on separately-stated standards. However, the literature oncomparative analysis explains that a combined concept containing counterbalancing fea-tures allows the participant to form causal relationships between the distinguishing com-ponents. This type of mental process brings the causal knowledge into working memory.

research. The implication is that accountants decision making could be enhanced by arevised professional standard reminding them to jointly consider the goals of unbiaseddecision making and justiable client advocacy.

2014 Elsevier Ltd. All rights reserved.

to preserve a legacy of honor and integrity for future gen-erations of CPAs. The profession must buitional values such as rigorous commitme

Professional integrity is a desirable social norm, and it isrelevant to the wider debate on principles- versus rules-based accounting, with Sunder (2010) arguing that what-ever written rules exist, there is a need to give social normsa stronger role in restoring professional and personalresponsibility in accounting. This sentiment is consistentwith the AICPA presidential viewpoints on professional

1 These AICPA standards were also adopted byAccounting Oversight Board as its interim standards

ing and ones current level of moral development (Herron & Gilbertson,2004; Scoeld, Phillips, & Bailey, 2004). Bailey, Phillips, and Scoeld (2005)

D. Fatemi et al. / Research in Accounting Regulation 26 (2014) 2639 27ld upon its tradi-nt to integrity.

the Public Companyin April 2003.

assert that more research is needed to identify factors that mediate moralreasoning for accounting-related judgments. Given the costly and timeconsuming aspects of the DIT, the present study does not examine the DIT.Instead, our study builds on prior research regarding accountants propen-sity towards implicit biases and tests how standards in the current AICPACode of Conduct might inuence accountants decision making.47) reports a speech by (then) AICPA President Barry Mel-anchon stating that the professions leadership must act

Results, so far, have shown that the effectiveness of DIT applications to theaccounting environment vary with participants functional area of account-Introduction

Current professional standards issued by the AmericanInstitute of Certied Public Accountants (AICPA) includetechnical guidance aswell as a Codeof Professional Conduct.The Code applies to all AICPA members whether they areemployed in public practice, industry, government or edu-cation. After reviewing the history of the AICPAs develop-ment of a set of ethical rules, Voynich (2005) notes thatthe existence of a code of conduct is unique to a professionand should guide behavior. Likewise, Libby and Luft (1993)reiterate the importance of the professional Code, and arguethat the primary purpose of the Code is to inuence profes-sional decision making. To date, however, little is knownabout the behavioral effect of the professional standards.

The Code serves as the professions overriding ethicalguidance for applying technical standards whether theyare principles-based or rules-based (AICPA, 2010; Moehrle,Previts, & Reynolds-Moehrle, 2006), and it emphasizes theimportance of amembers integrity as it is the quality fromwhich the public trust derives and is a benchmark againstwhich a member must ultimately test all decisions (AICPACode of Conduct, Section 54-Article III). In addition to thespecic rule for integrity, the AICPA Code endorses Rule102-6, which allows a member to act as a client advocatein support of the clients position on accounting or nan-cial reporting issues, either within the rm or outside therm with standard setters, regulators, or others (AICPACode of Conduct, Rule 102-6).1 Prior literature documentsthe tendency for these standards to be perceived as segre-gated objectives: one for promoting unbiased reporting andthe other for allowing client advocacy (Jenkins & Lowe,1999; Roberts, 2010). We posit that these two standardsare best understood when they are aggregated.

Although there has long been interest in measuringaccountants ethical attitudes and behavior (Loeb, 1971),it was a series of corporate accounting scandals that ledthe U.S. Securities and Exchange Commission to requiredisclosure regarding whether companies have adoptedcodes of ethics, and if not, why not. In addition, the Associ-ation to Advance Collegiate Schools of Business (AACSB,2004) began to require that students learn ethics as a partof their business degree at an AACSB-accredited institu-tion. Immediately after Enron, academe and practice in-creased their focus on accounting ethics. Smith (2003, p.integrity expressed a decade ago and the importance ofrestoring the publics trust in accounting.

While the Code prohibits any action or lack of actionthat knowingly misrepresents nancial information, it alsocondones client advocacy. Given the complexity of techni-cal guidelines for accountants, expertise in interpretingnancial rules is clearly a valued service. AICPA Rule 102-6 allows professional accountants to be an advocate fortheir clients when it is appropriate to do so. In fact, priorresearch on conrmation bias in nancial accounting deci-sion making has documented the tendency for some pro-fessionals to exhibit pro-client tendencies when clientpreferences are made explicit (Hackenbrack & Nelson,1996; Hateld, Jackson, & Vandervelde, 2011; Kadous,Kennedy, & Peecher, 2003; Roberts, 2010). In the absenceof stated client preferences, research has found that expe-rienced professionals tend to respond with conservative,income-decreasing outcomes (Francis & Krishnan, 1999;Jenkins & Haynes, 2003; Lord, 1992).

If decision makers in accounting have either conserva-tive or pro-client tendencies, unbiased judgments thatare promulgated by the AICPA could be affected. Nelson(2003) argues that professional standards for accurate,unbiased reporting should be designed to incentivizeappropriate decision making, yet no study to date hasexamined whether the professions standards for advocacyand integrity impact nancial decision making.2

The present study examines the impact of current pro-fessional standards for integrity and advocacy on upper le-vel undergraduate students who are not yet familiar withthese standards but soon will be faced with how to apply

2 Some prior research on accounting ethics has been motivated byKohlbergs (1973) theory of cognitive moral development and operation-alized through the use of Rests (1979) Dening Issues Test (DIT). There arenumerous accounting ethics studies using the DIT (e.g., Abdolmohammadi,Read, & Scarbrough, 2003; Bay, 2002; Shaub, 1994; Tsui & Gul, 1996). Thesestudies tend to report a resulting P score by which the ethical attitudesabout a particular sample may be inferred and compared to other samples.

vative rather than pro-client, and this parallels prior

28 D. Fatemi et al. / Research in Accounting Regulation 26 (2014) 2639them. The reaction of novices to the professions standardsis appropriate for measuring initial responses to the word-ing of the standards. Using student subjects is also bene-cial as their reactions are likely to approximate those ofjurors who could be asked to judge whether professionalaccountants had complied with these standards. In con-trast, experienced professionals are likely to have stronginternalized beliefs based on their personal experiencesand workplace norms. Consequently, once new recruits en-ter the profession, they will undoubtedly be inuenced bythe rival dynamics specic to their own discipline (e.g.,audit versus tax) and by the associated implicit and expli-cit incentives such as the inuence of client preference(Cloyd & Spilker, 1999; Hackenbrack & Nelson, 1996;Schweikart, 1992).

In this study, we experimentally explore the effects ofcounterbalancing professional standards for advocacy andintegrity. When confronted with ambiguous issues requir-ing the exercise of professional judgment, one standardmay dominate the other on reporting decisions, or thepresence of both standards could mitigate a directionalbias. Theories on availability suggest that having immedi-ate, explicit access to relevant information allows moreefcient processing of cognitive encoding (Bhattacharjee,Maletta, & Moreno, 2007), and Payne, Bettman, and Johnson(1993) explain that targeted information is more likelyto be used if it is brought into working memory. Hence,participants in the present study are assigned primes forthe separately-stated standards. In addition, research hasshown that using a joint presentation of dissimilar butcomparably relevant concepts draws attention to the dis-tinctive features of each, allowing them to become con-text-dependent by providing boundaries for each concept(Payne et al., 1993; Sumer & Knight, 1996). Using a primethat combines two counterbalancing standards, our studytests the impact of this aggregated presentation.

The results of our experiment show that student partic-ipants tend to be conservative in their reporting choices,which is consistent with prior accounting research usingexperienced accountants. Whether participants are onlyexposed to AICPA Rule 102-6 allowing client advocacy orSection 54 calling for integrity, participants still indicatea conservative tendency consistent with the control group.However, when the combined standards (advocacy andintegrity) are presented to participants, the responses tendto be less biased (operationalized in this study as beingneither pro-client nor conservative on an ambiguous is-sue). Thus, the combination of counterbalancing standardsresults in a signicantly more neutral position than theother groups, reecting neither conservative nor pro-clienttendencies. The implication is that professional decisionmaking could be enhanced by a global professionalstandard recognizing the importance of simultaneouslycomplying with the guideline for professional integritywhile allowing for client advocacy when it is appropriateto do so. Our result showing the effectiveness of acombined professional standard is consistent with thepsychology literature on availability and the literature oncognitive development (which asserts the importance ofcomparative distinctions for more efcient mentalprocessing).research nding that auditors and nancial accountantstend to exhibit conservatism and professional skepticism,perhaps when it is unwarranted (Jenkins & Haynes, 2003;Kieso, Weygandt, & Wareld, 2010; Lord, 1992). Third,when the contextual level is substantial, including sup-portive and non-supportive evidence, the conservative-natured decision makers discount the advocacy standardin favor of a more skeptical response, implying that theadditional facts could be used to challenge a favorable cli-ent decision. Fourth, in the presence of the integrity stan-dard as well as the advocacy standard, participantsresponses do not signicantly differ from those withoutexposure to the standards. The implication is that the iso-lated standards, as currently worded, maintain the statusquo tendency, which in this case is conservative decisionmaking. Especially important for the profession and societyis the fth contribution demonstrating that an aggregatedpresentation of the two counterbalancing standards resultsin the most neutral and least biased decision. The implica-tion is that standard setters should consider combining thetwo standards into one overarching rule that requiresobjective, unbiased decision making, while recognizingthe need to advocate for a client when warranted.

