uaud

ninParis,la Te

particular technology, the fraud triangle, which was initially developed in the aftermath of

CFOs cant be expected to peer into the souls of everyemployee or business partner, of course, but they do

surrounding thed clearly worthyhe desire fprevent it,

zations need to keep a closer eye on the soul omembers and, given that fraud ensues from lack of cit follows that the best way to keep it from spreading is tofoster control and cultivate specic skills in its detection. Inthis article, we examine how the notion of organizationalfraud has been conceptualized through a particular tech-nology: the fraud triangle.

http://dx.doi.org/10.1016/j.aos.2014.01.0060361-3682/ 2014 Elsevier Ltd. All rights reserved.

Corresponding author. Tel.: +33 149232029.E-mail addresses: jmorales@escpeurope.eu (J. Morales), yves.gendron

@fsa.ulaval.ca (Y. Gendron), henri.guenin-paracini@fsa.ulaval.ca(H. Gunin-Paracini).

Accounting, Organizations and Society xxx (2014) xxxxxx

Contents lists available at ScienceDirect

Accounting, Organiza

journal homepage: www.eby nancial gain, pressure to perform or a threat to theirpower, money or status; you need susceptible victims;and you need an absence of controls to prevent or haltthe scam. (The Observer, December 21, 2008)

assumptions often present in discoursesfraud triangle: an individual activity anof condemnation, fraud is motivated by tsonal gain, either in wealth or power. ToPlease cite this article in press as: Morales, J., et al. The construction of the risky individual and vigilant organization: A genealogfraud triangle. Accounting, Organizations and Society (2014), http://dx.doi.org/10.1016/j.aos.2014.01.006or per-organi-f theirontrol,Many people could, given the right set of circumstances,be capable of fraud. Experts in this area talk about thefraud triangle, whose three legs are: the opportunityto act dishonestly, an incentive or pressure to do so,and rationalization by the fraudster of their actions. Inother words, you need an offender motivated either

and the capacity to rationalize. When those elementsunite, fraud often erupts. (CFO Magazine, April 1, 2009)

These two quotations attest to the use in both the main-stream and specialized press of a particular expression,the fraud triangle, in reference to the activity of explain-ing, detecting or preventing fraud. They also illustrate keyIntroductionthe creation of the fraud examination discipline. We examine the genealogy of the fraudtriangle and follow various chains of translations underlying its construction. One of themain translation trajectories we uncover presents individuals as vectors of moral riskinessneeding to be vigilantly monitored and controlled by the organization. The organization isconceived of as a prime site in which fraud is to be addressed not annihilated but signif-icantly reduced as long as sufcient care is devoted to establishing layers of vigilance. Assuch, the fraud triangle redenes social, political and economic relations through a webof translations that both celebrate and normalize the use of organizational surveillance sys-tems to control risk ensuing from the individuals (alleged) frail morality. In the process,other visions of fraud, focused on the broader role of society, political agendas and power-ful economic institutions in engendering and preventing fraud, escape from view.

2014 Elsevier Ltd. All rights reserved.

need to be more cognizant than ever of the three sidesof the classic fraud triangle: pressure, opportunity,This article examines how a vision of organizational fraud has been constructed around aThe construction of the risky individorganization: A genealogy of the fra

Jrmy Morales a,, Yves Gendron b, Henri Gua ESCP Europe, CPO Department, 79 avenue de la Rpublique, 75543 Cedex 11b Facult des sciences de ladministration, Pavillon Palasis-Prince, 2325, rue de

a b s t r a c tl and vigilanttriangle

-Paracini b

Francerrasse, Universit Laval, Qubec City, Qubec G1V 0A6, Canada

tions and Society

lsev ier .com/ locate/aosy of the

2 J. Morales et al. / Accounting, Organizations and Society xxx (2014) xxxxxxBroadly speaking, our research connects to the power ofsystems of thought in shaping peoples interpretiveschemes along a particular angle, in inuencing social de-bates, and in delineating the boundaries of professionaljurisdictions. As illustrated in the quotations above, thefraud triangle promotes a vision where certain eventsand types of behavior are referred to and shaped as fraud-ulent and abnormal. The triangle also aids in the construc-tion and legitimation of a eld of intervention aroundorganizational deviance problems. We aim to understandhow the fraud triangle has been articulated through spe-cic conceptualizing angles.

For several decades, both academic and professionalliterature has taken great interest in the notion of fraudor, more generally, the dark side of organizations(Vaughan, 1999). Many scholars and practitioners havesought to better understand the causes of organizationaldeviance and assess different methods for detecting andpreventing it. Keen interest in the matter gave rise tothe emergence and propagation of a eld of knowledgespawned by the meeting of two disciplines that had beenrelatively distinct previously: accounting and criminol-ogy. We will show that the foundation of this hybrideld of knowledge, fraud examination, was built lar-gely on a particular image, that of the fraud triangle.Tracing the genealogy of the fraud triangle, therefore,leads us to examine the creation of the eld of fraudexamination.

Importantly, our primary interest is not in how thefraud triangle is technically used to detect fraud, but inhow its conceptualization and the underlying constitutionof a eld of knowledge have been structured around spe-cic angles. Following translations made in the academicand professional literature, this study highlights that a spe-cic vision of fraud characterizes how the triangle has beenconceptualized and conveyed through formal discourse.Thus, we contribute to an emerging eld of inquiry on hy-brid processes of knowing, spanning across the boundariesof organizations, professions and groups of experts (Miller,Karunmki, & OLeary, 2008), showing not only how ideasmove across disciplines, but also how a corpus that aspiresto be both new and distinct from existing bodies of knowl-edge may take root.

In our genealogy, we will trace the different transla-tions, developments and claims that have surrounded thespread of the triangle and possibly impacted the construc-tion of a new area of competence and its legitimizationaround the notion of fraud. Our empirical task, therefore,is to follow and reconstruct the chain of translations (La-tour, 1987) surrounding the fraud triangle. Every transla-tion entails a certain modication of the original concept(Gendron & Baker, 2005). Further, the term translationcomprises the unexpected twists and deformations thatmay occur in the spread of ideas. While ofcial historiesof elds of knowledge often ignore these elements ofunpredictability (Kuhn, 1970), they are nevertheless com-monplace. Translation also implies continuous work inextending the web of meanings surrounding a given con-cept. Thus, in our case, the fraud triangle does not have asingle, xed meaning; it is subject to continuous reinter-pretation. Even the core elements of the triangle changedPlease cite this article in press as: Morales, J., et al. The construction offraud triangle. Accounting, Organizations and Society (2014), http://dx.dosignicantly when they were imported into the nascentfraud examination discipline.

The translations we follow point to a distinctive way ofunderstanding the subject studied by organizational devi-ance professionals a subject that is no longer limited tothe behavior of individuals in and around an organization,but that now also extends to their character. Hence in addi-tion to the notion of translation developed by Latour(1987), we will make use of some of Foucaults writingson the normalization and constitution of the subject (Fou-cault, 1994, 2001, 2004), and in particular, his analysis ofthe intersection between psychiatry and criminology (Fou-cault, 1981). The spread of the fraud triangle has translatedinto the development of a discourse promoting the evalu-ation, monitoring and normalization of the character oforganizational members. This is because, together withorganizational control mechanisms, individuals moralityis presented as key to understand fraud risk factors. Be-yond usual auditing devices focused on conformity con-trols, the fraud triangle thus introduces morality andimmorality as a target for fraud ghting.

Before going further, we want to stress that our geneal-ogy is not aimed primarily at evaluating how the fraud tri-angle and its different articulations resonate withsurrounding and broader discourses (e.g., Miller & OLeary,1987), but rather to follow chains of translations andexamine the fraud conceptualizations they sustain. Also,while we distinguish translations along three main periods(roughly dened i.e., moving upstream; translations real-ized within the fraud examination community; movingdownstream), in our analysis we do not seek to highlightspecic variations within each of them. Our interest is tounveil what we view as the most important translationsduring each phase.

A key goal of our study is to trace the normativeassumptions that underlie the association between fraudand morality. We will show that they form the basis of adiscourse, not only about fraud detection and deterrence,but also about normality and deviance within organiza-tions. Dening an act as a transgression or fraud is ambig-uous (Berger, 2011). This is because fraud denitions relateto sociopolitical processes of labeling and are matters ofdisagreement and conict in society (Becker, 1963). Yet,the triangle is presented as a technical device that aimsto prevent acts deemed to be naturally unacceptable. Thisenterprise is clearly normative: through an ostensibly neu-tral technicality a specic point of view on organizationaldeviance is promoted. However, this is a normalizing view-point that conceives of fraud as a public issue caused byindividuals frail morality needing to be addressed by theorganization. In other words, the conceptualizations andrenements of the fraud triangle are based on implicit sep-arations between normality and abnormality, yet theboundaries between normal and abnormal are not dis-cussed as if the latter were natural and universallyaccepted.

We thus show how the fraud triangle traces links be-tween accounting, auditing and risk management on theone hand, and assumptions about normality and organiza-tional deviance on the other. In particular, our analysispoints to the fraud triangle extending the eld of organiza-the risky individual and vigilant organization: A genealogy of thei.org/10.1016/j.aos.2014.01.006

2004).

