05A1-08 - Analyze the Consequential and Nonconsequential Theory according to your chosen ethical dilemma

Reproduced with permission of M.E. Sharpe, Inc. in the format electronic usage via Copyright Clearance Center.

. Whistleblowing, Organizational Performance, and Organizational Control James L . Perry Recent headlines tell a story of a rising tide of conflict within organizations:

"NASA Critic Says Job Binds Him to Truth, Not Agency," "Tennessee Valley Authority Stung by Critiques of its Nuclear Program," "Troop Carrier's Flaws Masked, Memos Say," "Weapons Costs Underestimated for 30 Years, Official Charges," "Maverick at USDA Attacks Inspections." The headlines reveal the growing incidence of disputes between individuals and their orga nizations, often about matters with substantial implications for groups not party to the dispute. Headlines have also given a name to this phenomenon:

Whistleblowing.

Although whistleblowing is like other types of organizational disputes (for example, those between departments or between labor and management), it also has distinctive attributes. It may be characterized by three procedural and sub stantive properties that collectively distinguish it from other types of organiza tional disputes: (1) apparent altruistic behavior on the part of the initiator of the dispute; (2) asymmetric distribution of power; and (3) absence of well-devel oped, neutral dispute resolution mechanisms.

Whistleblowing is unique in that it involves individuals initiating disputes apparently on behalf of third parties not directly involved in the employment or contractual relationship--for example, taxpayers or stockholders. The disputes are apparently on behalf of third parties because the actual motivations for whistleblowing may reveal that self-interest rather than altruism (Mansbridge 1990) is the driving force behind a whistleblower's claim. Personal vindictive 4 80 JAMES L. PERRY ness toward a supervisor or an organization could plausibly be the trigger for charges about organizational wrongdoing or illegitimate activity. The important point is that the claim could appear outwardly to be altruistic, regardless of an individual's motivation.

Another feature of whistleblowing is that it typically involves conflict be tween two parties with unequal power. Weinstein (1979, 2) suggests that whistleblowing involves "attempts to change a bureaucracy by those who work within the organization but who do not have authority." The whistleblower also is likely to be at a power disadvantage because the resources that can be brought to bear in any conflict are asymmetrically distributed. Whistleblowers can rely on their personal resources and whatever additional resources they can persuade others to commit to the conflict. In contrast, any organization that chooses to ignore the claims of a whistleblower and contest the dispute can readily mobilize organization members, legal services, and its network of external supporters to resist responding to legitimate claims. Thus, the whistleblower can expect to face an opponent who is structurally guaranteed a power advantage, often of immense proportions. A third distinguishing characteristic of whistleblowing is that it usually occurs in the absence of well-developed, neutral dispute resolution mechanisms. Third party procedures are now used to resolve a variety of disputes, ranging from labor-management interest disputes, commercial contract disputes, consumer claims, and conflicts between community groups. The claims of whistleblowers seldom are afforded opportunities for third-party review and resolution. In some environments, such as the federal government, the whistleblower may have some organizational mechanisms for third-party review of the evidence and disposition of the claim. Even with such procedures, the whistleblower is at a disadvantage when compared to other dispute contexts. The recency of procedures for whistleblower disclosure has prevented them from becoming fully institutional ized. One indicator of lack of institutionalization is that the procedures are not widely accepted by the parties affected by them. For example, after an analysis of survey data from federal employees, Miceli and Near (1984, 704) reported that "finding the right encouragements or inducements for whistle-blowers might be problematic and certainly will require long term, concerted effort." Another problem is that, despite good intentions, the procedures are sometimes poorly designed for protecting the whistleblower and assuring disclosure of wrongdoing (Devine and Aplin 1986).

Research on whistleblowing has grown significantly in recent years (Near and Miceli 1986; Graham 1986), but much of the research has focused on the indi vidual whistleblower, particularly the whistleblower's motivations and cognitive and affective decision processes. Significantly less research has been devoted to putting the whistleblowing process into its larger social context (Johnson and Kraft 1989). This chapter begins to fill that void by analyzing organizational consequences of whistleblowing. WHISTLEBLOWING 81 Theoretical Framework In creating a theory about whistleblowing and its consequences, it is useful to start with an understanding of the scope of the phenomenon. Building from an earlier definition by Miceli and Near (1985), whistleblowing is conceived as the disclosure by individuals or groups of activities claimed to be illegal, immoral, or illegitimate. One implicit assumption underlying the definition is that the whistleblower is not a passive participant or bystander who merely initiates the process by reporting a claim of wrongdoing. Some scholars (Weinstein 1979; Zald and Berger 1978) have argued that the whistleblower has no legitimate authority to effect change but must instead work through unconventional chan nels. This contrasts with the situation in some organizations in which individuals who occupy specific roles are authorized bring illegal or illegitimate activities to the attention of authoritative actors, thereby creating a prima facie case for change. The assumption of the present study is that whistleblowers will derive their power from both formal and informal channels within organizations, and the phenomenon should not be defined to preclude either source of power.

