Transcript

The Dark Side of Organizations: Mistake, Misconduct, and Disaster Author(s): Diane Vaughan Source: Annual Review of Sociology

, Vol. 25 (1999), pp. 271-305 Published by: Annual Reviews Stable URL: http://www.jstor.org/stable/223506 Accessed: 07/01/2009 17:01
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=annrevs. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that promotes the discovery and use of these resources. For more information about JSTOR, please contact [email protected].

Annual Reviews is collaborating with JSTOR to digitize, preserve and extend access to Annual Review of Sociology.

http://www.jstor.org

Annu.Rev. Sociol. 1999. 25:271-305 Copyright? 1999 by AnnualReviews.All rights reserved

THE DARK SIDE OF ORGANIZATIONS: Mistake, Misconduct,and Disaster
Diane Vaughan
Departmentof Sociology, Boston College, ChestnutHill, Massachusetts02167; e-mail: [email protected] KEYWORDS:analogical rational choice culture, deviance, theorizing, cognition,

ABSTRACT In keeping with traditionalsociological concerns about order and disorder, this essay addresses the dark side of organizations. To build a theoretical basis for the darkside as an integratedfield of study, I review four literatures in orderto make core ideas of each available to specialists in the others. Using a Simmelian-basedcase comparisonmethod of analogical theorizing, I first consider sociological constructs that identify both the generic social form and the generic origin of routinenonconformity:how things go wrong in socially organizedsettings. Then I examine threetypes of routinenonconformity with adverse outcomes that harm the public: mistake, misconduct, and disaster produced in and by organizations.Searchingfor analogies and differences, I find that in common, routinenonconformity,mistake, misconbetween duct, and disasterare systematicallyproducedby the interconnection environment, organizations, cognition, and choice. These patternsamplify what is known about social structureand have implications for theory, research, and policy.

INTRODUCTION
Weber warned that a society dominated by organizations imbued with legalrational authority would suffer negative consequences. Tracing that historic transformation, Coleman (1974) affirmed Weberian pessimism. He observed

271 0360-0572/99/0815-0271 $08.00

272 VAUGHAN

that this change altered social relations:Individualsnot only interactedwith individuals as before, they also interactedwith organizations,and organizations interactedwith other organizations.Coleman's primaryinsight was that this structural transformation producedboth perceived and real loss of power for individuals.But the rise of formal organizationsalso wroughtnew possibilities for adverse societal consequences as a result of mistake, misconduct, and disaster.Surprisingly,these harmfulactions and the extensive social costs to the public-the darkside of organizations-are not claimed as centralto the domain of sociologists who define their specializationas organizations,occupations, and work, althoughprima facie, they would appearto fall within it. Organizationalsociologists have affirmedthat formalorganizationscan deviate from the rationalistexpectationsof the Weberianmodel; also, the pathologies that harmmembersare partof mainstreamorganizationtheory. But only recently have textbooks included harmful outcomes and organizationalpathologies that adversely affect the public (Perrow 1986, Hall 1996, Scott 1998), and collections addressedfailure,crime, and deviance in andby organizations (Anheier 1998, Bamberger & Sonnenstuhl 1998, Hodson & Jensen 1999). Ironically,many organizationalsociologists have been workingon the dark side for a long time, as this essay shows. Moreover,scholarsin medical sociology, deviance and social control,andthe sociology of science, technology, and risk have studiedmistake,misconduct,and disasterproducedin andby organizations, as have scholars in other disciplines. The irony here is that we learn much abouthow things go wrong, but absentthe tools of organizationtheory, the full set of socially organizedcircumstancesthatproducethese harmfuloutcomes remainsobscure. Specializationwithin and between disciplines segregates knowledge, with four consequences. First, the tensions and affinities in relevantwork arenot visible, theoreticalmattersfor debatehave not been identified, andthe dialogue essential to intellectualdevelopmentis absent. Second, both the social origins and cumulative significance of harmfulorganizational outcomes are masked. Third,the sociological basis for policy implicationsfor organizations,the public, and agents of social control-remains underdeveloped. Finally, a broadertheoretical issue is at stake. This topic moves us away from rationalchoice assumptionsabout means-ends oriented social action towardexplanationsof socially patternedvariationsfrom that model. A substantialbody of scholarshipexists, but the darkside of organizations is not an integratedfield. The purposeof this essay is to legitimate it as a field of study, bridging disciplinary boundaries by reviewing four literatures in orderto make core ideas of each availableto specialists in the others.Muchhas been written about improving and/or regulating organizationsto reduce the possibility of harmful outcomes, but unmasking causal structuresand profor cesses is the focus here. I begin by searchingthe sociological literature gen-

misconduct. Unanticipated consequences of purposive social action can be differentiatedinto consequences to the actor(s) and consequences to others that are mediated culture. Formalorganizationsare designed to producemeans-endsoriented social action by formal structuresand processes intendedto assure certainty.His point is that outcomes can be unexpected. legal mandates. I identify analogies and differences across these three types to show where progress has been made.Ourprimaryconcernis unanticipatedsuboptimaloutcomes.occurring in and/orproducedby aformal organization. and/or the response of some public to the incident. either in the fact of its occurrence or in its consequences.the social group whose expectations are violated. Actors can be inthroughsocial structure.I will define an event. Using a Simmelian-basedcase comparison method of analogical theorizing that legitimates theory building from comparing similar phenomena occurring in different social settings (Simmel 1950. so the first step is to invent a conceptualdefinition (tentativeand subjectto later revision)thathelps us considerthe formand originof unanticipated suboptimal outcomes.social expectations). conformity. 1968) thinkingis the foundationof any considerationof the darkside of organizations. or circumstance.thatdeviatesfrom bothformal design goals and normativestandardsor expectations. and produces a suboptimaloutcome as organizationaldeviance. . and disaster. Vaughan 1992.and civilization (1936:895).the consequencesmay be eitheroptimalor suboptimal. it encompasses mistake. THE SYSTEMATIC ROUTINENONCONFORMITY: PRODUCTIONOF ORGANIZATIONAL DEVIANCE Merton's(1936. Therefore.He observed that any system of action inevitably generatessecondaryconsequencesthatruncounterto its objectives.and goal attainment. and disaster. activity. This conceptualdefinition is sufficiently broad to include conforming behavior by individual members as well as deviant behavior by individual members.DARKSIDEOFORGANIZATIONS 273 eral principles-theories and concepts-that help explain both the social form and generic origin of routinenonconformity:how things go wrong in socially organized settings. A conceptual language that helps organize the literaturedoes not exist. misconduct. and speculate about futuredirections.one differing from the othernot in its general fit with this definition but in its particulars:the normative standardsor expectations violated (internalrules.Therefore. three types of routine nonconformity with adverse outcomes that harm the public. Then I review the literatureson mistake.the extent of harmfulsocial consequences. assess the adequacy of this work.the categories of the public harmed. 1940. dividuals or organizations. 1999a).

and (c) the cognitive practices of individuals within them. and the leash thatjoins them) ordinarilyfunctions in an acceptable way. Becker 1994. a breakdown in some normally wellfunctioningorganizationalsystem may occur. It follows that the same characteristics a system thatproducethe brightside will regularly of the darkside from time to time. gives chase. Mannis & Meltzer 1994.processes. Aubert 1965. Alternatively. and we are concerned with both form and origin. Analogically.the languageandpersonalexperience of mistakemight suggest that some types of deviance resultfromcoincidence. can be understoodas routinenonconformity:a predictableand recurring productof all socially organized systems. but we don't know how or to what extent chance explains social life. This finding is consistentwith Durkheim'sthesis thatthe pathologicalis an inextricablepart of every social system because the conditions of the normal are the preconditionsof the pathological ([1895]1966:47-75).both denotingthe causal origins of unanticipated tive outcomes. Not in all cases did the originatorsmake the link between . but this system produces an unanticipated suboptimal outcome when the dog is startledor surprisedby some aspect of the environment. Meltzer & Mannis 1992. Tilly 1996).so backs up. causing an embarrassingstumble or fall. owner. synchronicity. leading to unanticipatedsuboptimal outcomes. in its generic form.much organizationaldeviance is a routineby-productof the characteristicsof the system itself. it shows not nly thatunanticipatedsuboptimaloutcomes are a routine aspect of social life. Paget 1993. tasks). Finally. but also thata conceptualapparatus alreadyexists for studyingthem as socially organized phenomena. however. However.274 VAUGHAN To define organizationaldeviance is not to explain it. Carryingthe analogy further. the simple system that provoke enables the owner to walk the dog (dog. the sociological theories and concepts assembled below assert that the system responsible for the productionof routine nonconformity includes (a) the environmentof organizations. the literatureshows that the social origin of routine nonconformity is in the connectionsbetween them. ciologists Zelizer 1979. I use routinenonconformity and the systematic production of organizationaldeviance as internegachangeableconstructs. or circles so that the leash entangles the owner. my searchof the sociological literaturefor theories and concepts with explanatorypower shows that whateverthe role of chance or system breakdown. Althoughthese threelevels of analysis aretopically separatedfor this discussion. Organizationaldeviance.or chance.this review affirmsthataspects of social organizationtypically associatedwith the brightside also are implicatedin the darkside.(b) organizationcharacteristics (structure. For example. Soorganizational have written about chance (Sumner [1906] 1940. To the sociologically uninitiated. Only apparentlysimple. In this essay. the dog-owner-leash system also encompasses the external environmentand the interpretivecapacities of dog and owner. runsa zigzag.

aggressive and changing. conflict. Although Stinchcombeused liabilities of newness to explain the developmentof populations of organizationalforms.and suppliersmake them scarce. the focus shifts from organizationsadaptingto an uncertainenvironmentto organizationsactively defining. economic..Because organizations have difficulty accuratelypredicting circumstancesthat might affect their future activities.and as conveying jolts to unsuspecting organizations(Hall 1996). 1991). Perrow& Guillen). Traditionally in organization theory.two classic concepts relevantto routinenonconformity(Selznick 1949). In this competition.power is both a means andan end. creating. technological. explaining why new programs.the process of mediatingthreatsto stability from the environment by absorbingnew elements into the leadershipor policy-determiningstructure of an organization. and the general conditions of the social context (political.It has been described as complex. turbulent. When power is used as the central explanatory concept. because the activities of regulators. all organizations must compete for resources that help them meet goals. or even the problems of neophytes in organizations. e. legal. Cooptation. Pressman& Wildavsky 1973. unexpected negative outcomes are possible even when initial conditions are diagnosed as optimal (see. Poorly understoodsocial conditions even can lead to the failure of the organization to survive. costs in time.competitors. Often. worry. cf. these resources are scarce because their naturelimits supply. the absence of standardroutines.DARKSIDEOFORGANIZATIONS 275 their theoretical constructs and routine nonconformity.or services by individual organizationsget into trouble. and inefficiency. Stinchcombe's "liabilities of newness" (1965:148) explain comparativedeath rates of new and old organizations:A higher proportionof new organizationsare likely to fail thanold.products.his concept also may applyto analogicalcircumstances at other levels of analysis. In orderto survive. Wilson et al 1996. and shaping it to suit their needs (Perrow 1986. Environmentaluncertaintyis a concept frequentlyassociated with routine nonconformity.can resultin compromisethatdeflects an organizationfrom .I have drawn out the implications when doing so seemed theoreticallyappropriate. Power struggles may result in cooptation and goal displacement. or because of pre-existing commitments thatkeep organizationsfrom getting the resourcesthey want.g. demographic. the absence of stable ties with consumers. The liabilities of newness include the necessity of generating and learningnew roles. the necessity of relying on social relations among strangers. the environment includes both the organizationset (networks and other kinds of interorganizational relations). Environmentand RoutineNonconformity Less is known abouthow the environmentcontributesto routinenonconformity than is known about the role of organizationcharacteristicsor cognition and choice. and cultural).ecological.