Handbook: Prevention and Detection, Wells founded theAssociation of Certied Fraud Examiners (ACFE). The bookis part of the ACFEs examination preparation toolkit. Ithelped guide our documentary research upstream fromthe triangle and shed light on some of its origins. We also

J. Morales et al. / Accounting, Organizations and Society xxx (2014) xxxxxx 3Also, our analysis of the triangles main translation tra-jectories points to a climate of suspicion promoting a eldof practical intervention within the organization, which ismade responsible for establishing vigilant systems of con-trol aimed at preventing and detecting lapses in organiza-tional members morality. Therefore, we studyconceptualizations of organizational deviance, as epito-mized in the discourse of the fraud triangle, to show howthey relate to normalizing stances encouraging organiza-tions to take responsibility for controlling individual char-acter. Other visions of fraud, focused on the broader role ofsociety, political agendas and powerful economic institu-tions in engendering and preventing fraud, are systemati-cally downplayed and overlooked. In the process, theorganization is constructed as a prime site in which fraudcan be reined in not annihilated but signicantly reducedas long as sufcient care is devoted to establishing a bat-tery of vigilant controls. Conditions are then arguably inplace for social order, and the dominant interests it sus-tains, to reproduce.

In the following section we present our research meth-ods. Next, we move upstream in the chains of translationsthat have led to the construction of the fraud triangle. Fol-lowing that, we present the main architects of the triangleand describe the translations they made as well as thestrategies they followed to present the triangle as a legiti-mate way of thinking and intervening in organizationaldeviance. Then, we move downstream in the chains oftranslations to follow some of the successive shifts that oc-curred as the network of support around the fraud trianglewas being extended. In particular, we show that thesedownstream chains promote a climate of suspicion thatenrolls the organization as a key site of fraud control. Inthe discussion section, we emphasize one of the maininterpretive statements ensuing from our analysis, namely,that organizational fraud is conceived of as a dual failure,one in individual morality and the other in the organiza-tions endeavors to control for individual probity. Lastly,the conclusion articulates some of the main implicationsresulting from our work.

Research methods

Our genealogy of the fraud triangle is carried outthrough documentary study. Our study is especially predi-cated on a pivotal practitioner book, which articulates con-nections between the nascent eld of fraud examinationand the discipline of criminology: Occupational Fraud andAbuse, by Joseph Wells (1997). In addition to republishingthis book three times under the title Corporate Fraudtional surveillance beyond individuals behavior to includethe assessment of risks stemming from individual subjec-tivity. Thus, the fraud triangle reects and participates inthe construction of one of the main social forces of ourmodern age: the redenition of social, political and eco-nomic relations around systems that position the subjectas a vector of riskiness and danger (Foucault, 1981) andset up and generalize surveillance and the assessment ofrisks stemming from individual subjectivity (Foucault,Please cite this article in press as: Morales, J., et al. The construction offraud triangle. Accounting, Organizations and Society (2014), http://dx.dolooked into the ways in which the fraud triangle and someof its translations are articulated in ACFE literature. Down-stream, we studied how the triangle and some of its deriv-atives are discussed in a range of journalistic, academic andprofessional documents. All told, our documentary analysiscovers a broad range of successive translations, from theemergence of white-collar criminology to recent shiftsand renements in the notion of the fraud triangle. Ouranalysis is articulated into three distinct parts whichcan be viewed as three moments of translations.

Our analytical journey begins with the examination of astrand of criminological research devoted to white-collarcrime, which Wells (1997) presents as one of the concep-tual bases of fraud examination knowledge. Although werecognize that the initiators of the white-collar crimemovement built their ideas on previous knowledge andnetworks (Latour, 1987), the chains of translations couldhave been retraced ad innitum and we had to choose areasonable starting point. In particular, we discuss severaldebates among criminologists that we link to competingvisions of fraud.

Yet these debates were largely overshadowed when thefraud triangle was conceptualized in the nascent eld offraud examination the criminological foundations offraud examination being presented as an indisputableand generalizable knowledge area. We are particularlyinterested in examining how the ACFE built a body ofknowledge around fraud examination, and what sort ofvision of fraud it conveys. In addition to Wells (1997)book, we analyze a number of documents connected tothe ACFE, including Albrecht and Albrechts (2004) bookentitled Fraud Examination & Prevention, whose rst authorwas ACFEs rst president.

In order to investigate, downstream and in greaterdepth, the spinning off of the fraud triangle beyond theACFE, we made an inventory of publications on the subjectusing bibliometric tools. We searched for the term fraudtriangle without any journal specication in the followingdatabases: Business Source Complete, ScienceDirect, Emer-ald and Jstor. Through this process we identied 99 arti-cles. Of these, we considered 54 relevant for our study.1

By retracing bibliographies, we obtained ten additional arti-cles that mention the three dimensions of the triangle with-out specically using the term fraud triangle or used theterm but were not referenced in our selected databases.We should note that these articles were published primarilyin academic journals, but also in a number of professionaland hybrid journals. We read and analyzed all 64 articles.For each document, we listed (when relevant) the theoreti-cal framework(s) used, the denition of fraud (and related

1 We eliminated articles that only mentioned the term fraud trianglewithout really presenting the model and without indicating its dimensions.We also eliminated articles which were only returned by the databasebecause they cited another article whose title includes the term fraudtriangle.the risky individual and vigilant organization: A genealogy of thei.org/10.1016/j.aos.2014.01.006

but rather in processes of socialization and cultural trans-mission that allowed certain individuals to learn criminaltechniques and, at the same time, led them to align theirworldview with that of their profession internalizingits customs, codes and practices.2 In 1939, he was ap-pointed president of the American Sociological Associationand during his rst presidential address he coined theterm white collar crime, dened as the violation ofdelegated trust within an organization (Sutherland, 1940).A few years later, he put forward a systematic analysis ofit, stating that:

These violations of law by persons in the upper socio-economic class are, for convenience, called white collar

4 J. Morales et al. / Accounting, Organizations and Society xxx (2014) xxxxxxMoving upstream in the chain of translations: Thearticulation of an individualizing vision of fraud

A signicant component in the development of a net-work of support surrounding the fraud triangle is the foun-dation of a new eld of knowledge and intervention; thatof fraud examination as defended and promoted by theACFE. Through our genealogical work, we rst analyzehow criminology was initially translated into fraud exam-ination. In this respect, Wells (1997) begins his book bypresenting fraud specialization as grounded in criminolog-ical knowledge. He brings together the writings of EdwinSutherland (pp. 810) and Donald Cressey (pp. 1022) totie the eld of fraud examination to previous academicstudies in criminology. Presenting the fraud triangle(Wells, 1997, p. 11) comprising pressure, opportunityand rationalization as having stemmed from Cresseysnal hypothesis gives it rhetorical strength. However, asshown below, Wells makes signicant translations whenarticulating this relationship and promotes a specic visionof fraud that differs from the one advocated by these twoscholars. Generally speaking, we aim to show that theimportation of criminology into the nascent fraud exami-nation discipline is characterized by a signicant transla-tion bias that dismissed or downplayed segments ofcriminological knowledge that viewed crime from socio-logical or systemic angles.

In the book which made him famous, Sutherland (1937)endeavors to show that the causes of criminality shouldnot be sought in the psychological traits of the individual,references), the causes of fraud, the solutions to fraud, themethodology, the way the triangle is used (presence of allthe dimensions, dimensions changed or added, translationof each dimension), the specic eld(s) of application men-tioned (e.g., employee fraud, misstatement, corruption, etc.),and the main argument. Four main translation trajectoriesemerged as signicant: focus on internal control, emphasison morality, assertion of suspicion, and promotion of organi-zational surveillance. We followed these four themes down-stream chains of translation. Further, as auditing standardswere often quoted in these articles to provide denitionsof fraud, we analyzed the translations comprised in SAS 99and ISA 240. Although these standards do not mobilize theexpression fraud triangle, both reproduce the threedimensions mentioned by Wells as constituting the triangle.

Finally, we examined several key documents in the riskmanagement, professionally-based literature. We limitedour analysis to documents issued by major accountingorganizations, such as the American Institute of CertiedPublic Accountants (AICPA), the Chartered Institute ofManagement Accountants (CIMA), the Center for AuditQuality (CAQ) and the Big Four. Accountants have beenparticularly involved in promoting their expertise in riskmanagement, appropriating and redening the concept ofrisk and its management along a variety of angles (Spira& Page, 2003). We identied the documents we analyzed,which are referenced below, either by questioning theInternet (using sentences such as managing fraud risk)or examining the organizations website.Please cite this article in press as: Morales, J., et al. The construction offraud triangle. Accounting, Organizations and Society (2014), http://dx.docrimes. This concept is not intended to be denitive,but merely to call attention to crimes which are notordinarily included within the scope of criminology.White collar crime may be dened approximately as acrime committed by a person of respectability and highsocial status in the course of his occupation. (Suther-land, 1983, p. 7)

Sutherland (1983) showed that criminology greatlyunderestimated (or even obscured) the violations of lawperpetrated by persons of the upper socioeconomic class.3

His hypothesis, now commonly referred to as differentialassociation theory, highlighted socialization and learningprocesses not pathological psychological traits as centralexplanations of criminal behavior. In adopting this stance,Sutherland discredited widely-held theories of his day thatattributed criminal behavior to poverty and its associatedpathologies (Shapiro, 1990, p. 346), what Braithwaite(1985, p. 2) calls the traditional theories of crime whichblamed poverty, broken homes, and disturbed personali-ties. The various cases (especially the biographical ones)presented by Sutherland are enlightening in this regardand show how the people concerned gradually lose theirattachment to a certain denition of honesty as they dis-cover that their professional success depends to a large ex-tent on performing unethical or even illegal acts. Given theinterpersonal competition that exists in the workplace,refusing to compromise their initial values may preventindividuals from being as effective in their occupation astheir less scrupulous colleagues. The individual then learns,along with techniques specic to the job, to adopt certainrationalizations which justify behavior that is neverthelessillicit (Sutherland, 1983, p. 245).