The definition does not presuppose the validity of the whistleblower's claim.

Much of the literature on whistleblowing assumes or limits the domain to dis putes that are grounded in matters of principle, that is, where dissent involves conscientious objectionto violations of standards of justice, honesty,oreconomy (Bowie 1982; Graham 1986). From an organizational perspective, however, the substance of a whistleblower's complaint is a matter that involves some degree of uncertainty. Although the authenticity of a whistleblower's complaint may be irrelevant for the organization that chooses to ignore it or to retaliate against the whistleblower, it is clearly relevant to the organization that wishes to respond appropriately. Responsive organizations are faced with investigating the com plaint to identify whether it is authentic or not.

Any theory of the consequences of whistleblowing must account for varia tions in the validity of claims that might potentially be brought by a whistleblower. Claims are likely to be of three types. The type that is most often assumed in the literature involves the reporting of wrongdoing that is accurately perceived by the whistleblower. At least two other types of claims, however, are likely to manifest themselves. One type of invalid complaint arises when the whistleblower's action is grounded in erroneous perceptions or incomplete infor mation about an activity. Another, more serious, instance involves whistleblow ing grounded in opportunistic self-interest, where the whistleblower fabricates a problem for self-protection or gain.

Because the validity of whistleblower claims may initially be uncertain, the content of the claim is not a given. Instead, the claim is subject to definition by the whistleblower, the targets of the claim, and other contending parties. It is conceptually inaccurate, therefore, to conceive of the claim as a "performance gap" or "problem" as Graham (1986) does with regard to the more limiting 82 JAMES L. PERRY concept of principled dissent. This study, therefore, will use the broader concept of "issue," referring to the controversy initiated by a whistleblower. Following Cobb and Elder (1983, 82), an issue refers to a conflict between two or more identifiable parties over procedural or substantive matters.

Underlying this research are several assumptions about the behavioral predis positions of individuals and organizations. One assumption is that an organiza tion will resist change attempts unless it has learned responses or designed itself in a way to facilitate positive or routine handling of disputes. Resistance to change is grounded in two general tendencies of individuals and organizations.

The first of these tendencies involves what Staw, Sandelands, and Dutton (1981) term the threat-rigidity effect. The general thesis of Staw et al . is that "a threat to the vital interests of an entity, be it an individual, group, or organization, will lead to forms of rigidity" (p. 502). The two most prominent processes contribute to rigidity in the face of threat are restriction of information process ing and constriction of control. Whether these responses to threat are functional or not depends on the magnitude of the threat itself. Rigidity in response to potentially radical environmental changes is likely to be maladaptive, but re sponse rigidities to incremental changes may be quite functional.

A second, related tendency may be described in terms of how people respond to different accountability mechanisms (Jos 1990). Staw (1980) has argued that individuals exhibit different forms of rationality in response to different account ability mechanisms. Decision makers exposed to predecisional accountability are likely to exhibit prospective rationality, in which the decision maker devotes considerable effort to finding the most defensible options. However, when peo ple are accountable only after a decision has been made, they are likely to resort to retrospective rationality-to cope by trying to justify their actions. These justification processes may range from simple verbal rationalizations to elaborate excuses designed to minimize damage to their social images (Tetlock 1985).

Thus, whistleblowing, because it represents a type of postdecisional accountabil ity, is likely to set into motion countervailing forces that motivate its targets to bolster previous decisions (Tetlock 1985). Whistleblowing and Organizational Consequences The organizational consequences of whistleblowing may be conceived in both micro- and macro-organizational terms. Outcomes at the micro level focus on the reactions of the organization to the whistleblower's claim and its resolution (Graham 1986). At the macro level, several studies, most of them outside the whistleblower literature, are suggestive of the possible linkages between whistleblowing and more general processes of organizational change and stabil ity. This study focuses on the latter consequences of whistleblowing.

Some insight about salient macro-level outcomes is provided by Zald and Berger (1978). Their central thesis is that much conflict in organizations, includ WHISTLEBLOWING 83 ing whistleblowing, occurs outside normal channels and such "unconventional opposition" may appropriately be understood using social movement analysis.

They contend that phenomena resembling social movements are important be cause they affect major outcomes in organizations, including the control of or ganizational resources and organizational survival. The Zald and Berger analysis suggests that the whistleblowing process and its aftermath might influ ence a broad range of outcomes, two of which are particularly salient: organiza tional control (Pfeffer and Salancik 1978) and organizational performance (Hirschman 1970).

The resource-dependence perspective (Pfeffer and Salancik 1978) suggests that organizational choice is socially controlled. The organization's dependence on the environment for resources creates the conditions under which external entities have power the organization. This power permits environmental elements to constrain and control organizational action. The likelihood of con trol, however, is not deterministic. It is contingent on the presence of conditions that encourage an organization to comply with the demands of extemal entities.