Our contributions have important implications for pro-fessional accounting rms as they seek to encourage pro-fessional behavior for their employees, and these ethicalguidelines will become increasingly important in the U.S.as the profession begins to adopt more principles-basedstandards with increasing reliance on unbiased decisionmaking. In addition, if the purpose of a code of conduct isto do more than window-dressing, then the exact languagein the current standards should be monitored to ensure theintended message is understood. Future research shouldexamine whether alternative wording for the advocacyand integrity standards would increase the effectivenessof the separately-stated standards. Lastly, educators andthose interested in ethics education should benet fromthis research by understanding the need for accountingmajors to study the professional codes and to ensure thatseparately-stated standards are not emphasized in isola-tion of other potentially counterbalancing standards.

The remainder of this article proceeds as follows. Thenext section provides background information and devel-ops our hypotheses. The third and fourth sections outlineour method and results, respectively, followed by someconcluding remarks.

Theory and hypothesis development

Priming

To examine whether AICPA ethical standards affectnancial reporting decisions, we investigate whether theThis study makes several contributions to the literature.The most important contribution is the demonstration thatthe current AICPA Code of Professional Conduct has the po-tential to inuence nancial reporting decisions. Second, inthe absence of a professional code, nancial reporting deci-sions made by the student participants tend to be conser-

affect participants decisions regarding whether the auditorclient relationship threatened the auditors independence ina hypothetical scenario. However, when the DIT was used tocategorize participants as having either low or high levels ofmoral reasoning (Rest, 1979), participants classied as low le-vel (stage 4 or lower on the DIT, representing rule-orientedindividuals) were more likely to believe independence wasthreatened when they had been given the rules-based ex-cerpts related to independence but notwhen given the princi-

D. Fatemi et al. / Research in Accounting Regulation 26 (2014) 2639 29current wording for integrity and advocacy act as effectiveprimes that inuence the level of potential bias in nancialaccounting decision making. Kunda (1990) describes prim-ing as any procedure that brings a particular concept,knowledge structure, attitude or feeling to mind. In addi-tion, Payne et al. (1993) note that primes become morereadily available when they are explicitly stated. Implicitprimes require more cognitive effort as the related conceptneeds to be moved from long-term memory to short-termor working memory. Explicit primes tend to be effectivebecause they reduce the cognitive effort needed to processthe targeted information. We test the effectiveness of twoAICPA standards, those for integrity and advocacy, by oper-ationalizing them as explicitly available primes.

Professional integrity

The unexpected demise of Enron and, later, ArthurAndersen led to a critical view of the accounting profession(Unerman & ODwyer, 2004). Consequently, the UnitedStates Congress enacted the SarbanesOxley Act (SOX).Although SOX started requiring disclosure of the existenceof corporate codes of conduct in 2002, many companieshad already adopted corporate codes. Unfortunately, theeffectiveness of these codes remains unresolved (Cleek &Leonard, 1998; Forster, Loughran, & McDonald, 2009;Hegarty & Sims, 1979; Simons, 2002). Given the rotelanguage of many corporate codes along with the varietyof purposes that different businesses face (e.g., marketingproducts, quality of products, employee rights, customercomplaints), effective corporate codes may be more elusivethan a code of conduct intended to guide professionalaccountants in their application of technical standards.Ultimately, all members of the AICPA, including those inpublic accounting, government, industry, and education(as well as student afliate members), are required to begoverned by its Code (Carnegie & Napier, 2010).

The effectiveness of a 1989 version of the AICPA profes-sional code was tested by Claypool, Fetyko, and Pearson(1990) who concluded that the ethical sensitivity of prac-ticing CPAs was affected by the Code, although they didnot present a copy of the Code to their participants. Instead,they presented six accounting dilemmas to the participantsand then asked them to identify which of six factors weremost important when considering the ethical implicationsof each dilemma. For the dilemmas involving independenceand condentiality, practitioners responded that the twomost important factors were indeed independence andcondentiality (mean rankings of 4.84 and 4.75 out of 5,respectively). In addition, the researchers asked some theo-logians to make the same evaluations, and the theologians(who had no knowledge of the AICPA Code) also rankedindependence and condentiality factors as very important(respective means of 4.71 and 4.45 out of 5).

Herron and Gilbertson (2004) compared the effects ofprinciples-based and rules-based ethical guidelines on thelevel of independence required in the client-auditorrelationship. Some of the student participants were givenexcerpts from the previous AICPA Code with a principles-based focus, and the others were given excerpts that wererules-based. Overall, the treatments did not signicantlyples-basedexcerpts. Conversely, participants classiedashighlevel (stage 5 or higher, representing principled-level reason-ing)weremore likely to believe independencewas threatenedwhen given the principles-based excerpts related to indepen-dence but not when given the rules-based ones. The implica-tion is that the guidance given in professional codes ofconduct is not universally internalized, because its effective-ness depends on whether the code is principles-based orrules-based and whether the participants level of DIT moralreasoning is high or low. Applying this result to practice isdifcult as each practitioners level of moral reasoning is notknown, yet the nding has important implications. Forexample, if the highest stages of DIT represent principledthinking, why do those at the highest level of moral reasoningnot maintain principled thoughts in the presence of rules-basedguidance?This result suggests thatevenethicaldecisionmakers may lose focus on the guiding principle when theybecome too engaged in the check-the-box (rules-based)guidance.

The above studies focused primarily on cases of inde-pendence between client and auditor, with the underlyingassumption that unwarranted, client-favorable decisionswill result in misleading nancial statements. Previous lit-erature, however, has shown that nancial reporting alsoerrs on the side of conservatismwhich, if unwarranted, rep-resents another form of biased reporting. When strong cli-ent preference or importance is not explicitly stated,studies have shown that experienced professionals, espe-cially managers and partners, tend to indicate income-decreasing outcomes involving nancial or auditing judg-ments (Farmer, Rittenberg, & Trompeter, 1987; Francis &Krishnan, 1999; Jenkins & Haynes, 2003; Lord, 1992; Trom-peter, 1994). This is not surprising given conservatismsearly inclusion in the AICPAs Conceptual Framework forFinancial Reporting, which has been discussed in nearlyall principles and intermediate accounting textbooks (e.g.,Kieso et al., 2010). Moreover, the concept has been incorpo-rated in many GAAP rules, such as the asymmetric recogni-tion of contingent losses relative to contingent gains, andthe lower of cost or market method for valuing inventory.3

In essence, prior research has documented both in-stances of biased reporting: overly-aggressive and overly-

3 Because of the tendency for conservatism to produce bias in nancialreporting, the Financial Accounting Standards Boards (FASB) revisedconceptual framework (cf. Statement of Financial Accounting ConceptsNo. 8) no longer supports the convention, arguing that conservatism iscounter to neutrality and faithful representation (FASB, 2010). However,given that the concept had been taught to the students in this study, that ithas been instrumental in the formulation of currently-existing rules underU.S. GAAP, and that its removal from the framework generated opposition(FASB, 2010, para. BC3.27), its inuence in decision making will likelyextend for years to come.

represents an overarching principle that is intended to

30 D. Fatemi et al. / Research in Accounting Regulation 26 (2014) 2639guide all of its members decision making.Nelson (2003) argues that fundamental principles

should be the goal and that overriding principles could beused to encourage accurate reporting. Consistent with Nel-sons goal, Sunder (2010) discusses the balance betweenuniform,written rules-based standards and principle-basedsocial norms, arguing that the latter should be given a stron-ger role in restoringpersonal andprofessional responsibilityin accounting. Jones, Massey, and Thorne (2003) questionwhether the professions ethical standards enhance or limitauditors ethical sensitivities, stating that more research isneeded to identify ways in which standards can enhanceethical awareness and behavior. In addition, Libby and Luft(1993) assert that the standards were written for the pri-mary purpose of inuencing professional decision making.

A prime for integrity implies that respondents will reactwith a discriminating, unbiased judgment, but Payne et al.(1993) assert that attitude and content-related experiencecould impact the posited priming effects. If participantshave a propensity toward advocacyor belief that substantialpayments to employee-shareholders should be classied asdividends, their responses might favor a net income-increasing decision for dividend reporting. An integrityprime is expected to temper the potentially biased belief to-ward a neutral stance, as long as the prime is sufcient andprior belief is not too ingrained. If participants are not in-clined toward advocacy and they are inclined to believe thatsubstantial payments to owneremployees should be clas-sied as salary, their responses are likely to favor a salarydeduction, a net income-decreasing decision. An integrityprime should result in a neutral stance if the prime is salientenough to offset the propensity toward conservatism.