2 Based on the memoir of a former thief, the author develops the ideathat theft is not a set of isolated acts but a social institution, and thatprofessional thieves constitute a community. A particularly importantnding of this book is to show how, unlike amateurs, professional thievesmanage to avoid legal sanctions thanks to the intervention of a xer (seeSutherland, 1937, pp. 6587).

3 This book was very controversial in the United States mainly becausethe author broke a taboo by demonstrating the magnitude of the crimesperpetrated by the countrys elites (both in large corporations and ingovernment) and also because he showed that many acts which are neverbrought before a criminal court should nevertheless be considered crimes.Accordingly, the strength of white-collar criminals is that they are able toinuence legislation and law enforcement and also that they are most oftenjudged by administrative commissions and civil or equity jurisdictions,which may allow white-collar criminals to avoid the stigma associated withcriminal identity.the risky individual and vigilant organization: A genealogy of thei.org/10.1016/j.aos.2014.01.006

J. Morales et al. / Accounting, Organizations and Society xxx (2014) xxxxxx 5Sutherland (1983) counters previous studies in crimi-nology by expanding the denition of fraud to include actsthat had not been sanctioned by a criminal court (andwhich do not, therefore, appear in the statistics used bymost criminologists). His argument is that the denitionof white-collar crime should include violations of non-criminal law (Berger, 2011, pp. 67). As Coleman (1985,p. 4) points out, one of Sutherlands objectives was to dem-onstrate that the violations perpetrated by persons of theupper socioeconomic class are not treated in the sameway as others because the powerful have the ability toinuence legislation and, most importantly, itsenforcement.

Sutherland recognizes violations of law but proposes abroader denition of violation, especially relating to re-straint of trade, rebates, infringement, misrepresentationin advertising, and nancial manipulations. He also in-cludes unfair labor practices (Sutherland, 1983, p. 14).His denition of nancial manipulations (dealt with inchapter 10, pp. 153173), for example, is particularlybroad:

The term nancial manipulation is used here to referto practices of corporations or of their executives whichinvolve fraud or violation of trust. These practicesinclude embezzlement, extortionate salaries andbonuses, and other misapplications of corporate fundsin the interest of executives or of the holders of certainsecurities; they include public misrepresentation in theform of stock market manipulations, fraud in sale ofsecurities, enormous ination of capital, inadequateand misleading nancial reports, and other manipula-tions. (Sutherland, 1983, p. 153)

Expanding this subject matter beyond what is statuto-rily dened as crime by existing criminal law did not gowithout controversies. This is because dening an act asdeviant, criminal, transgressive or fraudulent is intrinsi-cally ambiguous (Becker, 1963, 2011). Indeed, rules differfrom group to group and a person may break the rules ofone group by the very act of abiding by the rules of anothergroup. Dening a rule (and hence what constitutes rule-breaking), therefore, involves a sociopolitical process andits content is a matter of disagreement and conict to thepoint where ofcial rules may differ from what the major-ity of people consider to be appropriate (Becker, 1963, p.15).

This position contradicts a commonly held assumptionabout rule-breaking as obvious acts of deviancy ensuingfrom the perpetrators deviant morality:

What laymen want to know about deviants is: why dothey do it? How can we account for their rule-breaking?What is there about them that leads them to do forbid-den things? Scientic research has tried to nd answersto these questions. In doing so it has accepted the com-mon-sense premise that there is something inherentlydeviant (qualitatively distinct) about acts that break(or seem to break) social rules. It has also acceptedthe common-sense assumption that the deviant actoccurs because some characteristic of the person whocommits it makes it necessary or inevitable that hePlease cite this article in press as: Morales, J., et al. The construction offraud triangle. Accounting, Organizations and Society (2014), http://dx.doshould. Scientists do not ordinarily question the labeldeviant when it is applied to particular acts or peoplebut rather take it as given. In so doing, they accept thevalues of the group making the judgment. (Becker,1963, pp. 3-4)

According to Becker, the notions of rule-breaking anddeviance (and therefore fraud) cannot be intrinsically de-ned by character traits, living conditions (economic sta-tus, lifestyle) or even by particular acts, because all ofthese denitions would deny the political and divisive nat-ure of rule-making: deviance is not a quality of the act theperson commits, but rather a consequence of the applica-tion by others of rules and sanctions to an offender(Becker, 1963, p. 9). What is considered as fraud depends,therefore, on the values and position of the person speak-ing and it is important to examine his/her assumptionsto better understand what motivates him/her to expressa certain moral view.

Overall, it can be argued that disagreement over thenature (individual or collective) and cause (moral or socio-logical) of white-collar crime characterizes the eld ofcriminology (Benson & Simpson, 2009; Berger, 2011; Cole-man, 1985; Colvin et al., 2002; Poveda, 1994). Accordingly,Berger (2011, p. 27) suggests distinguishing betweenmacro-sociological, micro-sociological, and individualisticexplanations of white-collar crime. Macro-sociological the-ories focus on the broader historical, economic, and polit-ical factors that impact organizations, while micro-sociological theories focus on the link between individualactors and their immediate organizational circumstances.For their part, individualistic theories focus on the causesof crime that are located within the individual rather thanthe social environment. According to this classication,Sutherland provides micro-sociological explanations ofwhite-collar crime (Berger, 2011, pp. 3536), whereasCressey focuses on individualistic explanations (p. 30).Macro-sociological theory does not appear to have hadany inuence on the crafting of the fraud triangle (forexamples of macro-sociological theories, see Coleman,1985; Poveda, 1994).

Cresseys individualistic theory occupies a prominentplace in ofcial representations of how fraud examinationknowledge was founded. Although one does not necessar-ily expect practitioners to adopt a position in an academicdebate, the creation of a formal body of practitioner knowl-edge may be founded on a particular representation ofreality and, therefore, favor one orientation to the detri-ment of others. Certain representational choices have thusbeen made (explicitly or implicitly) in establishing theknowledge base that underpins the fraud examiners juris-dictional claim. However, the ofcial story presents thesechoices as obvious and uncontroversial, as if they naturallyensue from some inherent properties of fraudulent actsand fraudsters, and a long tradition of scientic researchin criminology. In other words, the foundation of a jurisdic-tional claim based on a view of fraud inspired by Cresseysresearch involves a number of assumptions.

Cressey, a doctoral student of Sutherland, studied thepsychology of the violation of nancial trust using a seriesof interviews conducted with prison inmates, who hadthe risky individual and vigilant organization: A genealogy of thei.org/10.1016/j.aos.2014.01.006

been convicted of embezzlement (Cressey, 1953). Hismethods for selecting people to interview reveal his re-search focus: choosing from inmates at three US prisons,incarcerated after being convicted of certain crimes; he se-lected those he considered to have initially accepted, ingood faith, the position of trust that was used to perpetratethe crime (Cressey, 1953, p. 20).4 In focusing on prison in-mates, Cressey eschewed the study of an entire range ofwhite-collar crimes considered to be the most relevant bySutherland, whose argument is that the strength of white-collar perpetrators lies in their ability to avoid criminal con-victions better than anyone else. In particular, Cressey ig-nored acts of tax evasion and collective fraud. Moreover,by choosing only people who, upon examination of their leor after interviewing them, seem to have initially acceptedthe job in good faith, he favored the emergence of a dis-

6 J. Morales et al. / Accounting, Organizations and Society xxx (2014) xxxxxxcourse where the people he met did not consider themselvesto be professional thieves (see Sutherland, 1937). The advan-tage of this method is that it results in a relatively homoge-neous set of cases, ruling out the possibility of variation byselecting cases he felt were the result of the same type ofbehavior and developed in a similar way.

Cressey assumed that embezzlement is an act commit-ted for individual and nancial motives. This position ap-pears quite clearly in his interpretation of the acts ofmisappropriation committed by employees against theiremployers (Cressey, 1953, pp. 5766). Among thenumerouscases presented, only one is not motivated by nancial con-cerns that of an employeewho committed an act ofmisap-propriation, not because of personal difculties, but becausehe resented his status in the organization: he felt that hewas underpaid, overworked and treated unfairly. Yet Cres-sey persisted in considering this behavior to be motivatedby an individual problem, which ignores any broader expla-nation, even though the individual interviewed points to aset of social and collective motives linked to relations ofauthority that he considered unequal and unfair.

When considering the spectrum of perspectives in theeld of academic criminology, Cresseys thinking appearsto be rather restricted.5 He does not examine every typeof white-collar crime, but only embezzlement perpetratedby an individual acting alone, motivated by personal gain,and for which he has betrayed a position of trust that hehad initially accepted in good faith.6 Examining his work

4 He obtained the list of people convicted for: embezzlement, larceny bybailee, condence game, forgery, uttering ctitious checks, conspiracy,grand theft, theft of government property, falsication of a bill of ladingused in interstate shipment and theft of goods in interstate shipment(Cressey, 1953, p. 23).

