M7D1 BUS 323

ETHICS PROGRAMS AND THE PARADOX OF CONTROL Jason Stansbury and Bruce Barry Abstract:

We analyze corporate ethics programs as control systems, argu- ing that how control is exercised may have pernicious consequences and be morally problematic. In particular, the control cultivated by ethics programs may weaken employees' ability and motivation to exercise their own moral judgment, especially in novel situations. We develop this argument first by examining how organizadon theorists analyze control as an instrument of management coordinadon, and by addressing the political implicadons of control. We discuss coercive and enabling control as variadons that help account for the disdnction between compliance-based ethics programs and values-based ethics programs. We then explore three potential drawbacks of ethics pro- grams: the specter of indoctrinadon, a politicizadon of ethics, and an atrophy of competence. Ethics programs that rely on coercive control may undermine their own effectiveness at stemming misbehavior.

O rganizational ethics programs have proliferated since the introduction of the Federal Sentencing Guidehnes for Organizadons (FSGO) in 1991 (Joseph, 2003).

The designers and administrators of these programs might view them as primarily intended to provide guidance to organizadonal members regarding ethical obligadons of organizadonal membership. For example, BellSouth's Office of Eth- ics and Compliance states that "Our top priority is to create an environment where BellSouth employees feel supported and protected when facing ethical situadons" (BellSouth, 2004). It is clear, however, that ethics programs also serve to dghten administrative control over a range of behaviors (Weaver, Trevifio, & Cochran, 1999b). We need not recount the growing htany of large and small corporate ethical failures in order to suggest that increased control over some risky behaviors may well serve both the firm's and the public's interests.

Control is indispensable to organizadon, because it makes cooperadon possible, and in tum enables the producdon of a range of complex goods and services. In classical management terms, it is the funcdon of the executive to redirect the activi- ties of an organization's pardcipants in such a way that cooperation yields sufficient rewards to keep the participants engaged (Bamard, 1938/2001). Notwithstanding its benefits, the manner in which control is exercised may itself have pernicious consequences. The argument we make in this paper is that forms of control culdvated by organizational ethics programs may weaken employees' ability and modvation © 2007. Business Ethics Quarterly, Volume 17, Issue 2. ISSN 1052-150X. pp. 239-261 240 BUSINESS ETHICS QUARTERLY to exercise their own moral judgment, especially in novel situadons. Accordingly, despite the apparent promise of organizadonal ethics programs as vehicles for dis- couraging business pracdces and employee behaviors that compromise integrity or threaten the collective good, it is important to examine the limitadons of such programs—the threats that they themselves pose to both the controllers and the controlled.

We organize the presentadon of our argument into four main secdons. First, we discuss how organizadon theorists analyze control as an instmment of management coordinadon of the enterprise, and we consider the pohdcal implicadons of control's applicadon. We examine two extremes of a condnuum of control—coercive and enabling (Adler & Borys, 1996)—as relevant for understanding die social effects of particular control mechanisms, as well as of ethics programs as coUecdons of control mechanisms. The second secdon provides an overview of research on the nature and consequences of these programs. Building on the disdncdon between compliance-based ethics programs and values-based ethics programs (Paine, 1994; Weaver & Trevifio, 1999), we draw connecdons between types of programs and forms of control. This sets the stage for the paper's third section—an argument regarding the potendal drawbacks of ethics programs, which we make in three parts:

first, the problem of the specter of indoctrinadon; second, the problem of polidcization of ethics; and last, the problem of an atrophy of competence.' In a concluding secdon, we discuss the prospects for enabling control to ameliorate the issues that we idendfy, and point to potendal imphcadons for the designers and administrators of ethics programs.

/. Control as a Technical and a Political Matter Control is a technical matter that has polidcal implicadons. It may be exercised along a condnuum from subde to blatant, and may independendy elicit responses from employees ranging from endorsement to rebelhon. In this secdon we define control and describe a stylized condnuum that is relevant and useful for interpredng the social consequences of control mechanisms and the corporate ethics initiatives that employ them.

Control Defined Definidons of control in contexts of management and organization generally rest on a platform of consistency and predictability. Garrison and Noreen characterize control as "the steps taken by management to ensure that the objectives set down at the planning stage are attained, and to ensure that all parts of the organizadon func- don in a manner consistent with organizadonal pohcies" (Garrison & Noreen, 1997:

360).