and individuals). attemptsat purposive action are embedded in concrete. unpredictable. Because the generalizedrules of the institutionalizedenvironmentareoften inappropriate specific situations.rapid decision-making in a crisis for but cumbersomeproceduresthat stall routinedecisions (Staw et al 1981). moreover. the embeddedness perspective is anothertool for explaining the systematicproductionof organizational deviance.the embeddednessperspective pected also applies intra-organizationally: The relational ties within organizations that generatetrustand control malfeasanceprovide memberswith opportunities for deceit and misconduct (Granovetter 1985:499-502).But centralizationand formalization. concepts relatingstructure routinenonconformityareplentito ful. but these structures and practicesmay be inefficient or inappropriate their tasks. also have their dark side. Both contingency and constraintexplain economic action. thus. They show that organizations sometimes incorporate structuresand practices that conform to institutionalizedculturalbeliefs in orderto gain legitimacy. so thatoutcomes deviate fromnormativestandards expecor tations. In addition to these theories. the number of written Organization Characteristics . Interestingly.institutedto keep things from going wrong. High levels of centralizationprovide greatercoordinationbut less flexibility. consistent policy that is inappropriate specific situations.276 VAUGHAN its originalgoals. so unexto adverse outcomes result. thus defining legitimategoals for them to pursueand affecting action and meaning at the local level. and Routine Nonconformity STRUCTURE Meyer & Rowan (1977) debunk the Weberian notion of efficient structurewith their "myth and ceremony" argument. Granovetterpoints out the ironic link between the bright side andthe darkside: The very concretesocial relationsand structures net(or works) in the environmentthatplay a role in generatingtrustand discouraging wrongdoing in economic exchange also increase opportunitiesfor deceit and deviance (1985:491-93). organizations.professions. Legitimacy in the institutionalrealmcan be not only a means to some end. The new institutionalismexplains thatorganizationalforms andbehaviors reflect prevailing values and beliefs that have become institutionalized. to outcomes may be suboptimaland. ongoing systems of social relations that affect them (1985:487). Culturalrules constitute actors (states. to some extent.For Granovetter. Complexity can make an organizationunwieldy so that the upper levels cannotcontrolthe subunits. To these concepts may be added two recent theories that locate causes of routinenonconformityin the environment:the new institutionalism(Powell & DiMaggio 1991) and the economic embeddednessperspective (Granovetter 1985). but an end in itself. Formalization can never cover all conditions. In contrast to the new institutionalism. agency is central to the embeddedness perspective.

A substantial literature connects routine decision making by organizationalelites with unanticipatedadverse outcomes for disadvantagedothers. eitherby the deletion of informaoccurs as an tion or by distortion(Guetzkow 1965).and (c) segregatedknowledge minimizes the ability to detect and stave off activities that deviate from normative standardsand expectations. and/oracceptabilityhave known associationwith the systematicproduction of organizational deviance (Blau 1955. More informationtypically is viewed by people in organizationsas the solution. perpetuatingpartial understandingand the possibility of unexpected negative outcomes.hierarchy. information can contribute to routine nonconformity both when there is too little and when there is too much (Feldman 1989).In his portrayalof how working class kids get working class jobs. vagueness. "Structural secrecy"refersto the way division of labor. Willis (1977) showed how educationpolicy designed to enable studentsto do betterafterhigh school in fact promotesrebellion that perpetuatestheir working class condition. Vaughan 1982a. Diamond (1992) revealed how the demandsof corporatebalance sheets andregulationsdistortthe daily rituals of nursing homes for the elderly. Structuralsecrecy implies that(a) informationandknowledge will always be partial and incomplete. Burawoy 1979). determining the relevance of informationby its social appropriatenessas well as its technical accuracy.277 DARKSIDEOFORGANIZATIONS rules and procedures. Confrontedby these obstacles.their recency.and specialization segregateknowledge abouttasks and goals (Vaughan 1996). PROCESSES Case studies hold memorablelessons about how organizational processes systematically produce unanticipatedoutcomes that deviate from formal design goals and normative standards. . power as process. Elsbach & Sutton 1992). Structuralsecrecy is reinforced as messages are as transformed they pass throughthe system. ratherthan structure. perceived relevance. with adverse consequences for patients. "Uncertaintyabsorption" technical language and classification schemes prevent some organization's kinds of informationfrom being communicated(March & Simon 1958:165). complexity. and Chambliss(1996) showed how the bureaucratic machineryof health care converts ethical decisions abouthospitalpatientcare to conflicts between occupational groups.Either explicitly or implicitly. The result is an informal network that excludes certain knowledge claims. Oliver 1991. eroding quality care. Most recently. They sortthroughknowledge claims. Unfortunately. decision-makersemploy the "micropoliticsof knowledge"(Lazega 1992). Classics show how informalorganizationcontributesto routinenonconformity at both the top of the hierarchy(Dalton 1959) and the bottom (Roethlisberger & Dickson 1947.is a central concept in these cases. (b) the potential for things to go wrong increases when tasks or informationcross internalboundaries.

in the organizationalbehavior literature. frequencyof performanceof the skill. . other scholars offer concepts that explain why attemptsto put out a fire may inadvertentlyfan the flame. Verifying this ironic consequence. Finally. His prescience aboutthe theoretical importanceof risk is borne out in the discussions of mistake. so some valuablepolicies arenever enactedand some socially harmfulpolicies may be (Fligstein 1990).Arguingthatunexpectednegative outcomes areindigenous to the work process. The response to errorincreases the rigidity of the organization. made a connectionbetween the brightside and the dark side of organizations. misconduct. Crozier(1964) found thata bureaucratic system of organizationcannotcorrectits errorsbecause the feedback process does not function well. althoughnot ignored. These responses to errorare systematic in origin and defeat the goals they are intendedto achieve. Learning. and role in the workplaceas a social system. spread. he theorizedaboutvariationsin their frequencyand probabilitydue to variationin amountof occupationalskill.and disasterthat follow. occupation is the key concept.is a process normally associated with the bright side. TASKS Hughes (1951) was the first sociologist to look for theoreticalprinciples associatedwith task-relatederrorsin the workplace. with major unanticipatedconsequences. Mechanic(1962) locateda cause of routine nonconformityin the power of lower participantsto subvert the formal goals of organizations.may produceconformityandpositive outcomes when thatpower is used to prevent organizationalacts with harmfulsocial consequences. How organizations learn the wrong thing is understudied(Marchet al 1991).For Hughes.surely a deviation from design expectations in most hierarchicalorganizations. Schulman(1989) found that decisions to correct errorscan spiralinto "error-amplifying decision traps. too. Parallelprocesses thatmake sense in terms of different subunits pursuing their own goals may produce a joint outcome not intended by anyone and directly counter to the interests motivating individual actions (Giddens 1984:9-14).perpetuatingthe productionof routine nonconformity. Risk is also a centralconcept for Hughes:the distributionof risk among occupationalroles and how systems delegate.278 VAUGHAN Power struggles aboutsubunitgoals can produceunanticipatedsuboptimal outcomes when the most powerful subunitor coalition consistently triumphs.The power of lower participants."A simple erroris transformed into organizationalpathology by resonate changes in organization structure: Efforts to correctand cover up involve more participantsand more actions that increase both the amount of deviation and the possibility of discovery.or concentrateboth the risk of mistakes at work and the losses thatresult from them.But he. "Deviance amplification"(Weick 1979) is the result of causal loops typified by interdependenceand feedbackthatmakes any small deviation grow.

For example.those who carry out the task (Collins 1974.several richly detailed analyses of work-relatedtechnology have laid bare its vast unanticipatedconsequences (Turkle 1995.the more elements unique to a work situationare lost: the ad hoc strategies. Thus. Following Perrow(1984). Rochlin 1997). or object but are true only in the sense that they shape understanding of it.it is acquiredby the "core set". a large body of work shows how the characteristicsof the technology contributeto negative outcomes (see section on disaster.with rules emerging from practiceand experience ratherthanpreceding it.concrete circumstances.and local knowledge that keep organizationsgoing (Star 1995). In conditions of uncertainty. so that routine nonconformityis a normalby-productof techno-scientific work. the core set cannot express essential understandingsto others. in her researchon human/computer the design of intelligentmachines. which in turnmay lead to unanticipatedsuboptimaloutcomes. workarounds. "Interpretive flexibility" is the capacity differentmeaningsby different for scientific or technological facts to be given actors (Pinch & Bijker 1984). and unanticipatedadverse outcomes. By definition. technology. Wynne (1988) notes that normal technology is "unruly": typically occurs underconditions of ambiguityin unclearcircumstances. . and routine nonconformityas a generic social form requirestreating technology both as an object with a determinate physical essence that has consequences thatarereal and as an artifactembeddedin social context within which it has meaning(KnorrCetina 1997). diagrams. fundamentalambiguitiesremain. Understandingthe link between tasks. and other artifacts that are means of making decisions in all kinds of technical work (Lynch & Woolgar 1988). Many concepts lend themselves to the study of the relationship between uncertainty.But the more general the formalrepresentation. Tenner 1996. records.)Taking the latter perspective. "Tacit knowledge" refers to intuitive understandingsaboutpractice that cannotbe articulated. Takingthe formerperspective. 1981). Despite these efforts. Suchman(1987) examinedthe discrepancy between "plans"as models for action and "situatedactions"as actions takenin the context of particular. convertinguncertaintyto certainty. These technologies stand in for and representsome incident.279 DARKSIDEOFORGANIZATIONS Tasks involve technologies. imperfect knowledge. decisionmakers outside the core set will always have imperfect knowledge. process.so thatadverse outcomes are always a possiEngineering bility. "Representational technologies" are models. all judgments are made under conditions of imperfect knowledge. actors convert disorderto order (Collins 1992). Because of scientific and technical uncertainty. scholars have examined scientific and technical practice using ethnomethodological and other interactionistperspectives to of focus on the social construction/production knowledge andmeaning(Knorr interactionin Cetina 1981).

which will feed back to promotesocial reproduction institutionalized of practicesbecause of routinization:"Repetitive activities. Culturalknowledge is instrumentalin complexity reduction (Sackmann 1991).280 VAUGHAN Cognition. organization. and RoutineNonconformity Psychologistsprovidemany insightsthatpoint to cognitive originsof unanticipated suboptimaloutcomes (see Heimer 1988. cognition.Culturalunderstandingsaffect interpretivework. individuals make the problematic . located in one context of time and space. DiMaggio 1997). It shows how culturalknowledge contributes to unanticipated negative outcomesby enablingindividualsto violate normative standards. even when the behavior in question is objectively deviant. unintendedby those who engage in those activities. ratherthan contradict.Tilly explains that responses to errorproduce stable social structureand processes because they are drawnfrom historicallyaccumulatedculturalunderstandings and embeddedsocial relationsthat modify.Asserting thaterroris incessantin social interaction. and action. This work is abundant.but Tilly (1996) supplies an answer. New institutionalists posit cultureas a mediating link: Institutionalizedculturalbeliefs have a complexity reduction effect that determineswhat individuals will consider rationalat a given moment (DiMaggio & Powell 1983. Other scholars have considered how institutionalizedculturalunderstandings mediatebetween the environmentandthe cognitive practicesof individuals.Choice. in more or less 'distant' time-space contexts" (1984:14). instead trying to track the problem of culture and cognition. choice. One line of thought is that by drawing on largely unconscious cultural knowledge. It begins to clarify the relationship between environment. Anotherstrandof theorytakes for granted institutionalizedcultural knowledge. and at threelevels of analysis. At the institutionallevel. have regularized consequences. However. How culturetravels from the institutional level to "manifest in people's heads" remains a central theoretical and empirical dilemma (DiMaggio 1997:272). theoreticallyrich. this discussion focuses on sociological constructsthat demonstrate the influence of social context on decision processes. so that people may see their own conduct as conforming. Giddens' structurationtheory locates the origins of routinenonconformityin the gap between the known and understanding about its significance (1984:xxiii). and routinenonconformity. Giddens leaves the details of this process to the imagination. Reason 1990. Zucker 1977). structuration theory and culturaltheory are explanatory tools that emphasize how aspects of the environment can shape cognitive limits to rationality.what came before. He arguesthatthis discrepancywill always have the potential to produce unanticipatedconsequences. thereby shaping and narrowingunderstandingso that unexpected adverse outcomes are one possibility.