5 For instance, beyond Sutherlands differential association theory,Merton crafted a strain theory (based on Durkheims concept of anomy),Cohen (1955) articulates a theory of subcultures, Sykes and Matza (1957)offer a neutralization theory, Cohen and Felson (1979) suggest a routineactivity theory (to include characteristics of the offended), Coleman (1985)offers an elite theory before advocating for an integrated theory(Coleman, 1987), Braithwaite (1989) suggests a theory of differentialshaming, Shapiro (1990) supports an offense-based approach (whenSutherland followed an offender-based approach), Vaughan (2005) speaksof a normalization of deviance, and Benson and Simpson (2009) suggestan opportunity theory. Berger (2011) also mentions rational choicetheory, control theory and criminogenic market structures theory.

6 All the persons he interviewed were men.Please cite this article in press as: Morales, J., et al. The construction offraud triangle. Accounting, Organizations and Society (2014), http://dx.doin greater detail, at the core of Cresseys thesis, we nd acombination of three elements that arguably help to explainthe process by which an individual comes to commit an actof embezzlement:

Trusted persons become trust violators when they con-ceive of themselves as having a nancial problemwhichis non-shareable, are aware that this problem can besecretly resolved by violation of the position of nancialtrust [perceived opportunity], and are able to apply totheir own conduct in that situation verbalizationswhich enable them to adjust their conceptions of them-selves as trusted persons with their conceptions ofthemselves as users of the entrusted funds or property[rationalization]. (Cressey, 1953, p. 30)

Of these three conditions, the most evident (to the pointwhere it may seem tautological) is that of perceived oppor-tunity: if a person commits an act, then he or she must nec-essarily have perceived the possibility to do so. Besides,Cressey (1953) gives opportunity less importance; therelevant chapter is particularly short (15 pages versus 44for the chapter describing what constitutes a non-share-able problem and 46 for the one dening rationalization).In this short chapter, he seeks above all to demonstratethat one should not consider opportunity as such butperceived opportunity, meaning that organizational con-trols are less important than individuals perceptions. Cres-sey provides additional details on opportunity in thepreface to the 1973 edition:

The general argument ordinarily is that embezzlementcan be eliminated by tighter accounting controls. Myresponse was, and is, that modern business necessarilyrequires conditions of trust and that, therefore,accounting controls rigid enough to eliminate embez-zlement (which, by denition, involves conversion ofmoney or goods with which the actor is entrusted) willalso eliminate business. [. . .] Accountants and otherbusinessmen continue to write as if the opportunityfor trust violation can and should be eliminated. [. . .]What is needed is a prevention program clearly basedon the fact that objective opportunities for embezzle-ment are necessarily present in modern business orga-nizations. (Cressey, 1973, pp. xii-xiii)

Therefore, Cressey is skeptical about internal controlsability to eliminate opportunity, and argues that oneshould assume that objective opportunities always exist.In the chapter on opportunity, he does address the issue ofhow to reduce opportunity but primarily seeks to under-stand why people eventually come to perceive their situa-tion as an opportunity to engage in embezzlement. Thus,Cressey focuses less on the link between individuals andorganizational circumstances than on phenomena locatedwithin individuals.

Of the three elements Cressey describes, the one thathas received the most criticism is the non-shareable nan-cial problem (Coleman, 1985). The authors idea is that,even when someone has had the possibility to embezzlefor several years, he will only do so because he is facing aproblem that he thinks he cannot share with anyoneand hence does not seek for others help. Again, however,the risky individual and vigilant organization: A genealogy of thei.org/10.1016/j.aos.2014.01.006

Cressey is not interested in identifying external factors thatmay explain why certain problems become non-share-able. Rather, he argues that misappropriation ensues fromthe individual viewing his problems as non-shareable.

sional work while grounding it in a formal body of knowl-

J. Morales et al. / Accounting, Organizations and Society xxx (2014) xxxxxx 7Finally, the third element is the notion of rationaliza-tion, borrowed from Sutherland, but redened accordingto a very different view.7 While Sutherland (1937) arguesthat rationalization helps to explain why professionalthieves proudly consider themselves as such, Cressey(1953) claims that embezzlers tend to rationalize theirbehavior in ways that imply that deep down, they are hon-est. Specically, Cressey contends that, before committing aviolation, the person convinces himself that the act will notcompromise his identity as an honest person and that it con-forms to a certain ethical view of himself. Cressey providesillustrations of the most common rationalizations and pointsout that this concept must not be confused with psycholog-ical justication carried out after the fact (Cressey, 1953, p.viii). According to him, these rationalizations are producedbefore the crime is committed and they are not used tomake the act intelligible for others, but rather for oneself(p. 95): they make the act possible because they make itmorally acceptable. However, even though this argumentis rather convincing, Cressey bases his ndings on inter-views carried out with prison inmates, i.e. after the factand in a situation of justication vis--vis a stranger. Hisinterpretation of the process of rationalization, therefore, issubject to criticism.

Overall, Cresseys (1953) ndings do not tell us muchabout sociological causes of fraudulent violations.Mainly, they focus attention on the idea that embezzle-ment is committed when an individual considers that hissituation makes the act of misappropriation feasible, nec-essary and acceptable. Cressey focuses on phenomena lo-cated within the individual perceptions, opinions andjustications that one gives to oneself. In so doing, his the-orizing is consistent with individualistic explanations ofcriminal behavior (Berger, 2011, p. 30), overlooking the mi-cro-sociological (ones immediate social and organizationalenvironment) as well as macro-sociological explanations(broader historical, economic, and political factors). Cres-seys work is thus based on a very particular conceptionof white-collar crime that marginalizes social environ-ments and circumstances.

Cressey thus offers a particular view of the offender:privileging individualist motives of transgression, hedownplays the role of sociopolitical explanations, creatingan individualizing vision of white-collar crime which, asshown below, is nevertheless capable of supplying argu-ments to dene and justify an area of intervention for cer-tain classes of professionals who claim jurisdiction overthe problem of organizational fraud (Abbott, 1988). Finally,it is worth noting that Cressey (1953) does not stress theword fraud in his book; the index indicates only one pagein which fraudulent checks is found, while embezzle-ment is found on 24 pages.

7 In fact, Cressey credits several scholars with authorship of this concept:not only Edwin Sutherland, but also George Herbert Mead, Herbert Blumer,Alfred Lindesmith and Charles Wright Mills (see Cressey, 1973, p. viii).Please cite this article in press as: Morales, J., et al. The construction offraud triangle. Accounting, Organizations and Society (2014), http://dx.doedge ostensibly derived from scientic criminology.Signicant translations were made in the process, espe-cially when articulating a specic denition of fraud thatresulted in a number of departures or biases from the ori-ginal criminological work. In what follows, we highlightthe main translations that Wells and other ACFE propo-nents realized in attempting to build a network to natural-ize the fraud triangle. Then, we examine some of the mainrhetorical strategies they used to secure support.

Translational biases in the initial formulations of the triangle

Wells played a leading role in formalizing the fraud tri-angle yet his position within denitional debates is notneutral. As indicated below, the way in which he translatedcriminological knowledge involves the creation of a systemof representations promoting a sense of causality betweenindividual character and fraud, which needs to be reined inby the organization. In his 1997 book, Wells promotes aconceptualization of fraud grounded in what is presentedas one of the cornerstones of criminological thought,namely, the embezzlement hypothesis as originally con-ceived by Cressey:

One of Sutherlands brightest students at Indiana Uni-versity during the 1940s was Donald R. Cressey(1919-1987). While much of Sutherlands research con-centrated on upper world criminality, Cressey took hisown studies in a different direction. Working on hisPh.D. in criminology, he decided his dissertation wouldconcentrate on embezzlers. Accordingly, Cresseyarranged the necessary permission at prisons in theMidwest and eventually interviewed about 200 incar-cerated inmates. [. . .] Cressey was intrigued by embez-zlers, whom he called trust violators. He wasespecially interested in the circumstances that led themto be overcome by temptation. [. . .] Upon completion ofhis interviews, he developed what still remains as theclassic model for the occupational offender. Hisresearch was published in Other Peoples Money: A Studyin the Social Psychology of Embezzlement. Cresseys nalhypothesis was: [see previous section, excerpt fromCressey, 1953, p. 30]. (Wells, 1997, p. 10)

The transition from Sutherland to Cressey is quiteabrupt in Wells depiction of the historical origins of theelds formal body of knowledge. He also downplays therole of the social context in Sutherlands work. Yet in pre-senting Cressey as one of Sutherlands doctoral students,he invokes a sense of progress and continuity in the devel-opment of criminological knowledge. Importantly, the crit-icisms that Cresseys work engendered in the eld ofRealizing translation through the constitution of aknowledge claim

Although often attributed to Cressey, the term fraudtriangle does not appear in his writings. In fact, it wasnot coined by a criminologist at all, but by Joseph Wells,a CPA. However, drawing on Cressey, Wells and severalother proponents laid claim to a particular area of profes-the risky individual and vigilant organization: A genealogy of thei.org/10.1016/j.aos.2014.01.006

criminology are not mentioned. Hence the above initialtranslation, from criminology to fraud examination, wasfundamentally partial in both senses of the word (i.e.,incomplete and biased).