The crucial point of the resource-dependence perspective for this research is that the actions of the whistleblower affect the stability of pattems of exchange in the markets of influence and control within the organization, which, in turn, affect environmental dependence.

A second macro-level outcome can be tied directly to the whistleblower phe nomena using Hirschman's (1970) theory of exit, voice, and loyalty. Hirschman sought to explain member responses to declines of performance in organizations and administrator responses to members' attempts to influence the organization.

He reasoned that declines in the rationality of organizations are likely and that the behaviors of members as responses to these declines can lead to corrections in inefficiencies or ineffectiveness.

The member response to performance decline that has typically been the focus of traditional economic analysis is exit, in which a member leaves an organization or ceases to buy a firm's product. Hirschman identified a second generic form of member response-voice-which represents a political re sponse. Voice involves the direct expression of dissatisfaction to management or to some other authority. Although Hirschman treats exit and voice as essentially mutually exclusive, Kolarska and Aldrich (1980) have argued that they represent four responses if they are considered jointly: (1) stay and be silent; (2) stay and be vocal; (3) exit and remain silent; and (4) exit vociferously.

It is obvious that the act of blowing the whistle indicates a choice between one of two of Kolarska and Aldrich's four responses. Whistleblowers opt to stay and be vocal or they exit vociferously. In some instances, the behavioral options will be used sequentially if the organization is not responsive to an employee's efforts to. effect change from within. The important conceptual point is that these responses are intended to correct declines in organizations and should, therefore, be linked to organizational performance. 84 JAMES L. PERRY Hypotheses A basic premise of this research is that organizations will resist change attempts unless the organization has learned responses or designed itself in a way to facilitate positive or routine handling of disputes (Staw, Sandelands, and Dutton 1981). The veracity of whistleblower claims will vary greatly, with some whistleblowers raising legitimate claims and others raising false issues to punish co-workers or superiors (Westin 1981).. How the organization is designed to respond to these different types of claims is important. Does the organization make type I (rejecting true claims) or type II (accepting false claims) errors?

Does it reward illegitimate claims? Does it punish legitimate claims? An organi zation cannot know in advance the mix of legitimate/illegitimate claims, but it can make strategic choices with respect to acceptable levels of error for each type. Hirschman (1970,42) suggests that institutions may be designed to lower the costs of voice and increase its rewards as a means for taking advantage of the recuperative powers of voice. Thus, the design choices that organizations make are likely to influence organizational consequences, which suggests the follow ing hypothesis:

Hypothesis 1. The more organizational policies tend to punish legitimate claims and reward illegitimate claims, the more likely organizational perfor mance will suffer.

Q'Day (1974) suggests that superiors in organizations are likely to engage in intimidation rituals to discourage reform-minded individuals from continuing to seek reform. According to Q'Day, the increasingly oppressive rituals have two purposes. First, the rituals are intended to control reformers so that they do not succeed in recruiting others to their cause. Second, the rituals are designed to maintain the image of authorities. The most direct manifestation of these rituals is the organization's hostility toward the whistleblower, under the guise of infor mal reprimands or more formal disciplinary action. Ultimately, the rituals shift the focus from the whistleblower's claim of alleged immoral, illegal, or illegiti mate activity to the whistleblower's deviance from or noncompliance with or ganizational norms (Harshbarger 1973). By obscuring the facts, the conflict diminishes the organization's capacity to make rational judgments. Thus:

Hypothesis 2. The greater the hostility toward the whistleblower, the less likely whistleblowing will improve organizational performance.

Another important dimension of the whistleblowing process involves whether the issue created by the whistleblower's claim is resolved. In some cases, closure will occur when' an authoritative source vindicates the whistleblower's position and punishes wrongdoers. Various other scenarios are . WHISTLEBLOWING 85 possible as well, among them that the issue will not be resolved and the contro versy will continue or dissipate without any formal resolution. Even if the whistleblower's claim has no merit, it is unlikely to be to the organization's advantage to let the controversy hang fire so that it is a continuing distraction and irritant to the whistleblower and other organizational participants, such as co-workers and supervisors. It can be concluded, therefore:

Hypothesis 3. The more able the organization is to resolve issues, the greater the likelihood that whistleblowing will improve organizational performance.

Organizations will strive to maintain their autonomy. The exercise of voice (Hirschman 1970) by the whistleblower threatens organizational autonomy. If the organization is sensitive to voice as a recuperative mechanism, the organiza tion is more likely to act upon the claims of the whistleblower and is less likely to lose any of its autonomy over goal setting or resource allocation. On the other hand, if an organization ignores the whistleblower or reacts with hostility, then the organization risks the intervention of third parties, possible imposition of a solution, and public disapproval and sanctions.