In Section 54 of the AICPA Code of Conduct, integrity isdescribed as the ultimate benchmark for all membersdecision making. The standard states that integrity canaccommodate . . .the honest difference of opinion; it can-not accommodate deceit or the subordination of principle.Since the standard is intended to mitigate a propensity tomake unwarranted client-favorable decisions or unwar-ranted conservative responses, this study examineswhether exposure to the integrity standard impacts thelevel of pro-client decision making. Thus, the rst hypoth-esis is as follows:

H1. Participants exposed to a prime for professionalintegrity will make more neutral, less biased nancialreporting decisions than those without any exposure to theprofessional standards.

Client advocacy

Prior research has shown that auditors can be inu-enced by client preferences and tend to permit aggressivereporting by favorably interpreting vague facts and stan-conservative. Carey (2008) argued convincingly in 1970that real independence means integrity and objectivity,and the current AICPA standard for integrity calls for unbi-ased, objective reporting. It does not condone a conserva-tive bias any more than it condones a pro-client bias. Itdards (Hackenbrack & Nelson, 1996; Salterio & Koonce,1997), and the same is true for tax practitioners (Ayres,Jackson, & Hite, 1989; Cloyd & Spilker, 1999; Jackson &Milliron, 1989). Cuccia, Hackenbrack, and Nelson (1995)found that tax practitioners responded to high precisiontechnical rules by interpreting the evidence more liberally to their clients benet. Thus, writing any standard,whether it be principles-based or rules-based, may notbe the only solution. Prior research, however, has notincorporated the inuence of an ethical standard or profes-sional code of conduct on judgments involving ambiguousissues (Cuccia et al., 1995; Hackenbrack & Nelson, 1996). Ifprinciples-based technical standards with an override foreconomic substance were to become the norm, an advo-cacy stance, as permitted by AICPA Rule 102-6, could stillresult in uncertainty regarding how an ambiguous transac-tion should be reported.

Bobek, Hageman, and Hateld (2010), in their review ofprior advocacy research, note that most prior studies havenot measured participants advocacy attitudes. Instead,researchers have experimentally created a treatment effectfor advocacy by stating the clients preference (Cloyd &Spilker, 1999; Hateld et al., 2011). Results have beenmixed in that some studies have found a signicant effectfor measured advocacy attitude on the respective tax judg-ments (Johnson, 1993; Levy, 1996), while others have not(Kadous & Magro, 2001).

Pinsker, Pennington, and Schafer (2009) found thatthe effect of decision context (tax versus audit) wasmoderated by the professionals attitude toward advo-cacy, and that advocacy attitude was signicantly lowerfor the auditors than for the tax accountants. The authorsattribute this to the professional skepticism that auditorsare generally expected to exhibit. Their study also re-ported that advocacy attitude signicantly correlatedwith the nal decision, regarding disclosure of a contin-gent liability, for the auditors but not for the tax profes-sionals. A study by Roberts (2010) posited a tendencytoward pro-client decision making in a nancial setting,and he found that auditors conform their professionaljudgments to a clients demand for earnings manage-ment when client preferences are explicit. The resultsof these studies conrm the need to examine the effectsof an advocacy standard not just in tax, but also in nan-cial decision making. Traditionally, accounting rms wereexpected to be unbiased in their audit responsibilitieswhile being advocates in their tax services, but accordingto Colson (2005), the culture has become more charac-terized by advocacy because of the large revenues gener-ated by audit services.

Although prior research has recognized the importanceof advocacy in some tax-related judgments (e.g., Pinskeret al., 2009), its legitimate status as a professional standardin nancial-related judgments has not always been per-ceived as a positive virtue (e.g., the accounting literatureon conrmation bias such as Hackenbrack & Nelson,1996; Hateld et al., 2011; Roberts, 2010). The AICPA hasrepeatedly conrmed the right of CPAs to be advocatesfor their clients, and AICPA Rule 102-6 continues to con-done the role of for advocacy in both tax and nancialreporting decisions.

discussed earlier, studies in nancial decision making have

D. Fatemi et al. / Research in Accounting Regulation 26 (2014) 2639 31documented conservative tendencies as well as pro-clienttendencies. The AICPA Code of Conduct reects a need tocounterbalance these positions by sanctioning client advo-cacy and professional integrity. The intent is to encouragedecision making that is as truthful as possible withoutbeing biased toward advocacy or conservatism. For ambig-uous contexts in which the technical guidance is unclear,the AICPA narrative accompanying Rule 102-6 suggeststhat advocacy can be justied, but there will sometimesbe a line in the sand that should not be crossed. In sum,the two standards are context-dependent as each rule actsas a constraint on the other: one should advocate for theclient, but not at the expense of making biased judgments.Professionals should make objective judgments consistentwith technical guidance but not at the expense of disre-garding the substance of the clients unique facts andcircumstances.

Related to the need to balance advocacy with integrity,Anderson, Marchant, Robinson, and Schadewald (1990)assert that a comparative analysis draws attention to therelevant components of each underlying principle. Suchcomparisons are effective, because they exacerbate thelevel of attention to specic attributes of each concept.A prime for advocacy (AICPA Rule 102-6) suggests thatparticipants will respond with a pro-client judgment, butas noted earlier, Payne et al. (1993) warn that attitudes,knowledge, and related experience could impact positedeffects. For example, if participants have a propensitytoward advocacy or toward classifying payments toemployee-shareholders as dividends, their responses toan advocacy prime is likely to result in either a mere con-rmation of their current response or an increased inclina-tion toward the net income-increasing dividendclassication. On the other hand, if participants are not in-clined toward advocacy or the classication of a substan-tial payment as a dividend, then the response to anadvocacy prime will either be met with resistance, or itcould result in a more client-favorable response. To testthe impact of AICPA Rule 102-6, an advocacy standard,on nancial reporting issues, our second hypothesis is:

H2. Participants exposed to a prime for client advocacywill make less conservative nancial reporting decisionsthan those without any exposure to the professionalstandards.

Combined counterbalancing standards

The AICPA narrative to Rule 102-6 discusses theaccountants need to balance multiple professional stan-dards. The concern is that clients could request profes-sional services that stretch the bounds of performancestandards, and this could then impair the reputation ofthe member and their rm in regards to professional integ-rity. Serious cases may even risk damaging public percep-tion of the accounting profession (Carnegie & Napier,2010). Notwithstanding these extreme situations, the needto balance multiple principles of professional conduct canoccur even in seemingly routine matters of judgment. AsExperts are considered to have more organized knowledgethrough their use of comparisons that encourage theformation of linkages among existing rules. Applied tostudies of knowledge acquisition, education studies havedemonstrated that using examples that are inconsistentor distinctive can be more helpful in the learning processthan are examples that are consistent or redundant(Gorman & Gorman, 1984). Chi, Lewis, Reinmann, andGlaser (1989) demonstrated the effectiveness of writingself-elaborations that include the distinctions betweeninconsistent or opposing statements.

Prior research in psychology has shown that decisionmakers are inuenced by available, comparative informa-tion. The presence of a comparison reduces the amount ofcognitive effort needed to process the information (Payneet al., 1993). The act of comparingconceptswithdistinct fea-tures maximizes attention to each of the components. Thecombined presentation aids the encoding process and min-imizes memory demands. In essence, the simultaneouspresence of two concepts reduces the likelihood that onlyoneof the conceptswill be recalledand therefore consideredin isolation. Kunda (1990) discusses how combinations ofcountervailing features are mentally processed. She indi-cates that people use causal reasoning when given conict-ing features of an object or principle. The resulting causalknowledge helps them resolve conicts among seeminglyincompatible concepts. The exercise of processing this cau-sal relationship increases its availability for use in subse-quent decisions. Applying this research to the presentstudy suggests that participants will better understand theintent of the advocacy and integrity standards when pre-sented in tandem. Rather than focus on only one direction,they are reminded to use integrity in reporting honestlywhile advocating for the client when it is appropriate to doso. As a result, when given an ambiguous context with un-clear technical guidance, combined exposure to both stan-dards should lead to a neutral response that is not overlypro-client or overly conservative for the facts at hand.

Although research has shown that decision makers willuse the information that is most readily available (Bhatta-charjee et al., 2007), when no information is explicitlyavailable (such as the absence of professional standardsfor our control group), the most readily available informa-tion is the subjects own knowledge or previous experi-ence. Prior psychology and accounting research has foundthat individuals are hard-wired to exhibit a conrmationbias (seeking conrmation of ones prior beliefs) and aselective information processing bias (seeing only whatone wants to see) when evaluating evidence (Cloyd & Spil-ker, 1999; Kunda, 1990; McMillan & White, 1993; Wheeler& Arunachalam, 2008). The potential bias, linked to experi-ence and/or prior beliefs, has led to increased attention toauditors levels of professional skepticism by the PublicCompany Accounting Oversight Board and audit research-ers (e.g., Nelson, 2009). For participants in our study, priorknowledge and previous experience most likely relates totheir pre-conceived notion of conservative nancial report-ing. In sum, participants in the control group are expectedto exhibit conservative tendencies on nancial reportingdecisions. In contrast, participants exposed to a jointpresentation of the standards for advocacy and integrity

will have readily available, comparative information toconsider when making their decision. Exposure to the

makers. Hence, our hypotheses are tested on a set of

32 D. Fatemi et al. / Research in Accounting Regulation 26 (2014) 2639respondents that has not yet been immersed in a workclimate that may foster certain norms.