The above excerpt, however, does not mention the termfraud triangle. We contacted the ACFE by email concern-

8 J. Morales et al. / Accounting, Organizations and Society xxx (2014) xxxxxxing the origins of this term and received the following re-sponse from an ACFE representative:

Dr. Cressey developed the three items, but he did notcall it the Fraud Triangle. Actually, Dr. Wells is the rstperson we know of to take the three items and put[them] in a triangle format. He was working on a videofeaturing Dr. Cressey in 1985, and he used a trianglegraphic in the video to illustrate the 3 factors that arepresent in most white-collar offenses. He began usingthe triangle graphic in training programs after thattime. People saw the graphic and began referring to itas the Fraud Triangle over the years. So although wehave never undertaken an extensive review of its use,as far as we know, thats how it came about.8

This quotation is noteworthy in several respects. Frompractical experience, it is condently maintained that thethree factors are present in most white-collar offenses.This generalizing attitude may be linked to a systematicshift from limited empirical cases to a broader label, asif the original analysis had taken all types of white-collarcrimes into consideration and could, therefore, be general-ized to every form of fraud. We found a similar tendencyto generalize in other ACFE documents, suggesting thatACFE representatives are highly condent in the knowl-edge base that allegedly underlies the fraud triangleconcept.9 The latter also tends to be naturalized in thesedocuments, as indicated by the capital letters being usedin the last sentence of the excerpt (Fraud Triangle),implying that the triangle is now understood as arecognized doctrine, a proven theory or, quite simply, aninstitution. Thus, the knowledge base claimed by the ACFEis presented as universal:

Perceived pressure, perceived opportunity, and ratio-nalization are common to every fraud. Whether thefraud is one that benets the perpetrators directly, suchas employee fraud, or one that benets the perpetratorsorganization, such as management fraud, the three ele-ments are always present. (Albrecht & Albrecht, 2004, p.20)

The triangle provides the core elements of the fraudexamination disciplines body of knowledge. Importantly,a vast range of enigmas and peripheral questions emergewhen application is taken into account. How is fraud com-mitted? What mechanisms have been or should be used byorganizations in trying to rein in fraud? How effective arethese mechanisms? One of the most signicant questions

8 Email received on May 27, 2011 from a representative of the ACFE. In2010, Joseph Wells received a Doctorate Honoris Causa from the YorkCollege of The City University of New York, hence the use of the title Dr.in the ACFEs email.

9 See Gabbioneta, Greenwood, Mazzola, and Minoja (2013) for a study offraud practitioners overcondence.Please cite this article in press as: Morales, J., et al. The construction offraud triangle. Accounting, Organizations and Society (2014), http://dx.dois what kind of individual is more likely to commit fraud?In the previous excerpt, for example, the main distinctionis between employees assumed to commit frauds againstthe organization and management assumed to commitfrauds for the organization. Proling knowledge is enlistedto address this question. Accordingly, the ACFE developedcategorizations of fraudsters and their acts along variousdimensions. For example, Wells (1997) mobilizes the re-sults of a large-scale survey of Certied Fraud Examiners(CFEs), who were asked to provide information on actualfraud cases. He presents series of graphs (chapter 1), whichcollectively associate the imagery of the fraudster to theprole of a married male, for instance. Predictive abilitiesare emphasized in the text, through statements such as:One of the most meaningful trends of the survey is the di-rect and linear correlation between age and median loss(Wells, 1997, p. 38).

One key emphasis in the Associations body of evolvingknowledge is to associate fraud with moral issues. By hisdenition, Wells (1997, pp. 3-6) assumes that fraud consti-tutes a dishonest act perpetrated by an individual for per-sonal enrichment. In other words, fraud is being rooted inindividuals frail morality (and not as an effect of broadersociological and cultural inuences) and as a problemnecessitating the surveillance of individual ethics by theorganization:

Fraud is a generic term, and embraces all the multifari-ous means which human ingenuity can devise, whichare resorted to by one individual, to get an advantageover another by false representations. No denite andinvariable rule can be laid down as a general proposi-tion in dening fraud, as it includes surprise, trickery,cunning and unfair ways by which another is cheated.The only boundaries dening it are those which limithuman knavery. (Albrecht & Albrecht, 2004, p. 5)

This denition is centered on individual acts of moraldeviance. One of its most striking features is the use ofderogatory and persecutory language in qualifying theseacts. In a similar way, Wells moralizing tone infuses theseries of short cases he provides to demonstrate linkagesbetween certain behavioral patterns and fraud. Most often,these cases question the morality of specic individuals,without considering the surrounding network of institu-tions and social interrelations. For instance, the practiceof background checks, which Wells actively promotes, sin-gles out a category of individuals as problematic hires (seealso ACFE, 2010, p. 81):

The news that the hospital had unknowingly hired aconvicted felon distressed [the director of internalauditing in charge of investigating the case of anemployee who regularly overstated the number ofhours on his time sheets]. He discovered that the hospi-tals ability to conduct thorough background checks onprospective employees was restricted by money andaccessibility to records. [. . .] Harkanell [i.e., the personwho defrauded the hospital] remained at large for sev-eral months. But luck was on the hospitals side. Or, per-haps more accurately, stupidity was on Harkanells. Justas he had done with his time-sheet fraud, he left a cluethe risky individual and vigilant organization: A genealogy of thei.org/10.1016/j.aos.2014.01.006

behind, this time concerning his whereabouts. (Wells,1997, p. 263)

The overarching lesson from this example seems to be

J. Morales et al. / Accounting, Organizations and Society xxx (2014) xxxxxx 9that organizations should be reluctant to hire people witha criminal record, regardless of the circumstances in whichthey were convicted. This is reinforced in an article pub-lished in the ACFEs newsletter, in which an actual case isdeveloped through text and video to promote the view thatmoral deviants are unlikely to rehabilitate. As stated by theACFE president: Ive had fraud perpetrators tell me that itbecame an addiction for them. [. . .] Trust equals opportu-nity and it is indeed common for a fraudster to perpetratea second fraud when they get in a position of trust (Patt-erson, 2008).10 From this perspective, moral deviancy con-stitutes an incurable illness and the best that can be donein the face of it is to strengthen an organizations internalcontrols to catch fraudsters and place them in the hands ofjustice. Our analysis, therefore, points to fraud being concep-tualized, in the ACFEs community, from an angle that viewsthe problem of fraud as one of morality lapses upon whichthe organization inescapably needs to intervene.

Thus, from the triangles perspective, individual moral-ity is perceived as a chief source of fraud and the organiza-tion is presented as having prime responsibilities inestablishing a proper structure to control moral deviancies.In other words, the triangle provides fraud specialists withan investigative template that individualizes fraud andholds organizations responsible for controlling it. As a re-sult, fraud is circumscribed to the realm of the specic:individuals subjected to pressure and somehow able torationalize the act should not be left in a position to commitfraud. From this perspective, it is incumbent on the organi-zation to ensure that the three legs of the triangle are prop-erly overseen through a rigorous structure of internalcontrol. Fraud is thus constituted as a problem at the con-uence of the individual and the organization; it is certainlynot represented as a social, political, or historical problem.

The translations developed in the ACFE community rep-resent a signicant departure from Sutherland and evenfrom Cressey. For instance, Wells (1997, p. 17) translatesCresseys notions of position of trust and violation oftrust to emphasize the need for systematic controls:since [an employee] is in a position of trust (read: noone is checking) it can be violated. Although Wells(1997, p. 11) mentions the notion of non-shareable prob-lem briey, this central aspect of Cresseys hypothesis is re-placed with the notion of pressure. This notion misses acentral element of Cresseys work whereby a sequence ofevents is present when a violation of trust is committedand not present when no transgression is observed (Cres-sey, 1953, p. 12). Indeed, Cressey resolutely maintains thatpeople in a position of trust in an organization, who aresubjected to nancial pressure for several years, do not vio-late this trust until their nancial problems are felt as non-shareable.

10 Patterson (2008) provides video recordings of an individual who, afterserving a sentence for nancial fraud, explains that, like an alcoholic, he isnow incurable: this individual asks that he no longer be trusted because heknows that he will once again betray this trust.Please cite this article in press as: Morales, J., et al. The construction offraud triangle. Accounting, Organizations and Society (2014), http://dx.doIn addition, Wells (1997, p. 24) relates perceivedopportunity to lack of deterrence by management, as ob-served through decient internal controls. The trianglealso translates rationalization in a very specic man-ner. Wells (1997, p. 17) claims that it is part of humannature: once the line is crossed, the illegal acts becomemore or less continuous. Overall, in the eyes of the tri-angles proponents, the likelihood of giving in to pressurecharacterizes certain groups of individuals. Cresseyshypothesis that fraud is rst and foremost explained byindividuals perceptions is thus replaced by another indi-vidualizing stance now focused on personality andmorality.