Lawler and Rhode note that "control systems are instituted in organizations because management and others feel they need informadon about what is going on in the organizadon so they can coordinate the activities of others" (Lawler & Rhode, 1976: 3). Perrow conceives of control along a condnuum of intmsiveness. ETHICS PROGRAMS AND THE PARADOX OF CONTROL 241 with direct controls including supervision and mles, bureaucratic controls includ- ing task standardization and division of labor, and premise controls bounding the possibilides and terms under which decisions are taken (Perrow, 1986). Perhaps because premise controls are the least obtmsive, they are the most effective at maintaining the consistency of subordinate behavior within the expectadons set by management, even in relatively unpredictable environments that are not amenable to direct or even bureaucradc control (Perrow, 1986).^ Control, in other words, is a mbric for various organizadonal mechanisms that foster consistency, predictabihty, knowledge acquisidon, and coordinadon in the pursuit of defined objecdves.

From an operational perspective, mechanisms of control may prevent accidents and other (possibly catastrophic) failures dirough discovery, monitoring, inves- dgadon, and sanctioning of problematic behaviors (Vaughan, 1990). Statistical process control is one example of such a system: production irregularities may be idendfied through inspecdons or warranty retums, tolerances subsequendy set and monitored for the relevant portion of the producdon process, and corrective acdon taken when those tolerances are exceeded, whether that means retraining or retasking an employee, sending a bad batch of parts back to a supplier, or tuning a machine. Analogous systems exist in accounting, where they not only classify and track dollars but also alert management to attempts to subvert die system. For example, the powers to approve and to execute financial transacdons are typically separated to prevent a single individual from both writing and signing checks to himself, or awarding business to a suppher that she herself owns.

The collapse of such a separadon within an organizadon is a waming sign to investors and regula- tors alike, and may prompt invesdgadons (e.g., Gullapalli, 2004).

The Politics of Control The selecdon of both means and ends is value-laden for most decision-makers (Etzioni, 1990), and therefore the controls applied to a process reflect the values of those who select it. In fact, the power of individuals or groups widiin an or- ganizadon is insdtudonahzed in part by their specificadon of what pracdces and outcomes should be measured, and how (Salancik & Pfeffer, 1977). For example, asking callers to register their sadsfacdon by touch-tone after conversing with a call center representative effectively assigns responsibility for customer sadsfacdon to call-center representatives, when in fact variadon in customer sadsfacdon may also be determined by the products and processes that cause customers to call. Simi- larly, a decision not to measure customer satisfaction at all implies that customer satisfaction is less important than cost per call or some other measure, or perhaps that a decision-maker who does not bear the cost of low customer sadsfaction has more influence in the organizadon than one who does.

The choice of control mecha- nisms represents a choice about where to place responsibility, and therefore about how (and how much of) a problem might be solved. Control is therefore at play in organizadonal polidcs, which "involves those acdvides taken within organizadons to acquire, develop, and use power and other resources to obtain one's preferred 242 BUSINESS ETHICS QUARTERLY outcomes in a situadon in which there is uncertainty or dissensus about choices" (Pfeffer, 1981: 7). Although control may be confused with mere coordinadon (Ez- zamel & Willmott, 1998), "[t]he organizadon and control of any labor process is . . . both an operadonal issue and a polidcal acdon" (Warhurst, 1998: 471).

It is worth nodng that diis polidcal interest is operadve at all levels of the or- ganization: factory operators may object to the installation of sensors that stop the producdon line if they do not operate a tool within the specified 20-second window, just as top managers may prefer a weak separation of financial duties that leaves their annual bonuses safe from the whims of the market or the decisions of outsid- ers.

Pohdcs is not the prerogadve of management alone, but rather exists wherever divergent preferences regarding allocations of finite resources vie for influence.

Idendfying polidcs as a dimension of control is not an indictment of either.

Pfeffer observes that "Once consensus is lost, once disagreements about prefer- ences, technology, and management philosophy emerge, it is very hard to restore the kind of shared perspective and sohdarity which is necessary to operate under the radonal model" (Pfeffer, 1981: 32), leading to a pohdcal mode of decision-making in which "conflict is legidmate and expected" (Pfeffer, 1981: 31). Pohdcal deci- sion-making enables the equilibradon of divergent preferences, and the exercise of control is an expression of pohdcal influence. Because the imposidon of control is a political action, its acceptance, rejection, or subversion by the controlled is also a political action. This is just as much the case for the "controlhng" elements of ethics programs as it is for administradve stmctures intended to elicit consistent or predictable behavior in the realms of accounting, production, customer service, and other funcdons.