Vaughan 1996). situprimacy ated and emergent meanings.negotiation.the "garbagecan model" rejects it in favor of organizationalanarchy(Cohen et al 1972). interpretative work. or no solutions may appearat all."Merton's "bureaucratic identified how bureaucratic sonality. so that in their view their action is acceptableand nondeviant prior to an act (Stokes & Hewitt 1976. spontaneousnonrationalaction.March & Simon (1958) emphasizedthat the norm of rationalityitself cannot be viewed as a set standardbut must be seen ratheras a guideline from which deviation is expected.Barnard(1938) arguedthat organizationelites set the premises of decision-makingby setting up routines. Symbolic interactionism. so that internalactivity is often the result of habit and routine-following rather than consideration of permultiple options.He demonstratedthat in . It gives can to agency. By drawingon culturalscripts.solutions may arise that no one envisioned. The metaphor of the garbage can reveals a decision-makingprocess in which problems and solutions are loosely coupled. and the symbolic foundations of thought. The effect is to legitimate deviant actions. many classics have exposed cognitive limits to rationality. Goffman. Alternatively. interpretation. so that one problemmay get transformedinto a different one or may disappeardue to inattention.and orderingof the social. individualengagementin routinenonconformity can encompass violation of normativestandardsand expectations that are either internalor externalto the organization. The constrainingqualities of the social world are influential. yet the emphasis on agency and meaning constructionunderscoresthe potential for unanticipated consequences as a normalproductof the ongoing creation. in all its variations. They located systematiclimits to rationalityin the inability of organizationsto provide adequate information for decisions and individual cognitive constraints that limit the ability to adequately assess the information at hand. for example. specialized in exposing the micro-origins of routine nonconformity. This perspective suggests that unexpected interactions and outcomes can occur in systems of varying complexity. Morrill et al 1997." Bounded rationalitymodifies rationalchoice theory. At the organizationallevel of analysis.281 DARKSIDEOFORGANIZATIONS nonproblematicby formulatinga definition of the situation that makes sense of it in culturalterms. Decision-making is typified by "bounded rationality" and "satisficing. individuals may justify organizational deviance in retrospectby constructing accounts that bring their actions into harmonywith social expectations. contests the assumption of an objective reality in which "truth" and "error" be reliably ascertainedby individuals. Veblen's "trainedincapacity." and Thompson's "bureaupathology" systems can create extreme rule-mindedness that deflects individuals from actions that are most beneficial to the organization. At the micro-level. Zucker 1977. theory in social psychology has relentlessly pursuedthe ambiguousbases of decision-makingand its outcomes.

three things are clear. Martinet al 1983. Because the actor's subjective understandingand the loose coupling between informationand action dominateinterpretiveprocesses. First.or- .and disasterin these same pages. so paradoxesand contradictions and are common. Errors in troubles. Because meanings cannot be separatedfrom social structureand social relations. intentionaland unintentional. unanticipatedsuboptimal consequences are constant possibilities. or idea is selected for attentionand linked to a more general form in order to make sense of it. how things go wrong in socially organized settings.and enactmentcentralconcepts. limited by the agendato consider mistake. organizing. These can be used in research on mistake. forming a condition for future action. in which an observation. Nonetheless. Summary To legitimate the darkside of organizationsas an integratedfield of study.we have searchedthe sociological literaturefor theories and concepts that help explain.The role of social context as interpretivedevice and constraint is exemplified by his concept of framing(1974). numerous theories and concepts exist that explain routine nonconformityas a generic phenomenon.loose coupling. environment and organizationare enactedmoment to moment as people perceive and select the objects and activities on which sensemakingwill be based. Misperceptionis a function of the normal selectivity of the process.are integralaspects of social interaction.282 VAUGHAN social interactionthings are never as they seem because impressionmanagement and deception. This search is by no means complete. Second. rationalityis a definition of the situation bestowed only in retrospect. generically. Also. Boden 1994).experience. dramaturgy. misconduct. Manning (1992) combines loose coupling. identifying patterns from an inductive analysis of this literatureshows that the social form of routinenonconformityincludes aspects of environment. because the goal is to identify relevant causal mechanisms for the first time."among framing(1974:308-21) may resultin whathe calls "ordinary them actions that have unanticipatedsuboptimal outcomes (cf.and semiotics to examine organizationalcommunication. the important question of the variablerelationshipbetween the principles identified and the frequencyandprobabilityof unexpectedadverse outcomes is assumedbut not discussed. Two theories combine aspects of symbolic interactionism explain cognito tive processes associated with routine nonconformity. misconduct. Emerson & Messinger 1977). and differences. communication consists of signs symbolically marking authority.Weick (1995) makes sensemaking. Misunderstandings incomplete knowledge are typical (Putnam 1986. For Weick. Communication is comprised of multiple realities emanatingfrom multiple organizationroles. power. and disaster so that similarities and differences can be identified and the development of general theory about the dark side of organizationscan begin across types.

MISTAKE. Emerson observed that instead of using criteriafrom outside the setting to examine mistake and error. shifting fromroutinenonconformity as a generic social form to these three types creates a new definitional difficulty.mistake. presenting central themes in orderto facilitate discourse across disciplinary boundaries.it provides an heuristicdevice for consideringmistake.However. Also. misconduct. or circumstance. looking for analogies and differences across the three types (Simmel 1950. Third. Equally importantis determiningwho has the rightto say what a failureis (Hughes 1951).whetherit is defined as mistake. and disasterare socially defined in relationto the norms of some particulargroup.and disaster. or disaster will vary by group. Mistake. organization characteristics. Now we turnto the three types of routine nonconformity that have unanticipatedconsequences that harmthe public. misconduct. Because this definition could include conformingbehaviorby individualmembersas well as deviantbehavior by individualmembers. personal communication). misconduct. tasks). misconduct. this researchhas only begun. Vaughan 1992.cognition. unpublishedobservations). As before.and disasterare defined only in retrospectwhen outcomes are known.Further. MISCONDUCT.processes.Whetheran incident or activity producingan unexpectednegative outcome is viewed as conformingor deviant.the theories and concepts found implicate the macro-microconnection in the systematic production of organizationaldeviance: Its origin is in the nested and dynamic interrelationshipbetween environment.DARK SIDE OF ORGANIZATIONS 283 ganization characteristics(structure. . occurringin and/orproducedby a formal organization. But indicative of a nascent field of study. activity.sociologists should investigatethe local notions of social settings:how accompetentperformancehonoredandused in particular torsjudge each others' decisions and formulatethem as "mistakes"(RM Emerson. andthat produces an unanticipatedsuboptimaloutcome. and how those understandings are produced and transformedinto misconduct or disaster (Star & Gerson 1987). AND DISASTER Recall that organizationaldeviance was defined as an event. 1999a). it will vary by level of analysis: Intentionalfudging at the individual level could be viewed as conformityat the grouplevel. and choice. I concentrateon both social form and origin. I use the levels of analysis and subcategories inductively derived from the searchfor general sociological principlesto organizethis discussion.that deviates from both formal organizationaldesign goals and normativestandardsor expectations. and these understandingsare historically contingent.eitherin the fact of its occurrenceor in its consequences. transformed mistakeat the organinto izational level and misconduct at the institutionallevel (S Zerilli 1998. and cognition/choice.

For a definition of mistake sufficiently broad to organize this discussion. Belli & Schuman 1996). thatproduceunexpected adverse outcomes with a contained social cost: e. organizationsusually are a barely visible backdrop.structures. a solid basis is there..and processes often unnamed. psychological consequences. to distinguish mistake from misconduct and disaster. The other complex treatment producedby structural is Freidson's (1970. no attempthas been made to integratethis work to build a sociology of mistake. However.for if mistakes are aggregated across all hospitals.g. we furtherspecify that definition to include acts of omission or commission by individualsor groups of individuals. mistakewas a small partof a multifacetedproject. I createbroadconceptualdefinitions of mistake.andmistreatment an agent of social control. which stresses the vioand lation of formaldesign goals and normativestandards expectations. . injury.with varying direct social cost to them. With the exception of Roth's (1991) helpful compendiumof researchand anecdote.Nonetheless. The definitionalproblemnoted above immediatelyappears. the regularegulatemedicine createdifferentialperformancestandards.acting in their organizationroles. Environment and Mistake The most complex treatmentof environmentand mistake analyzes wrongful conviction in the criminaljustice system.Theory and the theoreticaldebatesfound in developed areasof study are absent.One common pursuitis groundedtypologies thatoften go beyond descriptionto explanation (Singer 1978. 1975) analysis of how the various fragmentedbodies that Thus. Huff et al (1996) first show the relation between institutionalizedculturalbeliefs andmistake. mistake in a hospital may harman individualactor or small numberof actors. misconduct. tangible social harm: loss of life. MISTAKE The sociology of mistakeis in its infancy. Bosk 1979. For by many.Then. The few sociologists who have studied it have investigated mistake in nonroutinework in organizationswhen it has direct. we use the definitionof organizationaldeviance above. propertydamage. dividing the research into these types based on how scholarsrefer to their own work and a few common characteristicsof the phenomenathat are evident in the literature.thenhow mistakeis conditions of the system.A point on which all agree is thatmistakeis systematicallyproducedas a partof the social organization of work. their organizational-technical system origins and extensive harmfulsocial consequences might well qualify as disasters.284 VAUGHAN Therefore. and disaster below.to preserveopenness of meaning. their environments.

terrorists)have the official responsibilityof bringing about the act of individual confession.285 DARKSIDEOFORGANIZATIONS origin of mistake.since mistakes-in-progress elusive.individuals. Research shows more about the processes of power and hierarchy than its structure. The only other aspect of structurespecifically linked to mistake is complexity (Roth 1991).so thatbuildings are constructedby an accumulation of blunders. therapists. Not so are surprisingly. Dwyer (1991) found that work site control by the dominant group (managers. McKee & Black (1992) found economic concerns of hospitals caused cutbacks in staff. more is known abouthow the powerful respondto mistakethanto the patternsof interactionthatproduce it. profits.. associated with routine nonconformityin the last section.g. One is by analyzing some aspect of the competitive environmentwithout showing how it affects mistake at the micro-level: e. whereby professionals in organizations (priests. Competitive pressureshave been explored two ways. politics. Landau& Chisholm (1995) arguethat norms of efficiency drive industries and organizations.Zerilli (1998) analyzed ceremonialconfession as a form of institutionalized interaction. OrganizationCharacteristicsand Mistake STRUCTURE Little is known about the relationship between structureand mistake.not norms of effectiveness.by its failure. Guillemin (1994) notes the rapiddiffusion of untestedhospital-basedtechnologies in the medical market place duringperiods of institutionalexpansion. and conflicts of interest. The ritualistic tactics of confessional-makingare such that they may elicit mistaken accounts from individuals. Light (1972) showed that the explicit therapeuticfailure and possible error PROCESSES . so that junior doctors made mistakes because of inadequatesupervision. and funding). injuries. have not been studied. unions) and the methods of control are related to the incidence of workplacemistakes. Singer (1978) shows that errorsare quasi-institutionalizedbecause they are literally profitable for organizationsto commit. In building construction. with misdiagnosis and harmfuloutcome the institutionalresponse. Guillemin & Holmstrom (1986) found the institutionalgoal of keeping beds filled was one factorin latent experimentsin newbornintensive care units that went on without scientific protocol.and death..Reimer (1976) found that division of labor among subcontractors createsirreversibleerrors. The second is to identify competition for scarce resourcesbut instead analyze the micro-processesthrough which it materializesin the workplace:e. The workings of interorganizationalpower. most researchreduces environmentto a single dimension: competition for scarce resources (income. Howtory environment. deadlines and demands for productivity produce mistake (Reimer 1976). police. in building construction.g.becomes a structural ever.