These departures from Cresseys work are justied bythe age of the original research: The study [Cressey,1953] is nearly half a century old. There has been consider-able social change in the interim. And now, many antifraudprofessionals believe there is a new breed of occupationaloffender one who simply lacks a conscience sufcientto overcome temptation (Wells, 1997, p. 20). Wells evenreintroduces genetic causes and argues that, in contrastto white-collar crimes, what he calls street crimes aregenetically based (Wells, 1997, p. 9). Wells (1997, p. 24)further modies Cresseys nal hypothesis to make thefraud triangle consistent with the fraud scale (see alsoAlbrecht, Romney, & Howe, 1984, p. 21) where the threedimensions can offset each other, non-shareable problemsare replaced by situational pressures and rationalizationsby personal integrity. The fraud triangle has thus trans-lated Cresseys work into a device to be used in organiza-tions to measure and intervene on fraud risks associatedwith deviant morality.

In summary, our analysis indicates that the creation ofa fraud examination knowledge base entailed substantivetranslations from the eld of criminology, in that Cres-seys ndings are introduced as an indisputable and gen-eralizable foundation for the practice of fraudspecialization to thrive. The initial translation was meta-phorically built around the fraud triangle, a concept thatconstitutes the hard core of the discipline while providingenough room for imagination in developing ancillarytranslations that extend the scope of the triangle in termsof technical and application details. From these transla-tions emerges a broader discourse that denes fraud asa problem that fundamentally relates to malevolent actsperpetrated by individuals whose morality is at best frailand at worst nonexistent. Thus, when fraud examinerstalk about systems, the scope is tightly limited to theboundaries of the organizations internal control system.In short, the translations surrounding the fraud trianglepromote a vision that condently individualizes fraud tothe point where a stigmatizing judgment sometimes isreadily passed on moral deviants who succumb to nan-cial pressures and make use of an organizational opportu-nity in order to remorselessly defraud their employingorganization. The latter is rendered accountable in theprocess, being enlisted in establishing a proper controlstructure around fraudulent risk. As a result, both individ-ual and organization are likely to share blame whenfraudulent behavior takes place in a given organizationalsetting.the risky individual and vigilant organization: A genealogy of thei.org/10.1016/j.aos.2014.01.006

Building a network of support to naturalize the triangle

Our analysis indicates that the ACFE houses the mainarchitects of the initial fraud triangle concept. After work-ing for the FBI for about ten years, Wells started a profes-sional service rm in 1982, specializing in frauddetection and prevention. Wells founded the ACFE a fewyears later, in 1988. The ACFEs mission is to reduce theincidence of fraud and white-collar crime and to assistthe Membership in fraud detection and deterrence.11 Asof May 2011, the Association comprised over 55,000 mem-

examination practices, while its website indicates a researchunit comprising nine staff members (May 2011), as well as22 faculty members described as some of the mosthighly-rated speakers in the anti-fraud profession.17 TheACFEs efforts in developing a formal and practically-ori-ented knowledge base are numerous for instance throughthe creation of the Institute for Fraud Prevention, describedas a coalition dedicated to multi-disciplinary research, edu-

10 J. Morales et al. / Accounting, Organizations and Society xxx (2014) xxxxxxbers (65,000 in April 2013) with more than 30,000 beingCFEs and the remainder being students and associate mem-bers.12 Certication implies passing a compulsory examina-tion and completing two years in a fraud detection orprevention-related position. CFEs key abilities reportedlyinclude identifying an organizations vulnerability to fraud,examining records to detect and trace fraudulent transac-tions, conducting interviews to obtain information, and giv-ing advice on improving fraud prevention and deterrence.13

The ACFEs formal body of knowledge comprises a mix oftheoretical, methodological and deontological elementsrelating to accounting/auditing, information technology,criminology, ethics, fraud investigation/schemes, interview-ing skills, and law.14 Apparently, CFEs are held in high re-gard; statements by representatives of Robert HalfInternational, the U.S. Department of Defense and the U.S.Government Accountability Ofce attest to this.15

The ACFE uses an extensive network of resources topromote the claim that it possesses a formal body ofknowledge derived from criminology, in the hope of secur-ing its jurisdiction and, at the same time, legitimizing itsviews on the essence of fraud and how it should be con-trolled. Every year since 1989, the ACFE has bestowedthe Cressey Award for a lifetime of achievement in thedetection and deterrence of fraud on a member who bestcontributed to the ght against fraud. The Award isnamed in honor of one of the countrys foremost expertson fraud and a founding father of the ACFE, Dr. Donald R.Cressey (Association of Certied Fraud Examiners (ACFE),2013).16 Further, the ACFE organizes annual conferenceswith each of the latest ones reportedly attracting over2000 attendees. Roughly every two years, it publishes anextensive survey report entitled, Report to the Nation(s),which draws on information from a large sample of CFEsto provide an approximation of the cost of fraud in the econ-omy and detailed knowledge on fraud types and perpetra-tors. The ACFE also publishes the Fraud Magazine, abimonthly publication on white-collar crime and fraud

11 Read on April 12, 2013 on www.acfe.com/who-we-are.aspx.12 Taken from an e-mail communication with one ACFEs memberservices representative (May 23, 2011).13 Read on May 25, 2011 on www.acfe.com/documents/press-kit/acfe-certied-fraud-examiners.pdf.14 Read on April 12, 2013 on www.acfe.com/self-study-cpe.aspx.15 Read on May 25, 2011 on www.acfe.com/documents/about-acfe.pdf.16 It is worth noting that Cressey died in 1987, about one year before thecreation of the ACFE (Wells, 1997, p. 22). According to Wells, Cresseyexpressed, during a conversation with him, the idea that it was time for anew type of corporate cop one trained in detecting and detecting thecrime of the future: fraud (Wells, 1997, p. 22).Please cite this article in press as: Morales, J., et al. The construction offraud triangle. Accounting, Organizations and Society (2014), http://dx.docation and prevention of fraud and corruption.18 The ACFEhas been particularly involved in disseminating its represen-tations in colleges and universities through an anti-fraudeducation partnership in which institutions are encouragedto develop a course on fraud examination, using extensiveeducational material that the ACFE provides free-of-charge.In 2010, nearly 400 post-secondary institutions worldwidereportedly took advantage of this opportunity (Carozza,2010).

Typically, books, articles and documents produced byWells and the ACFE are peppered with survey results. Thisgives the impression that the fraud examination eld iscommitted to building a knowledge base predicated onmethodologically rigorous and generalizable knowledge(as opposed to the production of knowledge focused oncontextualizing). This kind of rhetoric is increasinglyappealing given the emphasis on generality, mobility, com-parability and standardization in todays society (Porter,1995; Reed, 1996). The ACFE has continued and expandedits categorizing endeavors, producing several Reports to theNation(s).19 In the 2010 Report to the Nations, informationobtained from a sample of CFEs, who documented a totalof 1843 cases of occupational fraud, is presented in numer-ous graphs and gures (ACFE, 2010). One of the most pecu-liar aspects of the Report is how far claims go to provide asense of effectiveness regarding some of the anti-fraud tech-niques advocated by the ACFE. For instance, the followingexcerpt describes a type of effective intervention:

The ability to report fraud anonymously is key becauseemployees often fear making reports due to the threatof retaliation from superiors or negative reactions fromtheir peers. [. . .] One would expect that the presence ofa fraud hotline would enhance fraud detection effortsand foster more tips. This turns out to be true. As seenon page [x], the presence of fraud hotlines correlatedwith an increase in the number of cases detected by atip. In organizations that had hotlines, 47% of fraudswere detected by tips, while in organizations withouthotlines, only 34% of cases were detected by tips. Thisis important because tips have repeatedly been shownto be the most effective way to catch fraud. [. . .] Perhapsmore important, [. . .] organizations that had fraud hot-lines suffered much smaller fraud losses than organiza-tions without hotlines. (ACFE, 2010, p. 17)

17 R e a d o n M a y 2 5 , 2 0 1 1 o n www . a c f e . c om / a b o u t /faculty.asp?copy=faculty.18 Excerpt from www.theifp.org/ May 25, 2011.19 The 2010 Report is the rst in the series whose data is drawn from fraudcases not constrained to the USA hence the use of the term nations. Thissuggests that the Association is increasingly involved on the transnationalscenery, along entities such as the International Federation of Accountantsand the International Accounting Standards Board.the risky individual and vigilant organization: A genealogy of thei.org/10.1016/j.aos.2014.01.006

J. Morales et al. / Accounting, Organizations and Society xxx (2014) xxxxxx 11Knowledge on the effectiveness of control mechanismsis required for sensible intervention to take place, therebyputting pressure on organizations to adopt such mecha-nisms otherwise they may be viewed as negligent. Signif-icant energy is also devoted to the development of formallists of symptoms and other tools designed to help recog-nize fraud. Diverse categories of symptoms have beenidentied: accounting anomalies, extravagant lifestyles,and a psychologically-predicated category, namely, unu-sual behaviors:

Research in psychology indicates that when people(especially rst-time offenders, as many fraud perpetra-tors are) commit a crime, they are overwhelmed by fearand guilt. These emotions express themselves in anextremely unpleasant physical response called stress.Individuals then exhibit unusual and recognizable cop-ing mechanisms. [. . .] People who are normally nicebecome intimidating and belligerent. People who arenormally belligerent suddenly become nice. (Albrecht& Albrecht, 2004, p. 99)