The extent to which whistleblowing affects organizational control is a func tion of the arena in which the whistleblower's claim is resolved. Externally derived solutions are likely to involve the imposition of mandates upon the organization to comply with certain modifications of organizational policies or practices. Even if externally derived solutions do not prescribe specific organiza tional changes, they are likely to increase an organization's vulnerability to external control (Pfeffer and Salancik 1978). This implies the following:

Hypothesis 4. Externally derived solutions to whistleblower claims are likely to reduce organizational control.

Another influence upon the extent to which whistleblowing affects organiza tional control is the power of the organization to defend itself against the whistleblower's claims. If organizations are driven to maintain autonomy, their ability to overcome threats to their control is dependent on the amount of power they wield. Therefore, Hypothesis 5. The greater an organization's power, the more capable it will be of defending itself against a whistleblower's claim and the less the effect on organizational control will be.

Hirschman's theory has implications for predicting organizational conse quences under varying assumptions about the responsiveness of organizations to member voice during declining performance. For instance, Hirschman suggests that an organization's recovery from performance decline depends on its respon 86 JAMES L. PERRY siveness to whatever recuperative mechanisms it is equipped with. One implica tion is that organizations that are equipped with voice mechanisms but are re sponsive to exit are less capable of recovering from performance declines and more likely to suffer permanent decrements to performance.

Organizations that discourage dissent are prone to the most serious conse quences when whistleblowing does occur. Although an organization may suc cessfully discourage voice, it does so by sacrificing opportunities to correct performance declines. In so doing, however, it assures that should a whistleblower succeed, the magnitude of any potential correction is likely to be greater than if a small number of corrections had been made incrementally. Thus, Hypothesis 6. Organizations that discourage dissent will suppress internal and external whistleblowing, but when whistleblowing does occur it is likely to lead to reductions in organizational control.

Methods The research used the case survey approach first described in Yin and Heald (1975) and further developed by McClintock, Brannon, and Maynard-Moody (1979) and Yin (1981a, 1981b). As described by Yin and Heald (1975, 372):

"The case survey calls for the reader-analyst to answer the same set of questions for each case study. The questions are close-ended, so that the answers can be aggregated for further analysis." The quantitative evidence derived from each case study permits the computation of statistical measures of association, over coming one of the shortcomings of case studies-the inability to aggregate across cases. Thus, the case studies can be reviewed with scientific rigor. Sample The sample of cases was drawn from agencies of the federal government. In order to obtain a sample that was representative with respect to the intensity of the dispute, it was necessary to stratify cases (Kish 1965) by whistleblowing channel. We originally set out to sample cases from three strata: intraorganiza tional/informal channels (the management hierarchy, co-workers, unions, and personnel offices); formal channels (inspectors general, hotlines, office of the special counsel); extraorganizational/informal (Congress, media). After initiating contacts with various sources to identify cases, we found that very few cases could be identified for the first stratum, intraorganizational/informal channels.

There appeared to be several reasons for this difficulty. First, we found that few informants considered disputes that were settled within the management hierar chy to be whistleblowing. Informan,ts implicitly associated whistleblowing with disloyalty to the organization or with violation of the norms of commitment to the organization or co-workers. Informants were probably also less likely to WHISTLEBLOWING 87 recall incidents occurring within intraorganizational/informal channels because these disputes might have involved low levels of conflict, been quite routine, and produced few, if any, artifacts (for example, case reports of adjudicators). Sec ond, because whistleblowing in the federal government is highly formalized, many individuals who seek to report illegal, illegitimate, or immoral activities prefer to use formal channels rather than the organizational chain-of-command.

Even in field installations, most complaints are processed through formal whistleblowing channels.

Because of the difficulties encountered in identifying cases from the in traorganizational/informal stratum, the sample that was finally selected was com posed of cases from only two of the originally defined strata. Although this resulted in a sample that was less diverse than anticipated, it was consistent with the realities encountered in the field research. A summary of characteristics of the sample is provided in Table 4. 1. The sample of cases represents a wide spectrum of the organizational hierarchy, with relatively equal representation from top management, middle management, and nonmanagement positions.