social worker errorin diagnosing abused childrenwas explained by high observationcosts. Choice. suppress organization. TASKS Cognition.cognition and mistake are the turf of cognitive psychologists (see Rasmussen 1986. Leibel 1991. resulting in professional misdiagnosis. and task complexity were first noted by Fox (1959). Mulkay & Gilbert 1982. Millman 1977. for cultureis not named nor are the theoreticalimplications drawnout. Bosk 1979).protectingthe statusof profession.Contemplating the string of medical misdiagnoses that later took her life. yet its richness is still to be mined. Jackall 1988. This researchreveals formal and informal patternsthat are institutionalized. The third illuminateshow social and culturalconditions affect cognition and choice.Reimer (1976) links uncertaintyto the transitionalnatureof the buildingconstructionwork site.and individuals(Millman 1977. For example. Paget (1988:58) called medical work "error-ridden activity" because it is inexact. Mizrahi 1984). The greatest concentration of research on mistake is fine-grained and ethnographicanalysis of tasks. How uncertainty.and lack of time to review and revise as more information became available (Munro 1996). Bosk 1979. Repeatedly.who institutionalizeda review process that preservedthe professional project and protectedthe practitionerfromblame. Historymatters:Tasks thatare structurally temporallyseparatecan generatemistake throughaccumulation. gradualaccumulationof information. risk.Threelines of inquiryhave developed. risk. until harm is irreversible. so that processes that "neutralize" people in the workplace see them as routine and nonremarkable(Freidson 1975. Uncertainty. Paget (1993) noted how action unfolds in time. so that one mistake is compoundedby others. but sociologists have begun a cognitive sociology of mistake. The second examines the construction of accounts by professionals when confronted with possible evidence of their own error(Stelling & Bucher 1973. and task complexity vary and how they intersect with cognition and mistake remainto be studied. Rosenthal et al 1999). and practicedon the humanbody.throwing workplace culture into broadrelief. Reason 1990). Edmondson 1996. The first explores and "normalize"certainkinds of mistake. uncertain. Research using ethnomethodology and labeling theory exposes how institutionalizedprofessional culturalbeliefs affect cognition and mistake:Rosenhan's (1973) researchad- .286 VAUGHAN inherentin patient suicide threatenedpsychiatrists. and Mistake Traditionally.researchshows how organizational hierarchies mistakes and deny responsibility.Hughes' (1951) prescience about mistake and risk as indigenous to the task are verified in this empiricalwork. Huff et al (1996) locate risk anduncertainty in the complexity of offense and offender.

too.""occupational crime."or "abuseoftrust"?The resultis researchin many conceptualtraditions. police.The possibilities for futureresearchand theorizing are unlimited. regardless of size. many offenses originate in formal and complex organizations.thrivingenterprise. MISCONDUCT In contrastto mistake.the sociology of misconductin andby organizationsis a mature."a study of basic conceptual equipmentwith which judges.This development is institutionalized. These patternsbegin to suggest the social origin and form of to mistake.and cognition and choice all need study. Environment." "abuse of power. or kind of organization. the lead news story.But deeperinquiryhas producedharderquestions.Sociologists have focused on mistakes that directly harm individuals or property.The most valuable contribution the darkside of organizationsis the field research that shows responses to mistakes as they are occurring in the work setting: These insights into ethnocognitionand organizationaldeviance have not been possible in researchon misconduct and disaster." "organizationaldeviance.productsthat cannotbe assembled or fail and have to be recalled. and probationworkers organize their daily work. creating theories that pointedly gave explanatoryprimacyto the organizationallocus of wrongdoing. Beginning in the mid.1960s. scholarstrainedin the sociology of organizationsand of deviance and social control began combining the two. We need to know more about the human/technologyinterface and mistake.organization characteristics.287 DARKSIDEOFORGANIZATIONS mitting sane pseudo-patientsinto mental hospitals undetected." "corporatecrime. But a mechanism for coherence has emerged. omitting mistakes that are just as relentless but more mundane:mistakes in reports. complexity. Employees and organizations devote enormous resources to preventincidents of routinenonconformityfrom being publicly defined as mistake. This. researchhas been imbalanced. is worthy of research:The social organizationof clean-up work also has social costs that eventually are paid by the public. Key conceptualdebates remainunsettled:Is it "white-collarcrime. Summary The strengthof this work lies in its indicationof common patterns within each level of analysis. only a few have been tapped and none systematically. Sudnow's (1965) "normalcrimes. Despite differences. Of the sociological theories and concepts describedin the previous section that explain routine nonconformity. hiring. and about how the individual experience of mistake and emotion affects the propensity for subsequentmistakes.Pfohl's (1978) analysis of official reviews of inmates in institutionsfor the criminallyinsane by teams of mental healthprofessionals.Also.evident in (a) the systematic inclusion of or- . lawyers. the pornographicnovel stitched into the binding of an academic monograph. task.

powerful offenders can resist grass roots efforts at social controlbecause they constructpublic accounts that legitimate their actions and because individualsand organizations are dependentupon them for goods and services (see Edelman & Suchman 1997). (b) the shifting premises of theoretical debate (initially.The role of the regulatoryenvironmentin the social origin of misconduct is uncontested:Although effective in specific cases. the relationshipbetween individualactorsand organizationalactors. and fluctuatingeconomic conditions.(c) regulardiscussionof organizations textbooks.organizedcrime is excluded from this discussion.Power enters into this equationin the ability of regulatedorganizationsto affect both the structureof the regulatory environment and specific regulatory outcomes. most recently. marginaland failing firms will be more likely to offend. contests of . Because deviation from both formal design goals and normativestandardsor expectations is a requisite of routine nonconformity. or administrativeregulations on behalf of organizationgoals. we use the definition of organizationaldeviance above. Further. To distinguish misconductfrom mistake and disaster.288 VAUGHAN ganizations in causal theories. The results are contradictoryand weak (see Jamieson 1994. a prediction consistent with Merton's aboutthe working class and street crime but inconsistent with research on crimes of the powerful and the influence of power on the regulatoryenterprise. and (d) research using ever more sophisticated organizational analysis and theory. As with mistake. thereby underminingthe efficacy of deterrence.the relative explanatorypower of rationalchoice theory and orin ganizationtheory).when incidentsareaggregated.marketstructure. The relationshipbetween the competitive environmentand organizational misconduct is equally well established but still puzzling. The most frequent hypothesis. so social cost may be containedor diffused. is that organizationswith blocked access to opportunitiesfor economic success will violate: Thus. However.articles.Power-powerful organizationoffenders. qualitativeresearchsuggests this use of the Mertonianhypothesis is overly restrictive. and chapters. Environment and Misconduct The competitive and regulatory environments have received extensive research scrutiny.we furtherspecify that definition to include acts of omission or commission by individuals or groups of individuals acting in their organizationalroles who violate internalrules. For a definitionof organizational misconductsufficientlybroadto organize this review. laws. harmis extensive and social cost is high. Geis & Salinger 1998). derived from Mertonianstraintheory ([1938]1968). The extent of adverse consequencesand harmto the public will vary with the act. sources of regulatoryfailure are socially organizedand systematic across cases.Quantitativeresearchon corporateviolations separatesenvironmentinto complex industryvariables.

Organization Characteristics and Misconduct Theoristsuniformly hold that structures. and still others to keep from losing groundor droppingout of the competition altogether.not indicators of firm health.its role in the social origin of organizational misconducthas been located in manyplaces: in normativeambiguityfor white-collar offenses (Aubert 1952). the cultureof capitalism (Finney & Lesieur 1982). and industry cultures (Leonard & Weber 1970).processes.or its connectionto the actions of organizationsandtheirmembers. Michalowski & comprehending regulatory Kramer1987). . A new interest is network analysis linking organizationsand actors in networks of collusion that subvert the legitimate economy (Barlow 1993. suggesting that quantitative studies wrongly narrow competitive pressures to economic success and profits.(b) the means for carryingout violations.and business shows (a) competition not just for profits and economic success but for many kinds of scarce resources essential to survival. Qualitativeresearchsuggests revising the Mertonianhypothesis as follows: Given the universality of competitive pressures.g. marginal caught. LaFollette 1992. agency enforcementactions on corporations.governments. and tasks are opportunity structures for misconduct because they provide (a) normative support for misconduct.This possibility is reinforcedby Delaney (1992). paving the way for future studies of power struggles and the political environment. occupational cultures (Green 1997). Simon 1994. Cultureis an aspect of the environmentscholars hold important. Analysis of misconductby nonprofits. the cultural environmenttypically is noted as importantwithout precision in conceptual definition. Lackor failing business firms simply may be more likely to get ing power. its empiricalreferent. This researchbuilds upon theories of interorganizationalrelations. culturally approved success goals and anomie (Merton[1938] 1968). Baker & Faulkner1993. who debunksfirm financial figures as the productof strategiesand power relations. These findings call into questionthe datasources for quantitativeresearch:typically. and (c) concealmentthat minimizes detection and sanctioning. Grabowsky 1989.othersmay compete for upwardmobility.. The powerful may compete for scarce resources to sustainrank. Calavitaet al 1997).Often referred to as normative environment.DARKSIDEOFORGANIZATIONS 289 the power-is as essential to understanding competitive environmentas it is to failure (Snider & Pearce 1995. However. a tiered system of meaning in an industry(Denzin 1977). cultures of risk taking (Black et al 1995). all organizations may be structurallyinduced to violation. Fleischer et al 1992. and (b) misconduct not limited to marginaland failing organizations (e. regardlessof rankingin the organizational stratificationsystem. Zuckerman1977).

g. CEO background.organizationcultureis treatedunidimensionally:It provides normative support for wrongdoing. downsizing (Friedrichs1997). PROCESSES . this unidimensionalview of cultureis challenged by theory thatemphasizesculturalcomplexity andvariation(Vaughan 1983. Quinney 1963. STRUCTURE A centralcontributionhas been scholarshipestablishinghow the competitive. and escalating commitment(Ermann& Rabe 1995). Kramet al 1989). if pursued. regulatory. and variation in subunitvulnerability (Frey 1994. and more dramatically. Traditionally. Daly 1989. Now. are Top administrators responsible for performancepressures indirectly (by out-of-reachgoals or not providingsufficient resourcesnecessary establishing to attain goals) and directly (by actively conceiving and enacting violative behavior). leading individualsto engage in misconducton behalf of organization goals. and cultural environments translate into organizational processes.Also.. and responsibility (Stone 1975.g. Wheeler & Rothman 1982. the diffusion of knowledge.promises insights thattranscend types of organizationsand types of violations. Research shows how environmentalstrain materializes in performancepressures that affect individual actions and the development of internalculturethat supportsachieving the organization'sgoals illegitimately. Shulman 1997).number of componentparts. Punch 1985. or administrativeregulations. However. Unexamined is complexity: centralization. misconduct. Vaughan 1983. Bensman & Gerver 1963.by ethnographiesthat show conflicting culturesdue to differences in professions. much scholarshipaffirmsthatan individualoffender's position in a structure explicates the social organization of misconduct (e. and layers of hierarchyand how they vary in relation to misconduct. Shapiro 1990.ethnographiesthat show socialization and onthe-job trainingfor techniquesof rule violating behaviorused to execute routine tasks (e. Morrill 1995). Simpson& Koper 1997). geographicdispersion. This view is sustained by Sutherland's (1949) learning theory.and informalcliques (Jackall 1988. This latterresearchagenda. laws.decentralization.decentralizedmanagement..and productdominantstrategies(Simpson & Koper 1997). Shover & Bryant 1993). Clinard 1983. and researchshowing thatwillingness to use illegitimate means on the organization'sbehalf is sealed by a reinforcingsystem of rewardsandpunishments.290 VAUGHAN Scholars have studied the role of formal structurein "crimecoercive systems" and "crime facilitative systems" (Needleman & Needleman 1979). Chambliss 1996. scholars are studying the impact of administrativedecisions on misconduct in more detail: managerialsuccession. rank. which explains that employees are socialized into membershipin a groupwhere the normsfavorviolation of internalrules. Vaz 1979. Coleman& Ramos 1998).