Design, materials and procedure

The study uses a between-subjects design, varyingthe primed professional standard at four levels (none,

4 Approval for data collection was granted by the universitys institu-tional review board.combined standards is expected to produce less biasedjudgments. Thus, our third hypothesis is as follows:

H3. Participants exposed to a prime for both client advo-cacy and integrity will make more neutral, less biasednancial reporting decisions than those without anyexposure to the professional standards.

Method

Participants

A total of 113 undergraduates majoring in accountingfrom a university in the mid-west of the United States par-ticipated in the study.4 All participants in the study hadcompleted introductory nancial and managerial accountingclasses, 80% had completed intermediate accounting, only4% had taken an auditing class, and most were fourth-yearstudents. These participants represent inexperienceddecision makers, but the merits of using this population sub-set has been demonstrated in prior research (e.g., Ashton &Kramer, 1980; Maletta & Zhang, 2012). Although studentparticipants limit the studys generalizability, the effectson these potential new entrants into the profession are ofinterest to the profession as well as to educators. Clearly, fu-ture research will need to test our hypotheses on practicingaccountants, but students tend to be a more homogeneousgroup that may not differ on as many extraneous variablesas more seasoned participants would. In addition, the reac-tions of novice participants are likely to parallel those of jur-ors who could be asked to evaluate whether professionalshave complied with these standards.

Participants with strong internalized prior beliefs arelikely to be resistant to the inuence of a single exposureof a targeted professional standard. Furthermore, the useof experienced practitioners could lead to an extraneouseffect from the practitioners specic experience with theambiguous accounting issue used in the hypothetical case.Research by Jenkins and Lowe (1999) demonstrated thateven within one rm auditors signicantly disagree onwhether the auditors primary responsibility is to protectthe interests of investors and creditors (58.6% agreed while41.4% disagreed) and whether the auditors primaryresponsibility is to behave as a client advocate (32.8%agreed and 67.2% disagreed). Payne et al. (1993) assert thatheuristic biases, such as availability, priming, and hypoth-esis-conrming bias, are likely to affect novice decisionintegrity, advocacy, and combined advocacy and integritystandards). Participants were randomly assigned to a treat-ment group, and apart from those in the control condition,participants then read either a separate integrity prime, anadvocacy prime or a combined integrity and advocacyprime. The primed conditions were extracted directly fromthe AICPA Code of Conduct (2010):

Integrity requires a member to be, among other things,honest, and candid . . . Service and the public trustshould not be subordinated to personal gain and advan-tage. Integrity can accommodate the inadvertent errorand the honest difference of opinion; it cannot accom-modate deceit or subordination of principle. [Integrity]

A member or a members rm may be requested by aclient . . . to act as an advocate in support of the clientsposition on accounting or nancial reporting issues,either within the rm or outside the rm with standardsetters, regulators, or others. [Advocacy]

Pugrath, Martinov-Bennie, and Chen (2007) foundexperienced participants were inuenced by the inclusionof a lengthy, technical extraction from a combination ofAustralian principles-based and rules-based standards.Their student participants, however, were not inuencedby the presence of this guidance, and they concluded thatthe information seemed to overwhelm the inexperienceddecision makers. Consequently, the primes for advocacyand integrity in the present study did not include all ofthe content in the AICPA standards. The difculty withusing extracts from the standards is the decision regardingwhich words to retain and which to exclude. We focusedon the basic principles themselves, avoiding emotionally-laden phrases (e.g., stretch the bounds), and kept thelength of both standards fairly equal. Clearly, other ex-cerpts or revised phrasing could alter the strength of themanipulations, but our goal was to capture a portion ofthe original wording while maintaining the essence of eachstandard in a concise manner.

Participants were then asked to assume they were apublic accountant for a client who has asked for adviceon how to report an ambiguous issue on the nancial state-ments. The rst scenario is a skeleton version of the caseused by Pinsker et al. (2009) and is intentionally ambigu-ous with little guidance or evidence, pro or con. We stripthe case of the pro and con details to examine whetherthe ambiguity of the nancial reporting issue itself resultsin judgments that vary with exposure to the professionalstandards. This baseline scenario (shown in Appendix A)concerns whether a payment of $600,000 to the companyspresident should be deducted as compensation for servicesprovided or not deducted because it is a return of capital toa shareholder. The case makes clear that if the amount isdeducted (not deducted) then the clients nancial net in-come will be lower (higher). Participants were asked tomake a recommendation on a seven point scale anchoredby denitely deduct (+3) and denitely do not deduct(3), with a neutral response of zero. Thus, lower (higher)responses are more (less) pro-client.

After a decision regarding the salary-dividend classi-cation is made in a low-context case, a second salary-dividend

D. Fatemi et al. / Research in Accounting Regulation 26 (2014) 2639 33scenario is examined. This high-context case, used byPinsker et al. (2009) includes a substantial amount ofinformation that could be used as supporting and non-supporting evidence. They conclude that the ambiguouscase was indeterminable, theoretically resulting in anunbiased response. Prior research, however, has not con-trolled for prior belief on the targeted tax decision, in thiscase the classication of payments to an employee-share-holder, yet Payne et al. (1993) report that prior experience,knowledge, or belief can affect the targeted outcome. Thehigh-context case concerns a family-owned corporationwhere a widow of the founder becomes the president ofthe company and is paid a salary of $600,000. The casenarrative (see Appendix B) outlines the ofcers duties,and it provides nancial information and relevantguidance listing criteria for deducting (as compensation)and not deducting (if it is treated as a dividend or returnof capital) the amount. Participants are again asked tomake a recommendation on a seven point scale anchoredby denitely deduct (+3), which is less pro-client as netincome is lower, and denitely do not deduct (3), whichis more pro-client as net income is higher, with a neutraland balanced response of zero.

Research has shown that contextually-rich scenarioscan be interpreted differently depending on the incentivesof the decision maker (Cloyd & Spilker, 1999; Cuccia et al.,1995; Hackenbrack & Nelson, 1996; Johnson, 1993). Thedesign of the present study does not explicitly manipulateclient preference. Our context gives the participants an im-plicit client preference by noting that net income will behigher if the transaction is deemed to be a return of capitaland lower if it is deemed to be deductible compensation.Thus, participants may decide to report in a manner thatis believed to be what the client would prefer. Alterna-tively, participants may be inclined to report conserva-tively, consistent with prior studies in which clientpreference was not explicitly stated (Francis & Krishnan,1999; Jenkins & Haynes, 2003; Lord, 1992). Exposure tothe professional standards for advocacy and integrity areexplicit motivations that could affect the level of implicitmotivation either to produce client-favorable nancialstatements or the inclination to report conservatively.

After responding to the high-context case, participantswere asked what the most important factor was in theirdecision regarding whether or not to deduct payments tothe primary shareholder. Most made brief commentsregarding the nature of the compensation, but this sametype of comment was mentioned by those who deductedthe salary and by those who did not deduct it. Participantsalso answered the Pinsker et al. (2009) advocacy scale(adapted from Mason and Levys (2001) nine-item advo-cacy scale) and then proceeded to the background sectionwhere they answered a number of demographic questionspertaining to gender, prior coursework, and GPA (Radtke,2000). In addition, they responded to attitudinal questionsregarding the professional standards. When these variableswere tested as control variables that could affect the endresults, the results did not signicantly affect the out-comes, nor (as reported below) did they vary betweentreatment groups.Results

Preliminary analysis: attitudes and demographics

Demographics and attitude variables are presented inTable 1. The participants average age was 21.6 years, andmost were beginning their fourth year of undergraduatestudy. As previously noted, only 4% had taken an auditingclass. Parallel to the accounting departments malefemalestudent population, 40.2% were female.