About one-third of the sample suffered some form of reprisal; about half of these had been fired. Most cases for which information was available were resolved Data were collected on each case from archival records (memoranda, court cases, agency records, budgets), secondary sources, interviews, and a self-admin istered survey. An effort was made to build redundancy into the data collection procedures so that if information were not available from one source it could be obtained from another. The self-administered survey requested information (for 88 JAMES L. PERRY Table 4.1 A Comparison of Federal Government Whistleblowing Cases and the Effects of Several Organizational and Personal Variables on Outcomes Valid Valid Variables/Categories Frequency Percent Percent Cases Whistleblower's Position 87 Top Management 17 13.1 19.5 Middle Management 22 16.9 25.3 Supervisory 15 11.5 17.2 NonmanagementlNonsupervisory 33 25.4 37.9 Missing 43 33.1 Was Reprisal Taken 130 Yes 44 33.8 33.8 No 74 56.9 56.9 Unable to Ascertain 12 9.2 9.2 Organization Fired Whistleblower 109 Yes 24 18.5 22.0 No 85 65.4 78.0 Missing 21 16.2 Claim Resolution 107 Resolved 84 64.6 78.5 Partially Resolved 10 7.7 9.3 Not at all Resolved 13 10.0 12.1 Missing 23 17.7 Was Whistleblower Anonymous 128 Yes 43 33.1 33.6 No 85 65.4 66.4 Missing 2 1.5 How Long to Conclude Case 87 0-1 year 41 31.5 47.1 1-2 years 16 12.3 18.4 years 10 7.7 11.5 3-4 years 8 6.2 9.2 4-5 years 3 2.3 3.4 More than 5 years 9 6.9 10.3 Missing 43 33.1 Was Address Obtained 130 Yes 105 80.8 80.8 No 25 19.2 19.2 Was a Survey Returned 130 Yes 57 43.8 43.8 No 73 56.2 56.2 Source of Case 130 Public Case 53 40.8 40.8 MSPB/OSC 27 20.8 20.8 GAP (Government Accountability Project) 8 6.2 6.2 HHS (U.S. Department of Health and, 7 5.4 5.4 Human Services) Field Office Hotline 16 '12.3 12.3 HQ Office Hotline 19 14.6 14.6 WHISTLEBLOWING 89 example employee demographics) that was also obtained from other sources such as court cases or newspaper reports. The redundancy assured that if one of data were not available it might still be obtained using the alternative source. The data were used to create three data files for each case: (1) the case survey file, containing analyst-coded information based upon archival records; _ 90 JAMES L. PERRY which the organization acquiesced to dubious claims in past whistleblower cases.

An agency score was created from responses to an item from the 1983 Merit Principles Survey. External whistleblowing channels refers to those formal and informal chan nels that lie outside an· organization. This item was created from responses to items in the case survey and self-administered survey. Externally derived solutions are solutions to the whistleblower's claims that are imposed by external authorities. Solutions were categorized according to whether the party was internal or external. Organization power is the extent to which the organization possesses re sources that may be used to persuade other actors of its position on an issue. The indicator for this concept was the size of the budget for the parent organization measured in constant dollars in the year in which the whistleblower claim was made. The data were taken from published budget documents. Suppression of whistleblowing is the extent to which the organization takes actions to prevent whistleblowers from pursuing their claims both inside and outside the organization. It was measured using agency-level indicators of per ceived reprisal created from items in 1980 and 1983 MSPB surveys.

Results Multivariate statistical techniques were used to test the hypotheses. The regres sion results are presenfed in Tables 4.2 and 4.3. Table 4.2 presents three regres sions. The first regression includes only the indicator for punishment of legitimate claims, the second includes only the indicator for reward of illegiti mate claims, and the third incorporates both measures.

The organizational performance regressions provide little support for the hypotheses. The relationships for hostility and issue resolution are in the pre dicted directions, but none of the relationships reach the 0.05 level of signifi cance in any of the three regressions.

The F-statistic is significant at the 0.05 level for only one of the three regres sion equations. The adjusted R2 ranges from 2 percent to 6 percent for the three equations, which also suggests that the variables are relatively poor predictors of organizational performance.

Table 4.3 provides a contrasting picture for the predictors of organizational control. The F-statistics for two of the three independent variables and the over all regression were significant. Organization power was not significant in the regression. The results indicate that externally-derived solutions are more likely to be associated with reductions in organizational control. They also suggest that suppression of whistleblowing could be detrimental to organizational control.