Sociologists have investigated accountingtactics (Passas 1996). Most significant for understanding the dark side of organizations is recent research integrating these three levels of analysis. Some are studying the diffusion of these innovations (Calavitaet al 1997. decisions to violate. Abolafia 1996. and tasks). Simpson et al 1998. Choice. if benefits outweigh the costs. in organizational misconduct. and Misconduct Traditionally. Vaughan 1998).In what is possibly the most importantnew development. covert debate about what the appropriatelevel of analysis is (Vaughan 1999b). Ermann& Rabe 1997. disappearinginto costs.and culture with interpretation. Summary Research and theory suggest that. a formof rationalchoice theory:Confrontedwith blocked access to legitimate means to organizationgoals. organizational deviance. This directionhas potential. and espionage (Hagan 1997). the social form of organizationalmisconduct includes aspects of environment. decision-makerswill calculate the costs and benefits of using illegitimate means. meaning. and the probabilityof punishment. Zey 1993). Lofquist 1997.However.the rational choice assumptionsbehind the amoral calculatorhypothesis are being examined within the context of social psychological and organizationtheory (Coleman 1987. actorswill violate (Kagan& Scholtz 1984).organization characteristics(structure.DARKSIDEOFORGANIZATIONS 291 TASKS Cressey (1953) argued that the skills necessary to misconduct are simply the skills necessary to do thejob in the first place. This researchpromises advances connecting institutions. the concepts of risk and uncertaintywere associatedwith tasks. and individualaction. The reason the definitional is- . The transformationof routine tasks by these technologies contributes to misconduct because they are employed in new. organizations.Friedrichs(1996) notes the conspicuous rise of "finance crime"-large-scale illegality that occurs in finance and financial institutions-and "technocrime"-use of new technologies such as computers. Simpson 1998.processes. and the probablereaction of regulators. complex transactionsystems thatare difficult to monitorandcontrol(Vaughan 1982b). Whereasin the sociology of mistake. these concepts usually are not directly studied. and accounting technologies. like mistake.facsimile machines. etc) is an unrecognized. manipulationof marketsand financial capitalism(Levi 1981. and cognition and is systematically produced by these three in combination. electronic surveillance. Behind the unsettled debates about the appropriate conceptual definition (occupationalcrime. Baker & Faulkner1997). Cognition. they are located in cognition. but only recently has the relationshipbetween tasks and misconductbeen studied. decisions to violate have been explained by the amoral calculatorhypothesis. Reed & Yeager 1996. benefits.

or some combination(Turner& Pidgeon 1997:19). the accident would need to be large-scale. Mistake and misconduct often occur in the prehistoryof accidents and disasters. as this review shows. Historiresponse. unusually costly. complex. all levels of analysis apply. Turner (1976.the centralquestion still is: When do they produce conduct and when do they producemisconduct? DISASTER The study of disasters has deeper roots than its topicality suggests. Then in 1984. For an accident to be defined as a disaster. technologies. unusually unexpected. further .To be sufficiently broad to encompass both accidents and disasters. 1978) analyzed 84 accidents and disasters across industries. and tightly ordered such thatthe failureof a partcan resystems in which partsare interdependent sult in the failure of the system. Although recent researchhas begun to inquireinto variation. The existence of this covert debate accentuatesthe legitimacy the of each level of analysis. unusually public.Often. Disaster is a type of routine nonconformitythat significantly departs from normative experience time andplace.the lattertwo distinguished from mistake and misconduct by the social cost and quality of surprise. It is a physical. often possessing a dramaticqualitythatdamagesthe fabric curring of social life. we start with the previous definition of organizational deviance.the benefit of the disagreementis thateach conceptualtraditionhas producedknowledge at all levels. readyto be integrated. Perrow's Normal Accidents and Short's "The Social Fabricat Risk" chargedthe intellectualcommunityand changed the research agenda. Landau(1969) located the social origin in large. finding that social. scholarshave identified aspects of environment and organizations as "criminogenic. interestlay in the impactof disasterand the organizational quiryinto the social origin examinedhumanfactorsin accidents.Unnoticed in that arraywas prescientwork locating the cause of accidents in power and the structuraland cognitive limits embedded in organizationsand technologies. Erikson (1976) showed the corporatepower and cost/safety trade-offs behind the Buffalo Creek dam collapse. and administrativearrangementssystematically produced disasters." and showing how aspects of environment.organizations. cultural. The result is a body of work showing how institutions. incally.and emotionalevent infor a particular social loss. thus substantiating importanceof merging them.organizationcharacteristics. technical. Moreover.292 VAUGHAN sue remainsprovocative and lively is that. and cognitive practices contributeto accidents and disasters. cognitive practice normally associated with the bright side of organizations are systematically relatedto the darkside. This review encompasses disaster studies: research its authorsidentify as investigations into the origin of accidents and disasters.

and/orregulation. it is usually a dependentvariable. strategies.and cultural environments. In mistakeandmisconduct. But turbulencemay also matter.culturalenvironmentappearsas an independent variable. and knowledge for an effective response to crisis (Clarke 1999).in disaster studies. 1994) presents convincing evidence that power and politics contributeto accidents. Indeed.investigating. MacKenzie 1990). Equally interesting is how the political environmentpermeates competitive. costs of regulation. Environment and Disaster Research firmly links the social origin to political. Historic environmentalchange has multiplied the potential for accidents and disasters. unavailable. regulatory. Producersand users of science and technology have the power to affect what is culturallydefined as an acceptablerisk so that debates about hazardsgo on in an environment that is compatiblewith their interests. technology. but both demonstratethatit contributesto accidentswhen subjectto power-dependence relationsthatundermine(a) effective monitoring.or obsolete.Competitionand scarcity set the stage for accidents when they lead to cost/safety trade-offs. How a high velocity environmentcan be identi- . Eisenhardt(1993) argues that accidents are associated with "highvelocity environments"markedby rapid and discontinuous change in demand. Organizationsproducing and using risky technologies areboth victims of political shifts in regulatorypolicy (Tompkins 1990) and powerful actors that shape it (Kroll-Smith& Couch 1990.DARKSIDEOFORGANIZATIONS 293 specifying it as organizational-technicalsystems failures that include acts of omission or commission by individualsor groups of individualsacting in their organizationroles."official plans to respondto accidentsthatare culturallyreassuringbut lack appropriate resources. so safety tends to be the goal displaced. so they develop into accidents and disasters. and culturalenvironments.most notably as large technical systems developed with the capacity for normal accidents (Mayntz & Hughes 1988).and sanctioning of producersand users and (b) effective response to incidents. competitive. Verma & Marcus 1995). with outcomes that either in the fact of their occurrenceor consequences are unexpected. regulatory. such that informationis often inaccurate. The regulatoryenvironmentgets more attention from sociolegal scholars than disaster specialists. competitors.Culturallegitimacy also is gained by displacing blame from organizations and institutions to operator error (Perrow 1984) and by the creationof "fantasydocuments. Hughes 1983. the history of technology is a history of conflicts of interest that resulted in suboptimaldecisions about technology and technological products (Cowan 1983. and the safety record (Marcuset al 1993. Quantitativeresearchnow is looking at producers' ability to afford safety. Cable et al 1998). adverse. and of high social impact and cost regardless of numberof lives and amountof propertylost.Sagan (1993.

environmental strain translates into internal processes thatare associated with accidents:conflicting goals.informalorganization. and culturalbeliefs about hazards that together preventedinterventionthat might have staved off harmfuloutcomes. Marcus & Nichols 1999). misconduct. Osborn & Jackson 1988. Turner(1978:179) found that errorsat the top have greateraccident potential because errorscompound as they move down throughthe hierarchy. an outcome of types of interactionof system parts(complex or linear) and types of coupling (tight or loose). Braithwaite1985.First.294 VAUGHAN fled and its effects on accidentsand disastersdefinedboth remainto be worked out. Bella 1987. For example. institutionalarrangements. but one known result is the failureof formalagents of social controlto effectively monitor rapid changes in scientific and technical knowledge (Jasanoff 1986).In addition. favoring technologies that reinforce the power structureand result in unwarranted claims of operatorerror. Executive goals and resource allocations can trickle down. group and subunit conflicts can displace goals. impeding the efforts of people doing the risky work (Perrow 1984. Shrivastava 1987. analogous to mistake. andculturalunderstandings (Perrow 1984. The conjunctionof competition. OrganizationCharacteristicsand Disaster Perrow (1984) established complexity as a core concept in disaster studies: Accidents are normal in complex systems. Second. Shrivastava 1987. underminingknowledge about potential hazards (LaPorte 1982. history. resulting in less-than-optimaltechnology (Thomas 1994). Gusterson 1996. Freudenberg1992. Strikingand new are the discoveries about rules. STRUCTURE PROCESSES Two patternsrepeat in disasterresearch. Also salient are the effects of organizationcomplexity on information flows. Vaughan 1997. Bogard 1989.Routine nonconformityhas multiple logics: Rule violations occur because of mistake. Turner& Pidgeon 1997).reward systems that reinforce productivity and underminesafety. typified by rule violations. Snook (1996) identified "practicaldrift":an incremental uncoupling of practice from writtenproceduresdesigned to handlethe . Clarke 1999. history matters:Turner(1978) found that man-made disasters had long incubationperiods.Perrow(1983) explains why military and top industrialmanagementare indifferentto good human-factorsdesign. and rule violations suggests that only poorly run organizations have violations and accidents.but well-runorganizationsalso exhibitthis pattern. Perin 1998). Sagan 1993. analogous to or- ganizational misconduct. Clarke 1989. deadlines. and decline of resource slack. Hierarchy and power are profoundly implicated in accidents. Vaughan 1996. escalating commitment. discrepantevents that accumulatedunnoticed. performancepressures.

As the technical system varies. Choice. Both intelligent technology and intelligenthumanshave limited ability to cope with inconceivable occurrences. Many aspects of social context are salient for cognition (Short & Clarke 1992. and disastersunfold are rare. so does the ability of operators. TASKS One strandof task-orientedresearch locates risk and uncertaintyin the technology because it is complex. acUnfortunatelybut understandably. not modes of failure (Schulman 1993). and Disaster Scholarshipon cognition. organizations. Weick & Roberts 1993. This gap between writtenrules and action impairseffective response in a crisis. The guiding assumption is that all technologies have interpretive flexibility. These . This new work discloses "the operationalrealities of risk handling"(Carroll& Perin 1995:22). studies of ethnocognitionas incidents. How individuals constructrisk and uncertaintyand how they produce technical knowledge are the central problematics. Weick (1993).managers.expertsdisagreeaboutfacts both before accidentsand after. Roberts 1993. resources. not complex interactiveones (Meshkati 1991). in a stunning analysis of why firefighters died in the 1949 Mann Gulch disaster. Weick et al 1999). Recent researchhas opened up the unexplored world of the people who do the risky work. training is for single failures. showing how the technology mystifies and how organizationaland institutional factors affect the work process.such as "heedfulinterrelating" and "collective mind"(Weick & Roberts 1993. cidents.promoting"thereasonablechoice of disaster"(Lanir 1989). while those who perishedconformedto the organizationmandatealways to carrytheir tools. choice. contributingto failures. and institutionsto controlit (Perrow 1984). uncertain. minor flaws and errors are accepted due to deadlines (Pate-Corell 1990). conformingto rules also can contributeto accidents and disasters. revealed that the few who survived droppedtheir heavy tools and ran. and commitments to hardware produce questionable fixes rather than change (Starbuck& Milliken 1988a). when tight coupling is called for. resulting in disputed knowledge claims about the technical world (Pinch 1991).The productivealternativehas been participantobservationof people doing risky work safely. Cushing 1994. and the social origin of accidents and disasters locates risk and uncertaintyin interpretation sensemakingabouttechand nologies. and therefore inherently risky:The technology makes tasks difficult andaccidentslikely. Weick et al 1999). people are taught modes of success. Cognition. risk-handlingresources are inadequate(Perin 1995). which has yielded concepts relatedto cognition and effective teamwork. schedules. Ironically.DARKSIDEOFORGANIZATIONS 295 worst-case condition when subunits are tightly coupled. Social context is Risk and uncertaintyof a situationor object vary with the social all-important: location of the actor.