At the conclusion of the experimental task, the back-ground questionnaire included two questions to check forcomprehension of the professional standards. The ques-tions asked respectively whether AICPA Rule 102-6(2) onadvocacy and Section 54-Article III(2) on integrity requirea more or less pro-client interpretation of the standards.The response scale for each question ranged from muchless favorable (+1) to much more favorable (+7) to the cli-ent. Participants rst responded with a mean of 5.04(SD = 1.25) on the question pertaining to Rule 102-6 onadvocacy, indicating a favorable, pro-client interpretation,with 72.3% agreeing. On a subsequent question relatingto Section 54-Article III on integrity, the mean was 3.36(SD = 1.33) with only 17.9% responding with a pro-clientinterpretation and 58% indicating an interpretation thatis less favorable to the client. A paired t-test indicates thatparticipants perceived these two standards as having a sig-nicantly different inuence on client preferred outcomes(t111 = 10.651, p < .001). In addition, participants wereasked a third question regarding which of the two stan-dards would be more inuential for making nancialaccounting decisions. The seven-point scale was anchoredwith Rule 102-6 (+1) and Section 54 (+7). The mean re-sponse was 4.83 (SD = 1.55), indicating that integrity wasperceived as more inuential. In fact, 64.3% asserted thatintegrity would be more inuential than advocacy, whileonly 24.1% stated that advocacy would be more inuential.Responses to these three background questions did notsignicantly differ between the treatment groups(F3,108 = 1.346, p = .263; F3,108 = 1.576, p = .200; F3,108 = .340,p = .797, for the three questions, respectively), indicatinga general consensus on the interpretation of the standards.This data conrms that the subject pool had a propensitytoward conservative reporting decisions, as integrity wasdeemed to be more inuential than advocacy for theiraccounting decisions, and integrity tended to be perceivedas unfavorable toward the client.

To measure general attitude toward client advocacy, thePinsker et al. (2009) advocacy scale was used. The Cron-bachs alpha for reliability was .83 in Mason and Levy(2001) and .85 in Pinsker et al. (2009) using a ve-itemscale to measure advocacy attitude in nancial decision-making. Our study uses the ve-item scale with anchorsof 1 for disagreement to the statement favoring advocacyand 7 for agreement, and the Cronbachs alpha for reliabil-ity is 0.85. The scaled average has a mean of 4.12(SD = 1.35). A midpoint of 4 (for lack of agreement or dis-agreement) represents the lack of a strong tendency ineither direction. Thus, in the present study most werenot inclined toward being an advocate. This average

Table 1Demographics and summary statistics for attitude and reporting decision.

Panel A: Demographics (n = 113)

r less

l, to yo

the n

orabls of 1

l guidto +7

cting am Joh

34 D. Fatemi et al. / Research in Accounting Regulation 26 (2014) 2639AgeGenderPanel B: AttitudeAdvocacy/integrity

Do you believe this AICPA guideline requires you to be more pro client ointerpretation of the nancial auditing standards?a

AICPA Code, Rule 102-6(2) [Advocacy]AICPA Code, Section 54-Article III(2) [Integrity]

Relative inuence

Based on the above two standards, which one would be more inuentianancial decisions?b

Mason and levy advocacy scale

Average response to ve statementsc

Panel C: Reporting decision

How strongly do you think that the $600,000 . . . should be deducted on(reducing the companys nancial net income)?d

Low contextual informationHigh contextual information

a Responses ranged from +1 (much less favorable) to +7 (much more favb Responses ranged from +1 (Rule 102-6) to +7 (Section 54). Response

inuential.c The ve statements (e.g., I feel I should apply ambiguous professiona

and Pinsker et al. (2009). Responses ranged from +1 (Strongly Disagree)Disagree (Agree).

d A seven-point scale was used, anchored at 3 for denitely not dedu+3) were coded as Not Deducting (Deducting). The high context case is froshown in the appendices.response is similar to Pinsker et al. (2009) who reported amean of 3.90 for professional auditors, while tax profes-sionals had a mean of 4.90. Consistent with ndings fromprior research on income-decreasing reporting decisions,the participants in the current study demonstrated a ten-dency to respond with conservative judgments.

To measure the level of conservatism affecting ourdependent variable, we rst note that the mean for allrespondents was .47 (SD = 1.65) on the low-context caseand .60 (SD = 1.95) on the high-context case, as shown inPanel B of Table 1. The responses were close to zero, imply-ing that the participants were relatively neutral. However,the responses were provided after the experimentalmanipulations. To demonstrate what the reporting judg-ment was in the absence of the manipulations, we testwhether a control group responded conservatively on thedependent variables. The means were 1.09 (SD = 1.54) forthe low-context case and 1.09 (SD = 1.88) for the high con-text case, which indicates agreement with a conservativeposition that the payment should be deducted as salaryand therefore reduce net income. A score of zero repre-sents a neutral position, and one-sample t-tests show thatthe responses signicantly differed from zero (t22 = 3.396,p = .002 for the low-context case; t22 = 2.772, p = .006 forthe high-context case).5 The results conrm that withoutthe presence of professional standards, participants respond

5 Throughout the paper, one-tailed (two-tailed) p-values are reported fordirectional (nondirectional) tests.e). Responses of 13 (57) were coded as Less (More) Favorable to Client.3 (57) indicate the view that Rule 102-6 (Section 54) would be more

elines to the clients benet) were adapted from Mason and Levy (2001)(Strongly Agree). Average responses of under (over)+4.0 were coded as

nd +3 for denitely deducting. Responses of 3 through 1 (+1 throughnson (1993), and the low context case is an abbreviated version. Both areMean (SD) = 21.57 (1.94)Percent male/female = 59.8/40.2

Mean (SD) Less Favorable toClient (%)

More Favorable toClient (%)

pro client in your

5.04 (1.25) 9.8 72.33.36 (1.33) 58.0 17.9

Mean (SD) Rule 102-6 (%) Section 54 (%)

u, in most 4.83 (1.55) 24.1 64.3

Mean (SD) % Disagree % Agree

4.12 (1.35) 37.2 55.8

Mean (SD) % Not deducting % Deducting

ancial statement

.47 (1.65) 34.5 57.5

.60 (1.95) 31.9 61.1conservatively on the outcome variables. Establishingwhether the participant pool has a penchant toward conser-vatism in nancial reporting decisions is important tounderstanding the potential impact of an integrity standardthat is interpreted as less favorable for the client and anadvocacy standard that is interpreted as favorable for theclient.

The effect of integrity as a separately-stated standard

Hypothesis 1The rst hypothesis predicts that participants who read

a prime for professional integrity will make more conser-vative decisions than control group participants. As shownin Table 2, the results of an independent-samples t-testcomparing the integrity and control groups showed no sig-nicant differences for the low-context case, with means of0.77 (SD = 1.51) and 1.09 (SD = 1.54) respectively(t43 = .692, p = .754). Similarly, comparison of the integrityand control groups for the high-context case yields no sig-nicant differences, with means of 0.86 (SD = 1.64) and1.09 (SD = 1.88) respectively (t43 = .424, p = .663).

As noted earlier, if the participants tend to make con-servative nancial reporting judgments, the availability ofan integrity prime may conrm but not change their pre-disposition towards conservative reporting. This attitudecould have affected the outcome of the rst hypothesis. Gi-ven that the pool of participants had virtually the samecourses in nancial accounting, it is likely that they were

equally exposed to the conservatism concept (Kieso et al.,2010). Consequently, the nonsignicant outcome on therst hypothesis implies that including a brief prime forprofessional integrity to decision makers with a propensitytoward conservative reporting is not likely to have an addi-tional effect, as it merely conrms the existing inclination.

To provide more insight, we test whether the integritygroups responses to the nancial reporting decision aresignicantly different from zero. One-sample t-tests showthat the responses in the integrity treatment were signi-cantly positive for both the low-context case (t21 = 2.401,

tion, as noted previously, the background questionnaire

were not signicantly different from the other participantson attitude toward client advocacy (F1,108 = .115, p = .735),indicating that most were not willing to assume a pro-clientposition that was questionable. Nonetheless, for the low-context case (without potentially damaging evidence) thepresence of Rule 102-6 for the advocacy group resulted in

from participants exposure to the combined standardsare shown in Table 3. In the low-context case, the respec-tive means are .09 (SD = 1.69) for the combined standardsand 1.09 (SD = 1.54) for the control group (t = 2.375,

D. Fatemi et al. / Research in Accounting Regulation 26 (2014) 2639 35measured advocacy attitude by using the Pinsker et al.(2009) advocacy scale. The mean response in the advocacytreatment group was 4.08 (SD = 1.24), and the mean re-sponse across all other groups was 4.13 (SD = 1.39). There-fore, student responses in the advocacy treatment group

6 We supplemented our analysis through the use of two additional typesof statistical testing. First, we performed randomization tests because theyavoid assumptions about the sampling populations parametric distribu-tion, providing greater exibility relative to conventional techniques (Good,2005). Statistical tests using this method are performed by comparing theexperimental results to comparison distributions formed from randomdivisions of the data into 100,000 samples, resulting in direct determinationof a p-value. We also examined our hypotheses using the MannWhitney Utest, which converts scores to ranks before comparing two groups. Theseanalyses provide further evidence of a difference between the advocacy andcontrol groups in the low-context case, as statistical signicance isobserved for both the randomization test (p = .047) and the MannWhitneyU test (U = 188.5, p = .044).p = .013) and the high-context case (t21 = 2.468, p = .011).Thus, similar to the control group, the responses of theintegrity group differed signicantly from a neutral re-sponse of zero, indicating a propensity to reportconservatively.