Discussion This study began with a theoretical model that was grounded in a basic assump tion about organizational behavior. The assumption was that organizations will WHISTLEBLOWING 91 Table 4.2 Multiple Regressions for Organizational performance Standard Independent Variables Beta Error cance Hostility -0.16 0.13 -1.27 0.21 Issue Resolution 0.20 0.11 1.79 0.08 Punishment of Legitimate Claims -0.02 0.12 -0.20 0.84 Adjusted R2 = 0.06 F=2.87 Significance =0.04 N=87 Hostility -0.19 0.12 -1.69 0.09 Issue Resolution 0.15 0.11 1.32 0.19 Reward of Illegitimate Claims -0.05 0.11 -0.46 0.65 Adjusted R2 = 0.04 F= 2.31 Significance = 0.08 N=84 Hostility -0.17 0.14 -1.22 0.22 Issue Resolution 0.14 0.12 1.14 0.26 Punishment of -0.04 0.13 0.76 Legitimate Claims Reward of Illegitimate Claims -.05 0.11 -0.48 0.63 Adjusted R2 = 0.02 F= 1.48 Significance = 0.22 N=80 resist change attempts unless they have learned responses or designed them selves in ways that facilitate constructive handling of disputes. Although the empirical research did not investigate this premise directly, it assessed a series of hypotheses that flowed from the basic logic of this assumption and the results of previous research on whistleblowing. The hypotheses attempted to predict the organizational consequences of whistleblowing. A statistical analysis of cases of whistleblowing in the federal government provided support for some aspects of the theory articulated in the hypotheses, but it also disconfirmed several of the hypotheses. Contrary to one of the hypotheses, organizational performance was not af fected by the extent to which organizational policies punished legitimate claims and rewarded illegitimate claims. as measured by aggregated perceptual mea 92 JAMES L. PERRY Table 4.3 Multiple Regression for Organizational Control Independent Variables Beta Standard Error Significance Suppression Organizational Power External Solution 0.38 0.01 -D.37 0.10 0.10 0.10 3.92 0.12 --3.56 0.00 0.91 0.00 Adjusted R 2 = 0.23 F =Statistic (Equation) =9.46 FSignificance = 0.00 N=84 sures. Given the difficulty of measuring the concepts involved in this hypothesis, one reason for the absence of association may be shortcomings in indicators for the reward and punishment variables. Considering the retrospective nature of the study, however, the aggregated survey measures were the best sources of data available.

What do the results suggest about the voice-performance relationship derived from Hirschman's theory? Three potential inferences merit discussion: one in volves the assumptions of Hirschman's theory, another involves the context in which this study was conducted, and a third involves the efficacy of individual action. These explanations cumulatively could account for the weak association between voice and performance.

The logic of the Hirschman framework is that organizations will be attentive to information about performance decline. This assumes, however, some com monality of interests between the organization (and perhaps, more appropriately, the various interests represented within it) and members who engage in exit or voice. The commonality of interests assumes that conflicts that arise are of the consensus variety, about means rather than ends. The disputes that arise from whistleblowing are often of the dissensus variety, fundamental disagreements about direction of the organization or its social role. The fundamental nature of some disputes may make them poor candidates to be resolved through the dy namics envisioned by Hirschman. Thus, the nature of some disputes may attenu ate a voice-performance linkage. One whistleblower described the dilemma:

While working as a contract librarian, I stumbled upon many illegal, fraudu lent, abusive practices. When the administrator became aware that I knew of these practices, he terminated my position. The library has since become a shambles. Personnel are not given the benefit of up-to-date medical journals and articles .... Those who caused and condoned the illegal activities never answered for them at all. In fact, the chief offender is to this day drawing a large salary from the government. .

WHISTLEBLOWING 93 Another factor that may influence the voice-performance linkage is the clar ity of performance norms within an organization. Although Hirschman's theory was quite general, it relied predominantly on argumentation grounded in market based organizations, where the benefit of success is profitability and the cost of failure is bankruptcy. In federal agencies the criteria of "good" performance are less clear. Federal agencies are responsible for accomplishing particular mission goals, but they are also under substantial pressures to satisfy political interests and conform to constituent and societal expectations.

If good performance is ambiguous or is defined as conformity, then voice may be a relatively ineffective mechanism for correcting "performance declines." One whistleblower described the contrasting visions of performance that existed in a military installation:

I spent two years in the DODDs school system. In that period of time I witnessed rampant abuse both on the job and on the army base. I was forced to throwaway equipment that had never been used under threat of removal from my position. Bulk orders of supplies led to excess and shortages in teaching materials. A system of strict "chain of command" made resolution impossible. On the base, before inspection, hundreds of dollars of items were discarded so they would not come up on the supply sheet. Overall, I found the program of waste, fraud and abuse to be channeled toward petty things while major abuses were ignored.

Yet a third explanation for the absence of a voice-performance linkage is the sensitivity of performance to individual action. We know that individual exit over time-for example, customers who choose to discontinue use of a particular brand--can send a strong message to an organization about the need to improve its performance. It is not clear that the act of a single individual conveying the same message will be equally successful. Comparison with other individual action contexts is only suggestive, but Pfeffer and Salancik (1978, 12) note the limited explanatory power ofleadership in research on administrative performance:

Studies estimating the effects of administrators have found them to account for 10 percent of the variance in organizational performance, a striking contrast to the 90 percent of the intellectual effort that has been devoted to developing theories of individual action.

It is conceivable that the efforts of whistleblowers, even when they may be described as heroic, will have only small consequences for organizational perfor mance.