Third. Summary The achievementsof disasterstudies are several. rules.researchinto causes affirmsthat environment. Starbuck& Milliken 1988b. Two studies show how institutional and organizationalculturalbeliefs promote failures of foresight by affecting the interpretationof information:A "disqualificationheuristic" leads decision-makers to neglect information that contradicts their conviction that a technical system is safe (Clarke 1993). directing preventive strategiestowardthe institutions. enabling people to conform to institutional and organizational mandateseven when personallyobjectingto a line of action (Vaughan 1996).296 VAUGHAN studies of process show how betterorganizationscatchmistakesby employing advancedcollective cognition before things snowball into disasters. and arguesthat complex systems will inevitably fail. most scholarship uses organization theory. High Reliability Theory (HRT) studies safe systems. decision-makers saw "ill-structured" problems that afterwards became "well-structured" decision-makersand investigatorsalike when the to adverse outcome alteredtheirworld view. Hotly debatedbetween NAT and HRT are the effects of centralization. limiting what investigations can learn from failures (Gephart 1984. it is typical to worryabout"TheGreatDivide":Normal Accident Theory (NAT) studies failures. When assessing disasterstudies. and cognitionareall implicated.but research has not . in contrastto mistakeandmisconduct.organizations.fomentingan incipientrevolution. many scholars have respondedto the high demandfor sociological insights from organizationsthat produce and use risky technologies.like mistake and misconduct. industry.11). and culturecreatefailures. showing the socially organized origins of accidents and disasters in comprehensive complexity. They consult and lecture. and redundancy.Second. emphasizes process. tight and loose coupling. to accidents. and professional culturalbeliefs frame facts. researchhas new insights abouthow power. 1998).First.and systems that contributeto accidents. emphasizes structure.These differentorientationsshow in detailhow the same aspectsof organizationsthat contributeto the brightside also contributeto the darkside. After accidents.many practitioners are beginning to look beyond operator error.Finally. government. Not only do "fantasydocuments"affect societal culturalbeliefs about risky technologies. Turner(1976) found culturalbeliefs contributedto "failuresof forePrior sight:"a historyof discrepantevents thatwere ignoredor misinterpreted. knowledge aboutthe effect of cultureon choice in the preconditions of accidents and disasters is accumulating(see Turner& Pidgeon 1997: Ch.decentralization. Carroll 1995. However. and the "normalizationof deviance" can neutralize signals of danger. they also persuade employees of the safety of their own enterprise (Clarke & Perrow 1996).Miraculously. and argues for effective prevention.organizations.

ratherthan successes or failures (Carroll & Perin 1995). relevantaspects of competitiveandregulatoryenvironments. informationrichness introduces inefficiency. scholars studying high reliability organizationscompare failures and safe organizations(Weick 1990. The division of labor. regardless of orientation.However.teams have multiple points of view that enhance safety. 1993. and disaster as three types. and especially the characteristics the task.g. Roberts& Libuser 1993). 1993.and ethnocognition in tasks of people doing risky work.which has HRT mainly exploringprocess andNAT focusing more on structure.more carefully specifies social context: the unit of analysis (system. so the generalizabilityof all findings is unclear.can create them. subunit.process. the complexity and coupling of the technology.its complexity. examiningthe link between environmentalfactors. but as they become a cohesive group they share assumptions. Vaughan 1996).has been most problematic for developing broad causal principles upon which to base strategies for control. The debate about whetherredundancycontributesto accidents (NAT) or prevents them (HRT) may be a function of differentdisciplinarytrainingand research styles that result in two different definitions of redundancy:structural redundancy versus process redundancy. organization. and integrativeefforts comparevariationsin structure. how variationin structureaffects the possibility of collective cognition and its success or failure as a preventive.and accidents (Schulman 1993..Neiof ther "side"is doing this. scholars studying failure link structureto process and cognition (Clarke 1992. too little produces inaccuracy. scholarsstudyinga risky industry.work group). much currentscholarshipmediatesthese polarities. performance.then of mistake. perhaps indicatingthe growing maturityof the field: Debates illuminate similarities as well as differences (LaPorte 1994).so the "requisitevariety"importantto safety is lost. organizationcharacteristics.presence or absence of resourceslack. The greatest advances may come when individualprojectsstrive to merge levels of analysis. THE DARK SIDE THEORIZING This essay is a Simmelian examinationof the darkside of organizations.often used to prevent errors. misconduct. e. we need to know. Studying the dark side of organizations exposes the operational inadequacy of society's institutional . Weick (1987) has skillfully shown how the same processes that produce the bright side (safety) can produce the dark side (accidents and disasters): Training. For both theoretical and practicalreasons. Additional progress is possible if future research.first exploring the social form and origin of routine nonconformity. Marcus 1995).DARKSIDEOFORGANIZATIONS 297 sortedout underwhat conditionsthey will producesafety or increasethe probability of accidents.

advancesin one areacan provide theoreticalinsight and researchdirectionfor another. common origin is located in connections between these levels of analysis that systematicallyproduce orand ganizationaldeviance. Only a few sociological constructslinkedto the social originof routinenonconformity have been used in researchon mistake. misconduct.but also the decontextualized.. misconduct. but systematicproductsof complex structuresand processes.and socio-cognition.g.error-amplifying traps. I have analyzed four literaturesin orderto make key ideas of each available to specialists in the others. Using an inductive case comparison method that identifies social form and origin by attendingto both analogies and differences. practicaldrift.Also.g. misconduct. and second. and disaster are not anomalous events. mistake is "young")and researchaccess and interestratherthan differences of form.core set) could be extended to study the three types. and disaster could guide the study of routine nonconformity or be swapped between types (e. anomie. the trickle-downeffect.theories and concepts associated with mistake. this review affirmsthatpolicy for preventive strategiesmust go beyond individuals to institutional and organizationalfactors that shape individual cognition and action.system complexity. goal displacement. To lay a foundationfor the darkside of organizationsas an integratedfield of study. two centralanalogies stand out: first. It increases our understanding social structure. Debunking myths of operatorerrorand individual wrongdoing. Reciprocally. routinenonconformity and the threetypes have a common form thatincludes aspects of environment.liabilities of newness.power andpolitics areemphasizedin misconductanddisaster . the embeddednessperspective.298 VAUGHAN bases. Analogous structures processes are easy to identify because patternsclearlyrepeatacross routinenonconformityand its types.the disqualification heuristic. means-endsorientationof rationalchoice theory. requisitevariety). so organizationsproduceunanticipated negative consequences that deviate from formal design goals and normativestandards and expectations.. showing that behavior is rationalwithin situationalcontexts and that social context can decouple rational choice from outcomes.g.tasks have priority in researchon mistake and disasterbut are neglected in misconduct. nonconformity.showing that routine of mistake. But some differences can be singled out as directionsfor futureresearch. Consequently. unTheoriesand concepts (e. organizations. and disaster. the regulatoryenvironmenthas been extensively studied in organizational misconductbut virtuallyignoredin mistake and disaster. But drawing conclusions about differences within these categories is more tentative because the absence of some characteristicmay reflect substantivetopic differences in researchmaturity(e.trainedincapacity..For example.it challenges not only Weberiannotions of rationality.environmental decision certainty.science and technology are centralin disasterstudies but unexploredin mistake andmisconduct.

and cultural sociology.organizational. ACKNOWLEDGMENTS to My thanksto Sal Zerilli for conversationsfundamental my thinkingand for the literatureon which this essay is based. . for researchon markets.What is the social organizationof this clean-up work. Klein 1998. Reason 1997. there is the unresolvedquestionof when conditionscombine to producethe brightside and when they culminate in the darkside of organizations. the variablerelationshipbetween structuresandprocesses identified and the frequencyandprobabilityof unexpectedsuboptimaloutcomes is an important question. 1997. rationalchoice sociology. misconduct. and an anonymous reviewer for thoughtfulcomments on the penultimate draft. Special thanksto the many kind colleagues who respondedto requests by sending bundles of their most recent work.Lee Clarke.g. and choice.and socio-cognitive elements combine to producevariationin individualchoice and action. and how is routine nonconformityconvertedto mistake. This discussion has been cast as a dark side/brightside dichotomy to emphasize variation from the usual treatmentof organizationalbehavior. Gaba et al 1987. Moe 1991. investigating changes in law. we might hypothesize that in virtually all socially organized settings. examining the disjunctionbetween rationalchoice and outcomes.institutionalizedinequality. Necessarily.cognition.studies of ethnocognitionare necessary. or disaster. and what is the effect on social structure? Third. 1976. Building the dark side of organizations as an integrated field also would include some topics thatare challenging methodologicallybecause the subject matteris. technology. understandings aboutwhat is culpable and what is tolerable.or disaster?Finally. organizationaldeviance.DARKSIDEOFORGANIZATIONS 299 but but minimally addressedin mistake. Karl Weick. Allison & Zelikow 1999). Much can also be learned fromthe strongtheoreticalinterdisciplinary work on the darkside of organizations (e.and social definitions of what is normativeand deviant at a particularhistoric moment. Jervis 1970. However. Answers are likely to be forthcomingwhen scholars examine how the conjunction of environmental. economic sociology. Risk and uncertaintyare fundamental are conceptuallyunderdevelopedin all three types. competition. this agenda would add to and draw from historical sociology. routinenonconformityis met with efforts to keep it frombecoming publicly identified as mistake. to RobertGibbons help assembling and the participantsin his MIT OrganizationalBehavior seminarfor their response to an early outline of these ideas. and John Carroll. First.who gets to decide what is and what is not.. Second. how do understandingsdevelop about what is an incident of routine nonconformity. andknowledge thataffect the incidence of unanticipated adverseoutcomes. for analyzing how culturemediates environment.organization characteristics. Lerner 1986. misconduct.and social costs. Sagan 1994. afterall. science.

CO: Westview Bosk C.Chicago: Univ. WhenThings Go Wrong. Policy Res. 1983. Pontell H.CT: JAI Calavita K. 1989. Sci.Boulder.CT: JAI Barlow HD. 58: 837-60 Baker W. Faulkner R. Chambliss D. Uncertainties. 17:1-25 Coleman JS. In Research in Social Problems andPublic Policy. 1985. The social organization of conspiracy. 1998. White-collarcrime and social structure. Anheier H. In Organizations. 1994. Changing Order. MA: HarvardUniv. 1997. Soc. 1999.MA: HarvardUniv.Ind. Organizational learning activities in high-hazardindustries. 1998. Crisis Q. Phila. Albany: State Univ. Beverly Hills. Sociol. J. 1996. Am. To Punish or Persuade. Chicago Press. L Clarke. 1979. Acceptable Risk? Berkeley: Univ. Organizations and systematic distortionof information. 1997. Stud. Context dependency and risk decision making. Stud. The disqualificationheuristic. Youn. Vol.AnnualReviews. The Bhopal Tragedy. Sociol. Law Soc. The Hidden Society. Subcultures and deviant behavior in the organizational context. CT: JAI ClarkeL. 1983. Olsen JP. Towardan integratedtheory of white-collar crime. 1998. Chicago: Univ. Research in the Sociology of Organizations. Sociol. T Youn. 58:263-71 Aubert V. Press Carroll J. 19:6-19 Collins HM. MissionImprobable. Garbage can model of organizationalchoice. Rev. Hastings D. eds. 1995. Prosaic organizational failures. eds. Rep. 1998. 1994. JF Short. Debating Corporate Crime. ed. Press Allison G. Essence of Decision.300 VAUGHAN Visit the Annual Reviews home page at http://www. 1993. Sociol. 3-34 Collins HM. Manage. Rev. ed. 1996. ed. 1981. TheBusiness of Talk. 1963. Presented at Int. 35:183-94 Bella D. ed. 1997. CorporateEthics and Crime. Chicago: Univ. LiteratureCited Abolafia M. The TEA set. Sociol. Crime and punishment in the factory. Q. Qual. 28: 588-98 Blau P. W Freudenberg. NSP95005.Prof. 1996. 5. MIT Cent. 1993. Chicago Press Braithwaite J. Networks Conf. 1992. 1989. 1974. Am. FaulknerR. Environ. Cincinnati. 'Foi por acaso': conceptualizing coincidence. Greenwich. 1955. Zelikow P. Ramos LL. Chicago Press Cable S. Chicago: Univ. Organizingand Managing for Safe Production. Chicago: Univ. Press Coleman JW. Admin. Boulder.Bristol. 2nd ed. 35:699-717 CarrollJ. CA: Sage Cohen MA. Calif. Behav.Berkeley: Univ. 1995. NJ: Bedminster BakerW.New York: Basic Books . Sci. See Bamberger & Sonnenstuhl 1998. 1999. 93:406-39 Coleman JW. W T Freudenberg. Press Clarke L. Perrow C. Greenwich. 1993. Incident reviews in highhazard industries. Rabe GA. GerverI. More WorkforMother. Chicago Press Clarke L. Schuman H. 39: 1040-56 ClinardM. Am. 1998. Press Becker HS. Lofquist WS. 1938. 1987. 1992. and Risk. The place of the 'core-set' in moder science. OH: Anderson Cohen MD. 1996. Forgive and Remember. 9:175-97 Carroll J. 4:165-86 Collins HM.org. Energy Environ. The Dynamics of Bureaucracy. 1972. Totowa. ThousandOaks. Making Markets. 19:423-30 Bensman J. Hist. Criminol. Cambridge. Sociol. Cambridge. Sci. Beyond Caring. Manufacturing Consent. 1965. Diffusion of fraud. 15. 1974. Q. Stamford.. Chicago: Univ. pp. Am. CO: Westview ClarkeL. pp. CA: Sage AubertV. 1979. Change 20:319-37 BarnardC. Boden D. The silenced majority. San Diego Bamberger PA. Sonnenstuhl WJ. 27-38. Big Money Crime.J. Calif. Shriver T.Am. The complexity of ignorance. Chicago Press Clarke L. PerinC. 1987. Vol. TheFunctions of the Executive. Power and the Structureof Society. 1952. March JG. J. Philadelphia:Univ. From fiddle factors to networks of collusion. New York: Longman. 113:360-70 Belli RE. NY Press Burawoy M. 2nd ed. In Research in Social Problems and Public Policy. Chicago Press Cowan RS. Issues Eng. Tillman R. Sci. UK: Polity Bogard W.