Coping mechanisms include insomnia, unusual irritabil-ity, inability to relax, inability to look people in the eyes,defensiveness, sweating and increased smoking (p. 100).Consequently, fraud investigators are encouraged to bealert to any behavioral changes surrounding them. Yetthe art of the skilled examiner is to separate the wheatfrom the chaff, which is recognized as being challenging(Albrecht & Albrecht, 2004, p. 103) given the impact ofunwarranted investigations on the examiners reputationand the professional liability potentially involved:

In putting forward your best efforts to detect fraud,youll be tempted to try too hard sometimes. You willweigh in your mind whether you should take an unau-thorized look at the suspects bank account; youll wres-tle with the dilemma of whether to secretly check thefraudsters credit records. Dont do it. Overreaching aninvestigation or fraud examination is the quickest wayto ruin it. Not only will you be unsuccessful in provingyour case, you will subject yourself to possible criminaland civil penalties. (Wells, 1997, p. 427)

Following the example of nancial auditing (Power,1994), but perhaps in a more pronounced way, the fraudexpertise discipline supports a climate of suspicion; it isas if to secure their jurisdiction, fraud specialists seek to re-mind people, organizational decision-makers especially, ofthe threat of fraud being constantly present in their sur-roundings. The concluding sentence in Wells (1997, p.528) book is particularly revealing in this respect: Andsince most people dont start their careers to become liars,cheats, and thieves, it is the fraud examiners job to ensurethey do not end up that way. That is, there is a potentialfor everyone to become a fraudster if the surrounding con-ditions, especially those within the organization, are con-ducive to this type of behavior. In a similar way:

Past research has shown that anyone can commit fraud.Fraud perpetrators usually cannot be distinguishedfrom other people by demographic or psychologicalcharacteristics. Most fraud perpetrators have prolesPlease cite this article in press as: Morales, J., et al. The construction offraud triangle. Accounting, Organizations and Society (2014), http://dx.dothat look like those of other honest people. (Albrecht& Albrecht, 2004, p. 18)

Such generalizing statements, however, mix with riskproling statements in ACFEs discourse. Fraud risk is bothwidespread and related to certain categories of people.Probabilistically, everyone is tied to some likelihood ofcommitting fraud yet it appears that the likelihood ishigher for certain proles. As such, the categorizingschemes institutionalized in the discipline emphasize thethreat of specic categories of individuals more likely toact in immoral ways: former prisoners, drug users, gam-blers, etc. Suspicion is, therefore, especially targeted to-wards certain categories of individuals, reecting somedeeply-ingrained beliefs that have historically stigmatizedcertain groups in society (Goffman, 1963). Further, somedata points to specic categories of employees as beingmore likely to be involved in fraudulent cases, while indi-rectly suggesting that fraud risk permeates every organiza-tion since these categories apply everywhere:

The six most common departments in which fraud per-petrators worked were accounting, operations, sales,executive/upper management, customer service andpurchasing. Collectively, these six departmentsaccounted for 77% of all cases reported to us. As thechart on the following page illustrates, the distributionof cases based on the perpetrators department wasremarkably similar to the distribution in our 2010study. (ACFE, 2012, p. 51)

The general impression that emerges from the materialwe analyzed is one of doubt and wariness that every em-ployee, citizen or corporation might harbor wrongdoing.Fraud, as conceptualized through the lens of the fraudtriangle and its intellectual ramications, can occur every-where. Yet some individuals are riskier than others. This iswhy, as the rhetoric goes, fraud can be everywhere;abnormal gestures or words should be construed assymptoms of fraud. Along these lines, Wells (1997, p.511) weaves wider webs of suspicion:

The cases we have seen on the preceding pages were, byand large, on the extreme edge of abusive conduct byemployees. In short, this data is merely the tip of theiceberg. How deep and massive that iceberg is variesfrom one organization to another, depending on a com-plex set of business and human factors. [. . .] Obviouslythe more rules within the organization, the moreemployees are likely to run afoul of them. [. . .] Tom R.Tyler, in his book Why People Obey the Law concludedoverwhelmingly that individuals obey only laws theybelieve in. If a rule makes no sense to the employees,they will make their own.

In the above quotation, Wells employs a series of per-suasive tactics in arguing that fraud constitutes a massiveorganizational problem that must be addressed. The meta-phor of the iceberg is interesting, since it surrounds fraudwith images of vastness and uncertainty. The clear-cuttone Wells uses is also noteworthy: employees willundoubtedly tend to break rules that do not makesense to them, if they have the opportunity to do so. Thisthe risky individual and vigilant organization: A genealogy of thei.org/10.1016/j.aos.2014.01.006

12 J. Morales et al. / Accounting, Organizations and Society xxx (2014) xxxxxxuncompromising assertion is authoritatively supportedwith a reference to a book written by an academic (Tyler,2006) specializing in psychology and law. Consistent withthis stance, organizations that are passive in terms ofimplementing specialized approaches and recipes aimedat controlling the risk of fraud are rmly criticized withthe benet of hindsight and wisdom:

If there is a lesson to be learned here, it is that auditfunctions are in place for a reason and should neverbe overlooked. Unfortunately for the [defrauded] chari-table organization, they were reminded of this lessonthe hard way. (Wells, 1997, p. 156)

Not only does the triangle offer a scheme to explainwhy fraud is committed, but it also provides a basic tem-plate for intervention thereby rendering the organizationresponsible for establishing a proper internal control struc-ture. The following metaphorical analogy explicitly estab-lishes a central linkage between fraud explanation andintervention:

Fraud resembles re in many ways. For a re to occur,three elements are necessary (1) oxygen, (2) fuel, and(3) heat. These three elements make up the re trian-gle, [. . .]. When all three elements come together, thereis re. Fireghters know that a re can be extinguishedby eliminating any one of the three elements. [. . .] Aswith the elements in the re triangle, the three ele-ments in the fraud triangle are also interactive. Withre, the more ammable the fuel, the less oxygen andheat it takes to ignite. [. . .] With fraud, the greater theperceived opportunity or the more intense the pressure,the less rationalization it takes to motivate someone tocommit fraud. [. . .] People who try to prevent fraud usu-ally work on only one of the three elements of the fraudtriangle opportunity. Because fraud-ghters generallybelieve that opportunities can be eliminated by havinggood internal controls, they focus all or most of theirpreventive efforts on implementing controls and ensur-ing adherence to them. (Albrecht & Albrecht, 2004, pp.20-21)

One of the key implicit messages in the above excerpt isto make the organization accountable for reining in the riskof fraud. As any responsible community cannot leave reunattended, any responsible organization must establisha proper structure of control around fraud; a failure to doso is bluntly viewed as carelessness. In this way, fraud spe-cialists promote the obvious necessity of their expertise,which is presented as making a difference in dealing withproblems that can have dire consequences on the negligentorganization. It should not be surprising, then, to see fraudexamination being highlighted as a growing careeropportunity:

As the number of frauds and the amounts of fraudlosses increase, so do the opportunities for successfulcareers in fraud-ghting. In fact, just recently, U.S. Newsand World Report identied fraud examination as oneof the fastest growing and most nancially rewardingcareers. The American Institute of Certied PublicAccountants recently touted fraud examination/fraudPlease cite this article in press as: Morales, J., et al. The construction offraud triangle. Accounting, Organizations and Society (2014), http://dx.doauditing as one of the six fastest growing andmost prof-itable opportunities for accountants (Albrecht & Albr-echt, 2004, pp. 13-14).

Further, the ACFE maintains that to establish a propercontrol structure, the organization should set up an appro-priate control environment:

Having an effective control structure is probably thesingle most important step organizations can take toprevent and detect employee fraud. [. . .] The controlenvironment is the work atmosphere that an organiza-tion establishes for its employees. The most importantelement in an appropriate control environment is man-agements role and example. There are numerousinstances in which managements dishonest or inappro-priate behavior was learned and then modeled byemployees. (Albrecht & Albrecht, 2004, p. 27)

The emphasis on control environment is presented asbeing experientially supported. The bar is set very highfor top managers, who should ensure not only thatappropriate control mechanisms are establishedthroughout the organization, but that they behaveappropriately at all times. Any temporary deviance fromthe dominant set of moral prescriptions is viewed ashaving the potential to engender dire cultural conse-quences, given the presumed behavioral isomorphismowing from top to bottom in the organization. Thepoint is that considering control issues seriously andmeticulously is a key attitude that top managers shouldmanifest at all times.