To summarize, it appears that several factors may have operated to attenuate the voice-performance linkages inferred from general theory. Among these fac tors are the dissensus nature of some whistleblowing disputes, the ambiguity of performance norms in the sample of organizations studied, and the limited ef fects of individual action on organizational performance. 94 JAMES L. PERRY The results of the research are far clearer with respect to organizational con trol than to organizational performance. An organization's control of its re sources and decision-making autonomy can be affected by a whistleblower's actions. The results indicate that issues that are resolved externally have a signif icantly greater probability of leading to reductions in organizational control than do those that are resolved internally. Although the relationship between suppres sion of whistleblowing and organizational control did not reach statistical signifi cance, the size of the association suggests that organizations face a tradeoff between lack of responsiveness to internal reformers and loss of organizational autonomy. The multivariate statistical results show that the amount of an organization's power does not immunize it from external control. What is pivotal in determin ing organizational control is whether the whistleblower engages in effective political behavior. Effective political behavior often involves some form of issue expansion as a means for moving the whistleblower's concerns to a formal governmental or systemic agenda (Cobb and Elder 1983). According to one observer interviewed for the study: As often as tradition and inertia win, it is also true that the powers that be fall frequently .... One needs to be very strategic within the organization and get the message to key constituents. Whistleblowers can succeed if they pursue significant matters-that is, politically appropriate constituencies.

The expansion to identification groups, attention groups, and the attentive public (Cobb and Elder 1983) is the primary means for arresting control from the organization.

Conclusions In general, what can be concluded about the efficacy of whistleblowing? The results tend to reinforce some existing notions about the limitations of whistleblowing as a mechanism for organizational change. Although perfor mance was not sensitive to organizational policies, the research indicated that appeals to external forces could decrease organizational autonomy and thereby increase accountability. Overall, one must conclude that whistleblowing, under existing institutional rules, has a very limited role to play in correcting specific abuses and promoting organizational change.

One of the difficult challenges encountered in this research involved identify ing relationships between a whistleblower's actions and organizational conse quences. The ease survey methods used in the research dealt adequately with the problems of quantifying and comparing cases, but they did not address the cri teria for attributing causality to the whistleblower's actions. Attributing causality is a risky endeavor even when available data are less constrained than they were WHISTLEBLOWING 95 in this research. Future research will have to employ different strategies to over come the problem. One strategy might be to conduct a small number of prospec tive, in-depth case studies to develop more completely a theory of voice and organizational change. The in-depth case studies would be useful for observing cross-level (micro vs. macro) changes and the dynamics associated with them.

Appendix: Operational Definitions Hostility toward the whistleblower is a count of the number of "yes" responses to retaliation questions in the case surveyor self-administered survey.

Case count of affirmative responses for the following items: poor performance appraisal; denial of promotion; denial of opportunity for train ing; assigned less desirable or less important duties in current job; transfer or reassignment to a different job with less desirable duties; reassignment to a different geographic location; suspension from job; grade level demotion; fired from job.

Self-administered count of affirmative responses to "Did this occur?" for the following items: poor performance appraisal; denial of promo tion; denial of opportunity for training; assigned less desirable or less import ant duties in current job; transfer or reassignment to a different job with less desirable duties; reassignment to a different geographic location; suspension from job; grade-level demotion; fired from job.

Resolution of the issue: the extent to which the claims of the whistleblower are resolved.

Case survey-Responses to question (II). (1) resolved (2) only partially resolved; major aspects of the allegation were unresolved (3) not at all resolved .

Self-administered survey-Responses to question (20). "What ended the con troversy?" categorized in the following way: (1) felt that my concerns had been completely addressed." (2) felt that my concerns were satisfactorily resolved by compromise." (3) stopped pursuing my claim because I became discouraged that any change would occur" or stopped pursuing my claim because I ran out of resources [e.g., money and time]" or "The controversy has not ended; it is ongoing." 96 JAMES L. PERRY Organizational policy's punishment of legitimate claims refers to the extent to which the organization has responded favorably to previous whistleblower claims. An agency index was created from aggregated employee responses to an item used in the 1980 MSPB Whistleblower Survey (item 22-8) and re peated in the 1983 Merit Principles Survey (item 20-6). The item asked respondents to indicate reasons for not reporting fraud, waste, or mismanage ment. Affirmative responses to the option did not think that anything would be done to correct the activity" were taken as an indicator of this variable. The 1980 departmental score was used for all cases originating in 1982 or earlier. The 1983 departmental score was used for cases originating in 1983 or later. The department score was created by summing the responses of all respondents from the organization and dividing by the number of re spondents.

Organizational policy's reward of illegitimate claims is the extent to which the organization has acquiesced to dubious claims in past whistleblower cases. It was measured using a departmental score created from responses to the follow ing item from the 1983 Merit Principles Survey: "To what extent, if any, are the following an obstacle to taking disciplinary action against employees who do not improve their performance?" possibility of whistleblower complaint (ranging from "to no extent" to very great extent"). The department score was created by summing the response s of all respondents from the organization and dividing by the number of respondents. Organizational performance is the extent to which the organization is perceived as fulfilling its mission.