Experiment and illusion in reproductive medicine. 1988. Communications in organizations. Chicago: Univ. J. In press Huff CR. Suchman MC. Inst. Everything in Its Path. Press Gusterson H. 1992.New York: Plenum Edelman LB. Sagarin E. 1992. Economic action and social structure. P DiMaggio. Stanford. 1996. Political Crime. Chicago: Univ. The National Collegiate Athletic Association: A Study in Cartel Behavior.Annu. 1987. 1996. Introduction. Convicted But Innocent. In The New Institutionalism in OrganizationalAnalysis. Sociol. Annu. Fatal Words. Press Denzin N. New York: Free Press Dwyer T.DARK SIDE OF ORGANIZATIONS Cressey D. 1959. Greenwich. WaywardGovernance. Englewood Cliffs. 1990. LesieurH. 1983. 1997. Calif. 1975. Deviance of organizational subunits. Acquiring organizational legitimacy through illegitimate actions. CA: Sage Hughes EC. 1951. Chicago Press DurkheimE. ed. 1991. See Bamberger& Sonnenstuhl 1998. 1994. Anesthesiology 66: 670-76 Geis G. Risk analysis and the organizationalamplification of risks. See March 1965. 53-68 Feldman M. 1965. Antitrustand organizational deviance. Calif. A contingency theory of organizationalcrime. 1993. DeAnda A. Frame Analysis. Glencoe. Rattner A. Strategic Bankruptcy. Making Gray Gold. Sci. Press Finney H. ed. 1996. 17:320-27 . New York: Harper& Row GrabowskyPN. Rev. ExperimentPerilous. Behav. Anesthetic mishaps. Other People's Money. 1970. Glencoe. Making sense of organizationally based environmental disasters. See Cohen et al 1997. Organizations. J. Occupational Crime. Mead & Co. 1986.Rev. M. Messinger S. Culture and cognition. Sociol. 1999. Heimer C. The legal environmentsof organizations. High reliability organizations meet high velocity environment. Focus 18:110-22 Friedrichs D. 1997. Probl. TheRules of Sociological Method.Chicago: Univ. pp.Law Soc. 534-73 Guillemin. 5: 1-21 Guillemin J. pp. 1997. 1953. Chicago Press Cushing S. Rabe GA. Granovetter 1985. Am. Chicago: Univ. Goff B. Organizational processes (not rational choices) produce most corporate crimes. 1997. 1992. Nuclear Rites. Rev. J. Chicago: Univ. Press Fox R. Manage. NJ: Prentice Hall. Chicago Press FreudenbergW. Press Hagan FE. Chicago: Nelson/Hall Guetzkow H. 1997. 1994.Am. Profession of Medicine. 1989. Jensen G. Boston: Allyn & Bacon Hall R. 1974. Risk 3:1-35 Frey JH. Annu. Sci. 42:905-20 Diamond T. Rabe GA.CA: StanfordUniv. Doctoring Together. New York: Oxford Univ. Sutton RI. CT: JAI Fleischer A. 1992. 71-110 GephartRP. Sociol. J. Chicago Press Dalton M. IL: Free Press Crozier M. TheBureaucratic Phenomenon. 14:491-519 Hodson R. Notes on the criminogenic hypothesis. Hum. MA: Wadsworth FriedrichsD. 1989. Trusted Criminals. Sociol. Criminol. Men Who Manage. Learning from mistakes is easier said than done. 1976. 1977. The micropolitics of trouble. Sociol. 32:5-28 EisenhardtK. Polit. In Research in the Sociology of Organizations. Can. 1984. 6th ed.Dev. 23:263-87 DiMaggio P. IL: Free Press Freidson E.Berkeley: Univ.psychology. Appl. 1977. Mistakes at work. Sociol. Criminol. New York: Dodd. Thousand Oaks. Maxwell MS. Canberra: Aust. Berkeley: Univ. [1895]1966. 1992. Chicago Press 301 of Fligstein N. 1996. Acad. Criminology27:769-94 Delaney KJ.MA: Harvard Univ. 1997. Cambridge. 1959. 1994. 1996. Chicago: Univ. Social structure. Freidson E. The downsizing of America and white collar crime. 91:481-510 Green GS. Rev. WorkOccup. 16:223-44 ErmannDM. 1989. Salinger L. 10:205-25 Giddens A. Soc. 25: 121-34 Erikson K. WW Powell. Crime at work. New York: Simon & Schuster ErmannDM. Calif. Manage. The Constitutionof Society. 1982. Nat. Order Without Design. pp.See Roberts 1993 Elsbach K. Change 26:351-66 Gaba DM. Belmont. J. 1984. eds. 1995. The Transformation Corporate Control. Mixed Blessings. 23:479-515 Edmondson AC. 35:699-738 Emerson RM. Holmstrom LL. and the estimation of risk. 1998. Behav. Press Goffman E. J. Econ. SB Bacharach. pp. Chicago Press DiMaggio P. Gender and varieties of whitecollar crime. Powell WW. Corporateconcealment of tobacco hazards. Life and Death at Work. New York: Wiley Daly K. 255-99. 1964. Berkeley: Univ. Tollison RD.

Stud. Soc. Greenwich. 1991. Princeton. Am. Does the current use of junior doctors in the United Kingdom affect the quality of medical care? Soc. 1994. Chisholm D. Calif. CT: JAI Marcus A. pp. 1991. 1989. Rev. 1970. CA: Sage Jasanoff S. Crisis Q. 1986. Psychiatry and suicide: the management of a mistake. Press JackallR. Scholz JT. 3:672-82 Merton RK. Q. K Hawkins. Economic and behavioral perspectives on safety. Chicago: Rand McNally March JG. 28:438-53 Mayntz R.rationality. Sources of power of lower participantsin complex organizations. 1972. The Organization of Corporate Crime. New York: Columbia Univ. Theory 12:1 March JG. 1997. Sci. 3: 67-78 LanirZ. Aldershot. Crisis Manage. Sitkin S. Berkeley: Univ. Am. 10:601-4 Leonard WN. 1969. Bus. Sociality with objects. The uniquenessparadoxin organization. 1994. 15. Stealing into Print.New York: Free Press Meshkati N. Yeager PC. eds. J Thomas. 1940. MD: Johns Hopkins Univ. Strateg. Social Structureand Social Theory. Ethn. Cambridge. CT: JAI LaFollette M. 1994. Networks of Power. Handbook of Organizations. July/Aug:346-57 LandauM. Automakers and dealers: a study of criminogenic market forces. See Cohen et al 1997. 1992. In Research in Corporate Social Performance and Policy. Bureaucraticstructureand personality. Sociol. eds. Cambridge. The Real Disaster is Above Ground. 1998. 1968. Cont. TamuzM.Lexington. Emergence and human contact. Ind. The Developmentof Large TechnicalSystems. CO: Westview McKee M. 1962. Social structureand anomie. 5:133-54 Hughes TP. 1990. The arrogance of optimism.Sci. Risk Management and Political Culture. Greenwich. A strawmanspeaks up. Inventing Accuracy. 1997. 1990. CP Wolf. Cont. 1981. Organ. McAvoy E. Sci.Admin. 1992.MA: MIT Press Manning PK. Vol. 1988. Press Kram KE.Rev.and the problem of duplication and overlap. J. Ky. 1983. Crisis Q. Meltzer BN. ed. Cambridge. New York: Aldine de Gruyter Leibel W. Baltimore. J. SproullLS. New York: Oxford Univ. Rev. 2:207-11 Lazega E. The Logic of Images in International Relations. 126: 333-42 Merton RK. CO: Westview LaPorteTR. 1970.New York: Sage Jervis R. Simon HA. UK: Gower Light D. 7:349-64 Meltzer BN. Am. Press Kagan R. Decisions and dilemmas. 4:407-24 LernerAW. Boulder. 32355. 1993. Sociol. 1992. 1992. ed. Sources of Power.302 VAUGHAN Levi M. Psychol. 2:1-13 Marcus A. Press Jervis R. Representation in Scientific Practice. Managing with danger. Hughes TP. The Micropolitics of Knowledge. In Research in Organizational Behavior. Environ. The Phantom Capitalists. 1995. Sociol. UK: Pergamon KnorrCetina K. Oxford. 1992. J. Human factors in largescale technological systems' accidents. 1999. 1989. 1986. and corporatecrime. 1984. Warnings heeded. 1988. Redundancy. 1936. 21-54. J. The reasonablechoice of disaster. Law Soc. Mannis JG. Chance in human affairs. Moral Mazes. Press Jervis R. 1997. 1983. Organizations. 1982. Press Jamieson K. Med. Weber M.Admin. 1991. New York: Aldine Mannis JG.pp. Sci. Forces 17:560-68 MertonRK. Sci. Princeton. Reed GE. Hatch MJ. MA: MIT Press Knorr Cetina K. ed. NJ: PrincetonUniv. Soc. Black N. Q. In press Martin J. VB Shelanski. Nichols M. 1965. Woolgar S. structure. 12:479-93 LaPorteTR. Boulder. Nichols M. Feldman MS. pp. 1:894-904 Merton RK. Agency. 1995. ed. On the design and management of nearly error-free organizational control systems. Crisis Manage. 1-30 Lynch M. J. 1988. Organ. KY: Univ. MA: MIT Press MacKenzie D. 14:1-30 Kroll-Smith JS. 77:821-38 Lofquist WS. Ind. 1958. In Enforcing Regulation. Public Admin. System Effects. Couch SR. D Sills. New York: Wiley MarchJG. Sociol. Boston: Kluwer-Nijhoff Klein G. Press LandauM. Soc.NJ: PrincetonUniv. JE Post. Organizational Communication. Admin. When scientists are wrong. 1981. ThousandOaks. In Accident at ThreeMile Island. The criminology of the corporation and regulatory enforcement strategies. 1976. Perceptions and Misperceptions in International Politics. ed. J. S Bacharach. 1938. The Manufacture of Knowledge. Learning from samples of one or fewer. There is more than one way to be redundant. 18:334-59 . TheoryCult. 34:549-58 Mechanic D. Rev. 9:139-51 MarcusA. The unanticipated consequences of purposive social action.