One of the key points to retain from our analysis thusfar is that, by formulating a knowledge claim groundedin what is represented as uncontroversial, fraud specialistslegitimize and disseminate a peculiar discourse aboutwhat fraud is, what its causes are, and who is responsiblefor controlling it. The triangle is represented as the coreof fraud examination expertise, surrounded with a constel-lation of applied knowledge renements to ensure thatfraud examination expertise can be meaningfully used inconcrete situations. This constellation is always underdevelopment, consistent with the image of a discipline thatgenuinely aims for continuous progress in its quest todominate fraud. It focuses on character traits of perpetra-tors, probabilities associated with fraud risk, and controlmechanisms to be implemented within organizations.One of the main emphases is that the fraud problem can,and should, be controlled by the organization throughthe implementation of well-tried anti-fraud techniques although the limitations of fraud control are sometimesprudently recognized in ACFE rhetoric, perhaps in an at-tempt not to set the audiences expectations of effective-ness at an excessively high level:

The dream of many in the accounting community is todevelop new audit techniques which will quicklyand easily point the nger of suspicion. To those inno-cent souls, good luck. Regardless of the ability of com-puters to automate a great deal of drudgery, there areno new audit techniques, and there havent been anyfor the last several centuries. (Wells, 1997, p. 526)the risky individual and vigilant organization: A genealogy of thei.org/10.1016/j.aos.2014.01.006

The chains of translations produced around the fraudtriangle thus attempt to build a network of support arounda notion of fraud highlighting the responsibility of individ-ual morality and the necessity to build vigilant organiza-tions. In the next section, we examine how this networksubsequently escaped the limits of fraud examination

J. Morales et al. / Accounting, Organizations and Society xxx (2014) xxxxxx 13and colonized other elds of intervention.

Moving downstream in the chain of translations:extending the network

Intermingling with the AICPA

The fate of any claim, including knowledge claims, liesin others hands (Latour, 1987). The ACFEs claim to exper-tise, grounded in the imagery of the triangle, spread be-yond its connes, thus extending the chain oftranslations to surrounding communities. In particular,the following auditing standards were inspired by ACFEswork, although none of them explicitly mentions thefraud triangle: Statement on Auditing Standards 99,Consideration of Fraud in a Financial Statement Audit,introduced by the AICPA to replace SAS 82 after the intro-duction of the Sarbanes-Oxley Act in the United States(American Institute of Certied Public Accountants (AIC-PA), 2002); and International Standard on Auditing 240,The Auditors Responsibilities Relating to Fraud in anAudit of Financial Statements (IFAC, 2006).20

In these standards, fraud risk factors are grouped inthree conditions that indicate incentives/pressures to per-petrate fraud, opportunities to carry out the fraud, or atti-tudes/rationalizations to justify a fraudulent action(AICPA, 2002, paragraph 31, p. 16; see also IFAC, 2006, par-agraph 24, section A25, p. 174). As in Wells writings(1997), the non-shareable nancial problem has disap-peared and been replaced by an incentive or pressure tocommit fraud, while the notion of rationalization is linkedwith that of attitude, thus shifting closer to the idea of indi-vidual morality:

Three conditions generally are present when fraudoccurs. First, management or other employees have anincentive or are under pressure, which provides a reasonto commit fraud. Second, circumstances existforexample, the absence of controls, ineffective controls,or the ability of management to override controlsthatprovide an opportunity for a fraud to be perpetrated.Third, those involved are able to rationalize committinga fraudulent act. Some individuals possess an attitude,character, or set of ethical values that allow them toknowingly and intentionally commit a dishonest act.However, even otherwise honest individuals can com-mit fraud in an environment that imposes sufcientpressure on them. The greater the incentive or pressure,the more likely an individual will be able to rationalize

20 This is not the only inuence that Cressey had on auditing practices: healso participated (with Charles Moore) in a study commissioned by PeatMarwick & Mitchell whose ndings, submitted in 1980, underscored theimportance of tone at the top (Cressey & Moore, 1983).Please cite this article in press as: Morales, J., et al. The construction offraud triangle. Accounting, Organizations and Society (2014), http://dx.dothe acceptability of committing fraud. (AICPA, 2002,paragraph 7, p. 8)

Readers may nd this translation of Wells work by theAICPA (and the International Federation of Accountants IFAC) rather surprising. To begin with, the transpositionto accounting fraud is not questioned:

Even if Cresseys ndings for embezzlers were valid,there is little evidence to support the fraud triangle asa general theory of nancial crime. The convoluted con-nection between Other Peoples Money and SAS No. 99suggests that the endorsement of the fraud triangle bythe AICPA occurred without serious consideration ofother criminology perspectives. (Donegan & Ganon,2008, p. 3)

Second, in contrast with Cressey (1953) giving lessweight to opportunity, the notion that is central for boththe AICPA and IFAC is, indeed, that of opportunity, whosedimensions are close to those of internal control. Third,replacing the concept of a non-shareable nancial problemwith those of incentive and pressure means that the iden-tication of factors that may have led the individual tocommit fraud, is pushed even further into the background.The notion of incentive is based on the idea that the fraud-ster has a reason to commit fraud. Finally, rationalizationis overshadowed by an attitudinal stance that is prone tocommitting fraud: individuals with a character or set ofethical values which allow them to knowingly and inten-tionally commit a dishonest act. This reintroduction ofmoral traits, anchored relatively permanently in the iden-tity of individuals, as an explanation for fraudulent behav-ior is far removed from the spirit of many writings incriminology, including those of Cressey and Sutherland.

This redenition of the three branches of fraud identi-ed by Cressey allows us to better understand the defor-mation mechanisms at work in the translation ofcriminology research, such as it was redeployed in thefraud examiner and auditing community. The notion of anon-shareable nancial problem has been replaced withthose of incentive and pressure, and perception of opportu-nity with a failure in internal controls. These changes arenot only a matter of replacing subjective, hard-to-observeelements with procedures in documentation and measure-ment that have already been mastered and put in place;they also imply subjecting any effort to understand thecauses of fraud to the will to manage the related risks.Cresseys causal view has been abandoned in favor of thenotion of risk.

For example, in their appendices the two standards pro-vide examples of risk factors for fraud that are very similarto what was presented in previous standards, but this timethey are organized systematically according to the threedimensions of the triangle. They are also separated intotwo parts: the risk factors relating to misstatements aris-ing from fraudulent nancial reporting (AICPA, 2002, pp.4448; IFAC, 2006, pp. 186189) and the risk factorsrelating to misstatements arising from misappropriationof assets (AICPA, 2002, pp. 4850; IFAC, 2006, pp. 189191). The former concern accounting fraud and can beattributed more to the behavior of top managers, whereasthe risky individual and vigilant organization: A genealogy of thei.org/10.1016/j.aos.2014.01.006

the latter concern individuals or groups of individuals whodefraud their employer for their own benet; these arewhat Cressey (1953) calls embezzlement. The incentivesidentied for accounting fraud relate to the highly compet-itive environment of the organization, the pressure facedby managers to reach nancial targets and the nancialstake that executives may have in their organization.Opportunities relate to the complexity and lack of trans-parency in the organizations operations and to decien-

their professional jurisdiction, while accounting and audit-ing professional associations were trying to reassert,through the triangle, their legitimacy as guardians of trustafter accounting fraud scandals. Although these groupsarguably had different views on how and why to use thefraud triangle, they shared an interest in promoting itand its underlying vision of fraud.

All in all, it appears that the turning point in the early2000s created a context that was favorable to an increaseduse of the fraud triangle. As such, our analysis indicatesthat the triangle has become an important concept inestablishing jurisdiction over fraud and in making theorganization responsible for the control of its membersmoral deviancies. Of course, instances of triangle use arenot limited to mere diffusion; the chains of translations ex-tended beyond the ACFE community. Now, we trace, inmore detail, the web of downstream translations, usingthe 64 articles that we identied (mostly in academic/edu-cational journals but also in a number of professional andhybrid journals) as specied in our research methodssection.

Justifying the emphasis on internal control

A number of recent articles incorporate the fraud trian-gle within the core of their argument. Some articles aim to

14 J. Morales et al. / Accounting, Organizations and Society xxx (2014) xxxxxxcies in internal controls. Rather tricky to observe, theattitude of management may be evaluated by examiningits level of commitment and communication on ethical val-ues, the attention paid to short term results and its ten-dency to commit to unrealistic nancial forecasts.21

As stated in the standard, fraud risk factors do not nec-essarily indicate the existence of fraud; however, they of-ten are present in circumstances where fraud exists(AICPA, 2002, paragraph 31, p. 16; see also IFAC, 2006,paragraphs A23A27, p. 161).22 Like the symptoms of a dis-ease, the presence of risk factors implies that one must beconscious of the possibility of fraud, even though no causalrelationship can actually be established between fraud andthe underlying risk factors. Since risk, by denition, canmaterialize anytime, the standards tend to promote a cli-mate of suspicion justifying the reinforcement of organiza-tional surveillance.

Our analysis also indicates connections between theACFE and auditing standard setters. Indeed, Wells partici-pated in drafting an exhibit (Management AntifraudPrograms and Controls. Guidance to Help Prevent, Deter,and Detect Fraud) associated with SAS 99 (AICPA, 2002,pp. 5777) and issued jointly by different associations,including the ACFE. Further, SAS 99 recommends callingin fraud examiners identied as key participants in theantifraud effort to work with the organizations adminis-trators and auditors (AICPA, 2002, pp. 59; 76). This alliancebetween two associations must be placed within the con-text of the nancial scandals of 20012002:

When the AICPA belatedly recognized the need to con-sider why so many accountants were committing fraudthey turned to the ACFE, which, in effect, meantembracing Cresseys perspective. Adopting the programof a group of dedicated fraud ghters was doubtlessly auseful way to burnish the professions tarnished imagein 2002. The triangle had the added advantage ofexplaining fraud as the action of a loner driven by need,taking advantage of a lack of internal control. Thus thedeterrent to fraud was more internal control, and thesearch for the culprit could focus on individual offend-ers, not on the culture that may have encouraged andrewarded their actions. (Donegan & Ganon, 2008, p. 3)

21 We see here how auditing standard-setting bodies translate thetriangle into actionable knowledge which gives them power over reality.Through various problematization and categorization proce