The effect of advocacy as a separately-stated standard

Hypothesis 2Hypothesis 2 predicts that a prime relating to client

advocacy (Rule 102-6) will lead to less conservative nan-cial reporting decisions than a control group. As shown inTable 2, Hypothesis 2 is supported in the low-context case,with respective means of 0.30 (SD = 1.69) for the advocacyprime and 1.09 (SD = 1.54) for the control group(t44 = 1.644, p = .054).6 In contrast, Hypothesis 2 is not sup-ported in the high-context case, with respective means of1.09 (SD = 1.86) for the advocacy group and 1.09(SD = 1.88) for the control group (t44 = .000, p = .500). Theimplication is that the advocacy standard causes the partic-ipants to be more pro-client (less likely to deduct nancialincome) when a low level of contextual data is present.The lack of support for Hypothesis 2 in the high-context caseis likely related to the participants inclination to make con-servative reporting judgments when additional informationmight be perceived as contrary evidence. The advocacytreatment group exhibits a level of a conservatism that isconsistent with the control group on the high-context case.As discussed earlier, participants in the advocacy group didnot differ from the other subjects on their perception thatRule 102-6 generally favors a pro-client judgment. In addi-66

p = .010), which supports Hypothesis 3 for the low-contextcase. Hypothesis 3 is also supported for the high-contextcase, with respective means of .02 (SD = 2.05) and 1.09(SD = 1.88) (t66 = 2.169, p = .017).

Supplemental tests demonstrate that the deductiondecisions in the presence of both standards did not differfrom a neutral position of zero (low contextual informa-tion: t44 = .353, p = .726; high contextual information:t44 = .073, p = .942). The neutral results suggest that thepresence of counterbalancing standards enables partici-pants to better comprehend each, consistent with the workof Payne et al. (1993), Anderson et al. (1990) as well asSumer and Knight (1996). When both standards arepresent, participants balance the support for advocacyand integrity in analyzing an ambiguous case, resultingin a more neutral, unbiased assessment of deductibility.

7 For participants in the other three groups, responses to the low contextcase did not signicantly differ from responses to the high context case. Inthe control group, the respective means were 1.09 (SD = 1.54) and 1.09(SD = 1.89) (t22 = .000, p = 1.000). In the integrity group, the respectivemeans were 0.77 (SD = 1.51) and 0.86 (SD = 1.64) (t21 = .216, p = .831). Inthe combined standards group, the respective means were 0.09 (SD = 1.69)and 0.02 (SD = 2.05) (t44 = .393, p = .696).a less conservative decision compared to those withoutexposure to the standard.

Regarding the high-context case, a propensity towardconservative nancial reporting is likely to lead to moreresistance to the advocacy prime, as the context has incre-mental details that could be interpreted as non-supportive.Participants responses to the two cases were, respectively,.30 (SD = 1.69) for a low level of contextual informationand 1.09 (SD = 1.86) for high contextual information, indi-cating a higher level of conservatism in the high-contextcase. A paired-samples t-test shows the assessments tobe signicantly different (t22 = 2.313, p = .015). Kunda(1990) recognizes that constraints can limit the extent towhich one is willing to seek conrmation of a preferredoutcome. The results suggest that the professional stan-dard for advocacy in the absence of contrary evidencesomewhat mitigated the conservative propensity.7

The effect of communicating combined professional standardson reporting decisions

Hypothesis 3Hypothesis 3 predicts that the presence of both stan-

dards will lead to less conservative (more neutral) deci-sions than those in the control group. Results arising

ion

36 D. Fatemi et al. / Research in Accounting Regulation 26 (2014) 2639Table 2Separately-stated standards for advocacy and integrity.

Panel A: Reporting decision mean (SD) [n]Contextual informat

Groupa LowDiscussion and concluding remarks

This study provides a foundation for future researchseeking to explain how the AICPA Code of Conduct(2010) can affect the behavior of its members. Even at

Advocacy 0.30(1.69)[23]

Integrity 0.77(1.51)[22]

Control 1.09(1.54)[23]

Panel B: Hypothesis testsH1: Participants exposed to a prime for professional integrity will make mor

without any exposure to the professional standards.

Test: Independent-samples t-test comparing reporting decisions in the contr

Context t (df)

Low .692 (43)High .424 (43)H2: Participants exposed to a prime for client advocacy will make less conser

the professional standards.Test: Independent samples t-test comparing reporting decisions in the contro

Context t (df)

Low 1.644 (44)High 0.000 (44)

Other terms are as dened in Table 1.a Before reading the two cases and making reporting decisions, participants

(Section 54-Article III). The control group did not receive either of these primes

Table 3Combined standards for advocacy and integrity.

Panel A: Reporting decision mean (SD) [n]Contextual inform

Groupa Low

Advocacy & integrity 0.09(1.69)[45]

Control 1.09(1.54)[23]

Panel B: Hypothesis testH3: Participants exposed to a prime for both client advocacy and integrity will

without any exposure to the professional standards.

Test: Independent samples t-test comparing reporting decisions in the contro

Context t (df)

Low 2.375 (66)High 2.169 (66)

Other terms are as dened in Table 1.a Before reading the two cases and making reporting decisions, participants i

Section 54-Article III. The control group did not receive these primes.Highthe corporate level, studies on the inuence of a code haveyielded conicting results, with some indications that acode can be effective (Hegarty & Sims, 1979) and other re-sults asserting the contrary (Cleek & Leonard, 1998; Forsteret al., 2009). The inuence of the AICPA Code of Conduct is

1.09(1.86)[23]

0.86(1.64)[22]

1.09(1.88)[23]

e more neutral, less biased nancial reporting decisions than those

ol and integrity groups.

p

.754

.663vative nancial reporting decisions than those without any exposure to

l and advocacy groups.

p

.054

.500

in the Advocacy (Integrity) group rst read an excerpt from Rule 102-6.

ation

High

0.02(2.05)[45]

1.09(1.88)[23]

make more neutral, less biased nancial reporting decisions than those

l group to the advocacy & integrity group.

p

.010

.017

n the Advocacy & Integrity group rst read excerpts from Rule 102-6 and

should hold, but this is a research question that will need

D. Fatemi et al. / Research in Accounting Regulation 26 (2014) 2639 37important for a professional organization that expects togain the publics trust (Moehrle et al., 2006), yet it hasrarely been tested.

The current study represents one step toward under-standing the impact of the AICPA Code of Conduct byexamining the effect of standards for advocacy and integ-rity on nancial reporting decisions. In addition to explor-ing the inuence of the standards in isolation, this studyinvestigates the effect created by a joint presentation ofcounterbalancing standards. One is associated with pro-client behavior, while the other distances itself from theclient by emphasizing objective, unbiased decision making.Hence, if taken separately, the two standards appear to cre-ate competitive guidance in decision making.

We nd that when participants are given an ambiguousfact pattern and must decide on whether to deduct pay-ments to an owner from nancial income, they exhibitconservative tendencies. This result, though obtained fromstudent participants, is similar to the income-decreasingpatterns observed from experienced professionals (Farmeret al., 1987; Francis & Krishnan, 1999; Jenkins & Haynes,2003; Lord, 1992; Trompeter, 1994). Exposure to AICPARule 102-6 for advocacy, however, decreases the conserva-tive posture in a case with scant contrary information,although when incremental but irresolute details are in-cluded (high-context case), the participants continue to re-port conservatively. Given the already conservative natureof the participants decision making, priming with AICPACode of Conduct, Section 54-Article III for integrity hasno effect on their nancial reporting decisions. This impliesthat exposure to a standard for unbiased decision makinghas little effect on decision makers who are already proneto conservative judgments.

When the two standards are jointly presented to partic-ipants, the ensuing decisions reect a neutral decision,absent a prevailing pro-client or conservative sentiment,and this result obtains with both low and high contextualinformation. The nding is consistent with the literatureon availability and the use of comparative analysis to facili-tate mental processing of each standards distinctive quali-ties (Anderson et al., 1990; Chi et al., 1989; Payne et al.,1993). Instead of being primed to consider only one stan-dard (favoring the client or disregarding the client), the jointpresentation improves the participants ability to grasp howeach standard provides a boundary for the other.

The results of the present study are informative forunderstanding the inuence of specic components ofthe professional code, and they further suggest that aware-ness of a single principles-based standard that jointlyrecognizes the roles of advocacy and integrity can enhancedecision making. This, of course, is just one piece of thepuzzle. Once a set of principles is shown to be effectiveat inuencing nancial decision making, it will becomecritical that management espouses these standards, asprior work has found that managements tone and organi-zational climate must reinforce the principles (DAquila,2003; Hageman & Fisher, 2012). The ndings will aid inpredicting the inuence of principles-based standards,and they will contribute to the ongoing debate on the roleof such standards relative to those that are rules-based, asexplored by Benston, Bromwich, and Wagenhofer (2006),to be tested empirically. Interestingly, Pugrath et al.(2007) tested the presence/absence of excerpts from anAustralian code of ethics, and its presence had a signicanteffect on the professional accountants but no effect on thestudent participants. They note, however, that the codewas very lengthy and technical which may have been dif-cult for the students to understand. Besides experience,other factors could moderate the results for student partic-ipants versus professionals. For example, Roberts (2010)documented that when client preference is made explicit,nancial accounting decisions of experienced accountantstend to become more pro-client. If participants have a pro-clivity toward aggressive, pro-client reporting, the resultscould reverse (i.e., the advocacy treatment could equateto the control group while the integrity standard could de-crease the pro-client tendency).