New York: Basic Books PerrowC. Politics andthe theory of organization. Mistakes at Work. Snyderman E. 1982. J. Cultural Knowledge in Organizations. Rowan B.Cambridge. Tech. Predicting Dangerousness. ES Petry. Sociol. Westport.Amsterdam:Klewer Pinch TJ. ed. Davis Sackmann SA. Mistakes at work. Avoidable and unavoidable mistakes in child protection work. Operating as experimenting. RE Frederick. ed. 20:517-28 Oliver C. 20:763-94 Perrow C. On being sane in insane places. Information Processing and Human-MachineInteraction. Managing medical mistakes. 1991. Lloyd-Bostock S. 1976. 1991. Soc. Towarda political theory of organizational reliability. CT: Yale Univ. J. Crisis Q. Amsterdam: North-Holland Reason J. 1989. The space between laws. 1987. Medical Mistakes. J. 1997. I Thayer. Trappedin the Net. Sociol. London: Tavistock Putnam L. 1984. J. The Unkindest Cut. 1986. Accounting for error. dangeroustechnologies. Soc. ed. Princeton: PrincetonUniv. The New Institutionalism in Organizational Analysis.Newbury Park. Dawson EJ. A society of organizations. Press Passas N. MA: Ablex Quinney ER. Organ. Soc. LibuserC. 1977.Sci. 1991. Calif. Guillen M. Contradictions and paradoxes in organizations. Manage. DiMaggio P. Press Roethlisberger FJ. 3rd.New York: Random House. 1990. Lexington. Stud. 1973. Occupations structure and criminal behavior. 21: 31-33 . 1991. 1994. Secur. The social construction of facts and artefacts. Berkeley: Univ. Sci.Admin. Criminology34:357-82 Reimer JW. Organizationsas contexts. Chicago: Univ.Dep. The perils of proliferation. ComplexOrganizations. 1978. 1977. Princeton: PrincetonUniv.Int. It's not what you do.Philadelphia: Temple Univ. Philadelphia: Temple Univ.Chicago: Univ. 21:15-26 Rochlin GI. Probl. 1990. A ComplexSorrow. Science 250:1210-16 Perin C. Acad. 1983. Organizational aspects of engineering safety systems. JacksonD. ed. Organizational crime. ed. 1988. Calif. 83: 340-63 Michalowski RJ. Values 23:36-57 PerrowC. Environ. New Challenges to Understanding Organizations. Probl. Press Rosenhan DL. Leaders. Bijker WE. Sociol. Press Sagan S. Dickson WJ. or purposefulunintendedconsequences in the management of complex. Cambridge: CambridgeUniv. KramerRC. Human Error. 1996.CT: Quorum Pate-Cornell ME. Soc. Implementation. New York: Morrow MizrahiT. New Haven. Sociol. 11: 179-85 Rasmussen J. From Bhopal to banking. 1997. Dyn. 34:34-53 Millman M. 1993.. 1986. MA: Lexington Books Pinch T. TheLimitsof Safety. Admin. Normal Accidents. Institutionalized organizations.riverboat gamblers. The organizationalcontext of humanfactors engineering. Organizationaloffending and neoclassical criminology. Forum 12:519-43 Mulkay M. J. Chicago Press Morrill C. 1996. Crisis Manage. 1990. Hum. 28:521-41 Perrow C. 1. 9:152-74 Perin C. Press Punch M. 1993. Cont. 14:399-441 Powell W. 1984. 1995. TR LaPorte. CA: Sage Sagan S. In The Ethics of Accounting and Finance. 16:165-83 Munro E. 7:106-29 MorrillC. TheorySoc. 1973. Law Econ. 1993. 1988. In Organization Communication: Emerging Perspectives. 31:924-47 Paget MA. Rev. Chicago Press Pressman JL. Management and the Worker. Wkly. Press Paget MA.Norwood. London: Open Univ. but who you are. 1993. Med. New York: Macmillan RobertsKH. The 'logic' of organizational irrationality. 1994. Q. How do we treat technical uncertainty in systems failure? In Social Re- 303 sponses to Large Technical Systems. Am.26:793-808 Needleman ML. 1985.DARK SIDE OF ORGANIZATIONS Meyer JW. Q. 2:228-40 Schulman PR. Soc. Press Pfohl SJ. 1979. TheAIDS Disaster. UK: Ashgate Reed G. 1991. Gilbert GN.. Conduct Unbecoming. 1995. Accounting for fraud. 1991. WM Hoffman. Wildavsky A. Yeager PC. TheExecutive Way. 1999. Vol. Acad. Perrow C. Soc. Managing the Risks to OrganizationalAccidents. Sci. Org. 1986. JB Kamm. Strategic responses to institutional processes.MA: HarvardUniv. 1997. Mulcahy L. Manage. Probl. Press Roth JA. 1984. 1963. Soc. Sci. Ind. 16: 145-79 OsbornR. 19:135-46 Moe T. 1996. 1998. Science 179:1-9 Rosenthal M. 1947. The Unityof Mistakes. Aldershot. Br. Press Reason J. 18:66-107 Sagan S. 1991. Needleman C. 23:255-67 Roberts K. Univ.Sociology.

Collaring the crime. TheChallengerLaunchDecision.MA Snider L. 1976. 1998. Calif. Normal crimes. 39:80-102 Vaughan D. Toronto:Univ. 26:501-24 Stelling J. TorontoPress Star SL. Pearce F. PhD thesis. Wherethe Law Ends. Rev. Rev. J. ed. 1992. M Blankensmith. New York: Garland Shrivastava P. Press Shapiro S. Hewitt JP. 1973. Piquero N. Milliken FJ. 1997. CT: JAI Staw BM. Ariz. Plans and Situated Actions. Sandelands LE. March1965. Crimesof the CriminalJustice System. not the criminal. Press Turkle S. 49:711-25 Short JF. New York: New Am. Cincinnati. Towardunderstanding lawful organizationalbehavior.Sci. Sociol. Bhopal. Organizations. Transl. 142-93 Stokes R. New York: Knopf Thomas R. Soc. 1998. Library Sutherland EH. 1949. 41:838-49 Stone CD. Selznick P. Executives' perceptualfilters. Q. pp. 1978. [1906]1940. situated action. Soc. UK: Cambridge Univ. New York: Dryden Tenner E. See Bamberger & Sonnenstuhl 1998. 1993. Challenger: fine-tuning the odds until something breaks. PaternosterR. 1981. The social fabric at risk. the macro-microlink. Press VaughanD. Ecologies of Knowledge. Clarke L. In Understanding Corporate Criminality. Probl. New York: Simon & Schuster TurnerB. BucherR. 173-202. PhD thesis. CO: Westview Shover N. Chicago: Univ. 1997. pp. GersonE. Soc. New York: Cambridge Univ. 1982b. Calif. Rev. pp. Sociol. Transactionsystems and unlawful organizational behavior. 1999b.304 VAUGHAN Stinchcombe AL. Am. Vocabulariesof realism in professional socialization. Rev. 1965.OH: Anderson Simpson S. Stud. 1950. and the so- SchulmanPR. NJ: Prentice Hall. Milliken FJ. Top management characteristics. Chicago: Univ. 29:373-79 Vaughan D. Social structureand See organizations. New York: Free Press Simon D. HarvardUniv. New York: Cambridge Univ. Cambridge. Sci. 21:378-97 Turner B. In Whatis a Case? ed. Berkeley: Univ. Sociol. 2nd ed. London: Butterworth-Heinemann. Chicago Press. Chicago: Univ. Theorizing: Analogy. Admin. The trickle-down effect. Controlling Unlawful Organizational Behavior.Am. 1996. 1995. Berkeley: Univ. Corporate Crime. WhyThings Bite Back. Practical drift. 13:373-404 Simpson S. 1997. 141-76. 1996. The analysis of high reliability organizations. Soc. unVaughanD. and Risk. 1987. NY Press StarSL. 1993. Northwest. ed. Rational choice. ed. 1949. 4th ed. 1983. Assessing social error. The organizationaland interorganizational development of disasters. 1994. 25:319-40 StarbuckW. Univ. organizational strain. 1987. Simmel G. On risk communication as interorganizationalcontrol. Dutton JE. Chicago Press Vaughan D. and ComparativeSocial Oganization.Greenwich. ed. Soc. Cambridge: Ballinger Shulman D. 1998. Boulder. Law Soc. The social organization of workplace deception. Exploring the micro-macro link in corporate crime research. Cases. In Nothing to Fear.Sci. 1996. Press Simpson S. Tucson: Univ. TheSociology of Georg Simmel. Man-Made Disasters. K Wolff. Folkways. The managementand dynamics of anomalies in scientific work. 1995. WhatMachines Can't Do. 1992. Theory elaboration. Calif: Manage. DC Hambrick. In press Vaughan D. 76:661-75 . Rev. 1988a. Press Tilly C. Manage. ed. Probl. 1975. Cambridge. Quant. 32:23-61 Vaughan D. 1965. 35-70 Singer B. and the social control of organizations. 1990. Uncertainties. 1984. White-Collar Crime. Press Sudnow D. Med. 1998. C Ragin H Becker. 1982a.. 12:255-76 Sumner WG. Boundarywork: levels of analysis. 1978. In The Executive Effect. 1987. Law Rev. 203-39. TVAand the Grass Roots. 80:1377-1402 Vaughan D. New York: Harper Suchman LA. A Kirby.Upper Saddle River. Koper C. Organizations. 1988b.J. 55:346-65 Short JF. 1990. Am. The invisible elbow. 1997. Albany: State Univ. Forum 2:589-601 Tompkins PK. 1994.See Roberts 1993 Scott WR. Man-madeDisasters. Policy 9:27-34 Snook SA. 1999a. Soc. Threat-rigidity effects in organizational behavior. Pidgeon N. Q. Bryant K.Admin. Mich. 28: 147-69 StarbuckW. and antitrust offending. eds. Theoretical explanations of corporatecrime. 1995. 1976. Life On The Screen. Q. Criminol. pp. London: Wykeham TurnerB. Aligning actions. Why Corporations Obey the Law. pp. 1996. Chicago Press Vaughan D.

E Sagarin.DARK SIDE OF ORGANIZATIONS cial control of organizations. 1982. 1977. Obstfeld D. New York: Columbia Univ.Sci. Unrulytechnology. Beverly Hills: Sage . 1990. Roberts KH. 1988. 87-138. Boston College. 1979. Am. Greenwich CT: JAI. Institutionalizedrule violation in professional hockey. Q. 1995. 1996. Banking on Fraud.In Social Science. Manage. 1987. Hickson DJ. 42:726-43 ZuckermanH. Q. Soc. New York: Aldine Zucker L. MA thesis. RothmanML. IA: KendallHunt Verma K. Press Zerilli S. Stud.Admin. 18:147-67 Zelizer V. Sci. The collapse ofsensemaking in organizations. Sociol. Miller S. The organization as weapon in white-collar crime. ed. In press Wheeler S. Social Policy. 28:112-27 Weick KE. R Sutton. The role of institutionalization in cultural persistence. Reading. In Sport Sociology. In Deviance and Social Change. Rev. The Social Psychology of Organizing. Am. 305 Organizing for high reliability. How organizations can overbalance. R Kagan. Calif. 1979. Dubuque. Morals and Markets. MarcusA. MA Zey M. 1977. Environ.New York: Russell Sage Vaz E. 16:571-93 Weick KE. 38:628-52 Weick KE. P Ewick. Crisis Q. 1999. ed. 1993. 80:1403-26 Willis P. A Yiannakis. Four common tactics of confession-makers. CA: Sage Weick KE. Organization culture as a source of high reliability. Law Rev. MA: AddisonWesley Weick KE. 1995. Behav. Rev. Ind. ed. 1977. ThousandOaks. J. Causes and effects of rising production costs in US nuclear power industry. Collective mind in organizations. 1998. A Sarat. Sci. Press Wilson DC. Sutcliffe KM. 39:995-1010 Wynne B. ChestnutHill. In Research in Organizational Behavior. Deviant behavior and social control in science. Learning to Labor. 1979.Admin. Sensemaking in Organizations. Mich. Manage. 38: 357-81 Weick KE. 1993. The vulnerable system. and the Law. B Staw. 9:242-58 Weick KE. Sci. ed.pp. 1993. New York: Columbia Univ.