Case survey--Recorded responses to the following close-ended question:

"Did the organization's performance improve or decline as a result of the whistleblowing episode?" to the following item: "To what extent has the performance of the organization for which you worked at the time of the incident changed since you reported the incident?" WHISTLEBLOWING 97 solution equals "1" if response to ques tion (24) was "courts" or "Congress"; "0" if courts or Congress were not checked; "8" (not applicable) if respondent checked "The claim has not been resolved. " Organizational control is the extent to which the organization maintains control over major decisions.

Response to the following question in the case survey: "Did the organization lose control over any of its authority?" 98 JAMES L. PERRY Suppress internal and external whistleblowing is the extent to which the organi zation takes actions to prevent whistleblowers from pursuing their claims both inside and outside the organization. An agency index was created from aggre gated employee responses to an item used in the 1980 MSPB Whistleblower Survey (item 28) and repeated in the 1983 Merit Principles Survey (item 23). The item was: "Within the last 12 months, have you personally experienced some type of reprisal or threat of reprisal because of an activity you reported?" The 1980 departmental score was used for all cases originating in 1982 or earlier. The 1983 departmental score was used for all cases originating in 1983 or later. The department score was created by summing the responses of all respondents from the organization and dividing by the number of respondents. References Bowie, Norman. 1982. Business Ethics. Englewood Cliffs, NJ: Prentice Hall.

Cobb, R.W., and Elder, C.D. 1983. Participation in American Politics: The Dynamics of Agenda-Building. Baltimore, MD: Johns Hopkins University Press.

Devine, T.M., and Aplin, D.G. 1986. "Abuse" of Authority: The Office of the Special Counsel and Whistleblower Protection." Antioch Law Journal 4. Glazer, Myron P., and Glazer, Penina M. 1989. The Whistlehlowers: Exposing Corrup tion in Government and Industry. New York: Basic Books.

Graham, J.W. 1986. "Principled Organizational Dissent: A Theoretical Essay." Research in Organizational Behavior 8: 1-52. Harshbarger, D. 1973. "The Individual and the Social Order: Notes on the Management of Heresy and Deviance in Complex Organizations." Human Relations 26: 251-69.

Hirschman, A.O. 1970. Exit, Voice, and Loyalty: Responses to Decline in Firms. Organi . and States. Cambridge, MA: Harvard University Press.

Johnson, Roberta A., and Kraft, Michael E. 1989. Bureaucratic Whisrlehlowing and Pol icy Change. Washington, DC: United States Environmental Protection Agency.

Jos, Philip. 1990. "Whistleblowing, Accountability, and Professional Ethics." Paper de livered at the American Society for Public Administration Symposium on Public Ad ministration Research, April 7-8, Los Angeles.

Jos, Philip H.; Tompkins, Mark E.; and Hays, Steven W. 1989. "In Praise of Difficult People: A Portrait of the Committed Whistleblower." Puhlic Administrative Review WHISTLEBLOWING 99 Near, J.P., and Miceli, M.P. 1986. "Whistle-Blowers in Organizations: Dissidents or Research in Organizational Behavior 9: 321--68.

O'Day, R. 1974. "Intimidation Rituals: Reactions to Reform." Journal ofApplied Behav ioral Science 10: 373-86.

Pfeffer, J., and Salancik, G.R. 1978. The External Control ofOrganizations: A Resource Dependence Perspective. New York: Harper & Row.

Soeken, K.L., and Soeken, D.R. 1987. "A Survey of Whistle blowers: Their Stressors and Coping Strategies." Unpublished manuscript.

Staw, B.M. 1980. "Rationality and Justification in Organizational Life." Research in Organizational Behavior 2: 45-80. SlaW, B.M.; Sandelands, LE.; and Dutton, J.E. 1981. "Threat-Rigidity Effects in Organiza tional Behavior: A Multilevel Analysis." Administrative Science Quarterly 26: 501-24.

Tetlock, P. 1985. "Accountability: The Neglected Social Context of Judgement and Choice. Research in Organizational Behavior 7: 297-332.

Weinstein, D. 1979. Bureaucratic Opposition: Challenging Abuses at the Workplace. New York: Pergamon.

Westin, A.F. 1981. Whistleblowing! Loyalty and Dissent in the Corporation. New York:

McGraw-Hill.

Yin, R.K. 1981 a. "The Case Study Crisis: Some Answers." Administrative Science Quar terly 26: 58--65. ---. 1981b. "The Case Study as a Serious Research Strategy." Knowledge: Creation, Diffusion, Utilization 3: 97-114.

Yin, R., and Heald, K. 1975. "Using the Case Survey Method to Analyze Policy Stud ies." Administrative Science Quarterly, 20: 371-81.

Zald, M.N., and Berger, M.A. 1978. "Social Movements in Organizations: Coup d'Etat, Insurgency, and Mass Movements." American Journal ofSociology 83: 823--61.