Acknowledgments

We thank Laureen Maines and Jake Rose for theirhelpful comments. In addition, the support of IndianaUniversitys Kelley School of Business, the Monroe Shineand Company Faculty Fellowship, Northern KentuckyUniversity and the University of New Hampshire are grate-fully acknowledged.

Appendix A. Low contextual information case

Assume you are a public accountant, and a client,Smith and Brown, Incorporated, has asked you how toreport an ambiguous accounting issue on the nancialstatements. Once you have read the scenario, please cir-cle the number that best describes your agreement witheach statement.

Your clients nancial net income has not yet been nal-ized. An expenditure of $600,000 has created a controversyas to whether this payment, which was to the President ofthe corporation, should be deducted as compensation. Theconcern is whether it is reasonable compensation for ser-vices to a major shareholder (requiring its deduction) ora non-deductible payment for return of capital to a share-holder. If it is deducted on your clients nancial statement,the nancial net income will be signicantly reduced.Jamal and Tan (2010), and Agoglia, Doupnik, and Tsakumis(2011). Finally, the use of student participants providesguidance on the degree to which students can be inu-enced by professional standards, a topic of importance inits own right (AACSB, 2004). The results suggest that amore holistic study of accounting principles may be moreeffective than a study of isolated standards.

This study has several limitations. First, the manipula-tions for the standards may have been weak, leading tothe lack of stronger effects from the segregated presenta-tions of integrity and advocacy standards. By using briefexcerpts from the two standards, they may not have beenstrong enough to affect the outcomes. Second, the use ofstudent participants inhibits the ability to generalize theresults to accounting professionals. If professional accoun-tants have a propensity to report conservatively, as priorresearch has shown, then the results of the current study

lic accountant, an sas an ambiguous ac eon nts. Once you hav -na number that bes e

m e d m yfo o r l,w c e s

y u a a s.3

Cl D & s 9 c alin o o al

of Business Ethics, 9(9), 699706.

38 D. Fatemi et al. / Research in Accounting Regulation 26 (2014) 2639been applied when considering the reasonableness of com-pensation, including (1) the employees role in the com-pany, (2) a comparison of the compensation paid tosimilarly situated employees in similar companies, (3)the character and condition of the company, (4) whethera conict of interest exists that might permit the companyr services ratherhich is not deduthantiblas a. Somdividendetimes ar retuve-fan of capitactor test haust be (1) reasonabl , an (2) in fact pay ents purel

laries or other compe sati n for services: T e paymentRelevant guidelines: Typically, there is a two-prongedtest for the deductibility of amounts purportedly paid as

sa n o h squestion.Background: Your new client, Johnson and Sons, Inc., is a

family-owned corporation engaged in the waste pickupand disposal business that performs trash-hauling servicespursuant to contracts with various municipalities. Thebusiness was incorporated by Mr. and Mrs. Johnson. Afterthe death of Mr. Johnson, the board of directors (composedof Mrs. Johnson and her four sons) elected Mrs. Johnsonpresident of the company. Each son is an ofcer with thetitle vice president and one son also holds the title ofsecretary and treasurer. During the past ve years, thestock of Johnson and Sons, Inc. has been owned by Mrs.Johnson (46%) and her sons (13.5% each).

Duties: Mrs. Johnson works 40 or more hours per week,but her four sons run the day to day operations. Her dutiesconsist of (1) keeping the nancial books, (2) reviewingbills and signing checks, (3) attending board meetingsand voting on major proposals put forward by her sons,(4) engaging in extensive public relations activities, and(5) acting as co-guarantor (together with her sons) of bankloans to the company for major capital expenditures.

Financial information: Some nancial information forJohnson and Sons, Inc. for the current taxable year is pro-vided below.

Gross sales $25,400,000Net income $155,000Ofcer compensation:Mrs. Johnson $600,000Sons (each) $375,000the nancial statemerio, please circle thee read the scet answers thked you how to report counting issu

Assume you are a pub d a client hapendix B. High contex caseIn this situation, you need to make a recommendation.How strongly do you think that the $600,000 paid to anowner of Smith and Brown, Incorporated should be de-ducted on the nancial statement (reducing the companysnancial net income)?

3 2 1 0 1 2 3Denitely Neutral DenitelyDo Not Deduct Deduct

Ap tual informationCleek, M. A., & Leonard, S. L. (1998). Can a corporate code of ethicsinuence behavior? Journal of Business Ethics, 17(6), 619630.

Cloyd, C. B., & Spilker, B. C. (1999). The inuence of client preferences on taxprofessionals search for judicial precedents, subsequent judgmentsand recommendations. The Accounting Review, 74(3), 299321.

Colson, R. H. (2005). Advocacy and independence. The CPA Journal, 75(1), 80.Cuccia, A. D., Hackenbrack, K., & Nelson, M. W. (1995). The ability of

professional standards to mitigate aggressive reporting. TheAccounting Review, 70(2), 227248.

DAquila, J. M. (2003). Does tone at the top that fosters ethical decisionsimpact nancial reporting decisions: An experimental analysis.International Business and Economics Research Journal, 2(8), 4154.aypool, G. A., Fetyko,dilemmas: A study. F.,pertaPearing ton, M. A. (1certied pu90). Reablic acctions to ethicuntants. JournHow students studCognitive Science, 1and, 145se ex182.mples in le rning to solve problem

i, M., Lewis, M., Rei mann, P., & Glaser, R. ( 989). S lf-explanationbusiness professionals: Portraying the accounting profession afterEnron. Accounting, Organizations and Society, 35(3), 360376.Ch n 1 e s:to disguise compensation as a dividend payment, and (5)whether the compensation was paid pursuant to a struc-tured, formal, and consistently applied program.

Please answer the following question by circling a num-ber on the scale:

In this situation you need to make a recommendation.How strongly do you think that the $600,000 paid to Mrs.Johnson should be deducted on the nancial statement(reducing the companys nancial net income)?

3 2 1 0 1 2 3Denitely Neutral DenitelyDo Not Deduct Deduct

References

AACSB Ethics Education Task Force (2004). Ethics education in businessschools. Tampa, FL: AACSB.

Abdolmohammadi, M. J., Read, W. J., & Scarbrough, D. (2003). Doesselectionsocialization help to explain accountants weak ethicalreasoning? Journal of Business Ethics, 42(1), 7181.

Agoglia, C. P., Doupnik, T. S., & Tsakumis, G. T. (2011). Principles-basedversus rules-based accounting standards: The inuence of standardprecision and audit committee strength on nancial reportingdecisions. The Accounting Review, 86(3), 747767.

American Institute of Certied Public Accountants (AICPA) (2010). AICPAcode of conduct. New York: AICPA.

Anderson, U., Marchant, G., Robinson, J., & Schadewald, M. (1990).Selection of instructional strategies in the presence of related priorknowledge. Issues in Accounting Education, 5(1), 4158.

Ashton, R. H., & Kramer, S. S. (1980). Students as surrogates in behavioralaccounting research: Some evidence. Journal of Accounting Research,18(1), 115.

Ayres, F., Jackson, B., & Hite, P. (1989). The economic benets ofregulation: Evidence from professional tax preparers. The AccountingReview, 64(2), 300312.

Bailey, C. D., Phillips, T. J., Jr., & Scoeld, S. B. (2005). Does political biasin the DIT or DIT-2 threaten validity in studies of CPAs? BehavioralResearch in Accounting, 17, 2342.

Bay, D. (2002). A critical evaluation of the use of the DIT in accountingethics research. Critical Perspectives on Accounting, 13(2), 159177.

Benston, G. J., Bromwich, M., & Wagenhofer, A. (2006). Principles- versusrules-based accounting standards: The FASBs standard settingstrategy. Abacus, 42(2), 165188.

Bhattacharjee, S., Maletta, M. J., & Moreno, K. K. (2007). The cascading ofcontrast effects on auditors judgments in multiple client auditenvironments. The Accounting Review, 82(5), 10971117.

Bobek, D., Hageman, A., & Hateld, R. (2010). The role of client advocacyin the development of tax professionals advice. The Journal of theAmerican Taxation Association, 32(1), 2551.

Carey, J. L. (2008). The independence concept revisited. Research inAccounting Regulation, 20, 295302.

Carnegie, G. D., & Napier, C. J. (2010). Traditional accountants and

Farmer, T. A., Rittenberg, L. E., & Trompeter, G. M. (1987). An investigationof the impact of economic and organizational factors on auditorindependence. Auditing: A Journal of Practice & Theory, 7(1), 114.

Financial Accounting Standards Board (FASB) (2010). Statement ofnancial accounting concepts no. 8. Norwalk, CT: FASB.

Forster, M., Loughran, T., & McDonald, B. (2009). Commonality in codes ofethics. Journal of Business Ethics, 90, 129139.

Francis, J., & Krishnan, J. (1999). Accounting accruals and auditorsreporting conservatism. Contemporary Accounting Research, 16,

Lord, A. (1992). Pressure: A methodological consideration for behavioralresearch in auditing. Auditing: A Journal of Practice & Theory, 11(2),89108.

Maletta, M. J., & Zhang, Y. M. (2012). Investor reactions to contrastsbetween the earnings preannouncements of peer rms. ContemporaryAccounting Research, 29(2), 361381.

Mason, J., & Levy, L. (2001). The use of latent constructs method inbehavioral accounting research: The measurement of client advocacy.

D. Fatemi et al. / Research in Accounting Regulation 26 (2014) 2639 39135165.Good, P. I. (2005). Resampling methods: A practical guide to data analysis

(3rd ed.). New York, NY: Birkhauser Boston.Gorman, M. E., & Gorman, M. E. (1984). A comparison of disconrmatory,

conrmatory, and control strategies on Wasons 2-4-6 task. QuarterlyJournal of Experimental Psychology, 36(4), 629648.

Hackenbrack, K., & Nelson, M. W. (1996). Auditors incentives and theirapplication of nancial accounting standards. The Accounting Review,71(1), 4359.

Hageman, A., & Fisher, D. (2012). The inuence of client attributes andorganizational climate on tax professionals. Best paper award at the17th Annual Ethics Research Symposium of the American AccountingAssociation, National Harbor, MD.

Hateld, R., Jackson, S. B., & Vandervelde, S. D. (2011). The effects of prioraudit involvement and client pressure on proposed auditadjustments. Behavioral Research in Accounting, 23(2), 117130.

Hegarty, W. H., & Sims, H. P. Jr., (1979). Organizational philosophies,policies, and objectives related to unethical decision behavior: Alaboratory experiment. Journal of Applied Psychology, 64, 331338.

Herron, T., & Gilbertson, D. (2004). Ethical principles vs. ethical rules: Themoderating effect of moral development on audit independencejudgments. Business Ethics Quarterly, 14(30), 499523.

Jackson, B. R., & Milliron, V. C. (1989). Tax preparers: Government agentsor client advocates? Journal of Accountancy, 189(2), 7782.

Jamal, K., & Tan, H. (2010). Joint effects of principles-based versus rules-based standards and auditor type in constraining nancial managersaggressive reporting. The Accounting Review, 85(4), 13251346.

Jenkins, J. G., & Haynes, C. M. (2003). The persuasiveness of clientpreferences: An investigation of the impact of preference timing andclient credibility. Auditing: A Journal of Practice & Theory, 22(1),143154.

Jenkins, J. G., & Lowe, D. J. (1999). Auditors as advocates for their clients:Perceptions of the auditorclient relationship. The Journal of AppliedBusiness Research, 15(Spring), 7378.

Johnson, L. M. (1993). An empirical investigation of the effects ofadvocacy on preparers evaluations of judicial evidence. The Journalof the American Taxation Association, 15(1), 122.

Jones, J., Massey, D. W., & Thorne, L. (2003). Auditors ethical reasoning:Insights from past research and implications for the future. Journal ofAccounting Literature, 22, 45103.

Kadous, K., Kennedy, S., & Peecher, M. (2003). The effect of qualityassessment and directional goal commitment on auditors acceptanceof client-preferred accounting methods. The Accounting Review, 78(3),759778.

Kadous, K., & Magro, A. (2001). The effects of exposure of practice risk ontax professionals judgments and recommendation. ContemporaryAccounting Research, 18(3), 451475.

Kieso, D. E., Weygandt, J. J., & Wareld, T. D. (2010). Intermediateaccounting. Hoboken, NJ: John Wiley and Sons Inc.

Kohlberg, L. (1973). Collected papers on moral development and moraleducation. Cambridge MA: Laboratory of Human Development,Harvard University.

Kunda, Z. (1990). The case for motivated reasoning. Psychological Bulletin,108(3), 480498.

Levy, L. (1996). Evidence evaluation and aggressive reporting in ambiguoustax situations. Ph.D. dissertation, University of Colorado.

Libby, R., & Luft, J. (1993). Determinants of judgment performance inaccounting settings: Ability, knowledge, motivation, andenvironment. Accounting, Organizations and Society, 18, 425450.

Loeb, S. E. (1971). A survey of ethical behavior in the accountingprofession. Journal of Accounting Research, 9(2), 297306.Advances in Taxation, 13, 123139.McMillan, J. J., & White, R. A. (1993). Auditors belief revisions and

evidence search: The effect of hypothesis frame, conrmation bias,and professional skepticism. The Accounting Review, 68(3), 443465.

Moehrle, S., Previts, G. J., & Reynolds-Moehrle, J. (2006). The CPA profession:Opportunities, responsibilities and services. New York: AICPA.

Nelson, M. W. (2003). Behavioral evidence on the effects of principles-and rules-based standards. Accounting Horizons, 17(1), 91104.

Nelson, M. W. (2009). A model and literature review of professionalskepticism in auditing. Auditing: A Journal of Practice & Theory, 28(2),134.

Payne, J., Bettman, J., & Johnson, E. (1993). The adaptive decision maker.Cambridge: University of Cambridge Press.

Pugrath, G., Martinov-Bennie, N., & Chen, L. (2007). The impact of codesof ethics and experience on auditor judgments. Managerial AuditingJournal, 22(6), 566589.

Pinsker, R., Pennington, R., & Schafer, J. K. (2009). The inuence of roles,advocacy, and adaptation to the accounting decision environment.Behavioral Research in Accounting, 21(2), 91111.

Radtke, R. R. (2000). The effects of gender and setting on accountantsethically sensitive decisions. Journal of Business Ethics, 24(4), 299312.

Rest, J. (1979). Development in judging moral issues. Minneapolis, MN:University of Minnesota Press.

Roberts, M. L. (2010). Independence, impartiality, and advocacy in clientconicts. Research in Accounting Regulation, 22(1), 2939.

Salterio, S., & Koonce, L. (1997). The persuasiveness of audit evidence: Thecase of accounting policy decisions. Accounting, Organizations andSociety, 22(6), 573587.

Schweikart, J. (1992). Cognitive-contingency theory and the study ofethics in accounting. Journal of Business Ethics, 11(56), 471478.

Scoeld, S. B., Phillips, T. J., Jr., & Bailey, C. D. (2004). An empiricalreanalysis of the selectionsocialization hypothesis: A research note.Accounting, Organizations and Society, 29, 543563.

Shaub, M. K. (1994). An analysis of the association of traditionaldemographic variables and the moral reasoning of auditing studentsand auditors. Journal of Accounting Education, 12, 126.

Simons, T. L. (2002). Behavioral integrity: The perceived alignmentbetween managers words and deeds as a research focus.Organization Science, 13(1), 1835.

Smith, L. M. (2003). A fresh look at accounting ethics (or Dr. Smith goes toWashington). Accounting Horizons, 17(1), 4749.

Sumer, H. C., & Knight, P. A. (1996). Assimilation and contrast effects inperformance ratings: Effects of rating the previous performance onrating subsequent performance. Journal of Applied Psychology, 81(4),436442.

Sunder, S. (2010). Adverse effects of uniform written reporting standardson accounting practice, education, and research. Journal of Accountingand Public Policy, 29(1), 99114.

Trompeter, G. (1994). The effect of partner compensation schemes andgenerally accepted accounting principles on audit partner judgment.Auditing: A Journal of Practice & Theory, 13(2), 5668.

Tsui, J. S. L., & Gul, F. A. (1996). Auditors behavior in an audit conictsituation: A research note on the role of locus of control and ethicalreasoning. Accounting, Organizations and Society, 21(1), 4151.

Unerman, J., & ODwyer, B. (2004). Enron, WorldCom, Andersen et al.: Achallenge to modernity. Critical Perspectives on Accounting, 15,971993.

Voynich, S. S. (2005). The professions core values: Connecting our past toour future. Research in Accounting Regulation, 18, 303320.

Wheeler, P. R., & Arunachalam, V. (2008). The effects of decision aiddesign on the information search strategies and conrmation bias oftax professionals. Behavioral Research in Accounting, 20(1), 131145.

The impact of professional standards on accounting judgments: The role of availability and comparative informationIntroductionTheory and hypothesis developmentPrimingProfessional integrityClient advocacyCombined counterbalancing standards

MethodParticipantsDesign, materials and procedure

ResultsPreliminary analysis: attitudes and demographicsThe effect of integrity as a separately-stated standardHypothesis 1

The effect of advocacy as a separately-stated standardHypothesis 2

The effect of communicating combined professional standards on reporting decisionsHypothesis 3

Discussion and concluding remarksAcknowledgmentsAppendix A Low contextual information caseAppendix B High contextual information caseReferences