A Continuum of Control Adler and Borys (1996) drew a disdncdon between coercive and enabhng for- malizadon in bureaucradc organizadons. Coercive formalizadon is a substitute for commitment: it designs a process to direct the acdon of employees who are assumed to be disengaged. In coercively formalized environments:

Procedures are designed to highlight to superiors whether subordinates' ac- tions are in compliance. The procedures are not designed to help subordinates determine whether the process is operating well, nor to help them navigate the inevitable contingencies of the real work process. ... From the employees' point of view, the arrangement is experienced as "a way for the higher-ups to cover their asses." (Adler & Borys, 1996: 71) Enabling procedures, on the other hand, are premised upon employee com- prehension of the process being controlled. Employees are provided with both a conceptual understanding of the process and its purposes, and real-dme informadon on its functioning, with the assumption that they will then interpret these streams of information in a way that responds effectively to contingencies. While such an approach may seem naive in organizations with high power asymmetry, in orga- nizadons where employees enjoy more informadon and influence it enables more ETHICS PROGRAMS AND THE PARADOX OF CONTROL 243 dynamic responses to novel circumstances.

It presupposes, however, a high level of goal congmence between employees and principals (Adler & Borys, 1996).

Although formalization is only one form of control, the above disdncdon be- tween enabling and coercive formalization finds a parallel in Simon's observation that "[m]odem writers on administration have emphasized the need for restraint by recommending the use when possible of other means of influence, leading to conviction, rather than authority, leading often to nothing more than acquiescence" (Simon, 1947/1997:

186).

An analogy is found also in the psychology of social in- fluence: Kelman's (1958) distinction between mere behavioral "compliance" with an influence request, and "commitment" that brings underlying atdtude change along with behavioral adjustment. It may also be useful to think of enabling versus coercive control as a continuum rather than a dichotomy, because employees are often prompted to behave in a certain way through carrot-and-sdck inducements that gamer more than acquiescence but less than conviction.

Control is a necessary condition of organization: without coordination, orga- nization is impossible (Bamard, 1938/2001), and coordination imphes a selection of goals.

Effective control stmctures the cooperation of actors at every level of an organization. It is therefore not control per se that we challenge here. Rather, we argue that a stylized descriptive continuum of control practices from enabling to coercive is useful for understanding the effects that control has upon its subjects.

Enabling forms of control, which in a sense are like Perrow's (1986) premise controls because they rely on shaping the understanding of the employee, have a number of potential advantages over coercive altematives. Following our review of current scholarship on ethics programs, we will discuss these advantages below.

//. Organizational Ethics Programs Since the adoption of the 1991 Federal Sentencing Guidehnes for Organizadons, many U.S. firms have instituted corporate ethics programs (Joseph, 2003) to hmit their exposure to the regulatory fines, employee or activist lawsuits, and negadve publicity that can result from ethical misbehavior. Table 1 lists the elements of an effective ethics program according to FSGO.

Notably, an update to FSGO that took effect in November 2004 also stipulates that organizations "otherwise promote an organizational culture that encourages ethical conduct and a commitment to compliance with the law" (United States Sentencing Commission [USSC], 2004, Section 8.B2.1(a)(l)). Other updated stipulations are also consistent with a new emphasis on organizational culture, including a provision that "the organizadon shall hire and promote individuals so as to ensure ... the promotion of an organi- zational culture that encourages a commitment to compliance with ethics and the law" (USSC, 2004, Section 8.B2.1(b)(3), Application Note 4(B)), an extension of the training requirement to the goveming authority (i.e., board of directors) and top management, an insistence that intemal control systems enable employees to "seek guidance regarding potendal or actual criminal conduct" (USSC, 2004, 244 BUSINESS ETHICS QUARTERLY Secdon 8.B2.1(b)(5)), and a recommendadon that employees be given "appropri- ate incendves to perform in accordance with the comphance and ethics program" (USSC, 2004, Secdon 8.B2.1(b)(6)). Programs that comply widi die pracdces in the guidehnes are eligible for reducdons of up to 95 percent in the fines that the Federal Govemment will seek (Desio, 2004).^ TABLE 1 Elements of a Corporate Ethics Program Under Federal Sentencing Guidelines According to the U.S.

Federal Sentencing Guidelines for Organizations (FSGO), an effective ethics program as defined by the guidelines will have these elements:

Standards and procedures to prevent and detect criminal conduct.

Responsibility at all levels and adequate resources, and authority for the program.

Personnel screening related to program goals.

Training at all levels.

Auditing, monitoring, and evaluating program effectiveness.

Non-retaliatory intemal reporting systems.

Incentives and discipline to promote compliance.

Reasonable steps to respond to and prevent further similar offenses upon detection of a violation.

Source:

Desio 2004:

The scope of acdvides that define an ethics program can be understood broadly.

According to the 2003 Nadonal Business Ethics Survey (NBES) conducted by the Ethics Resource Center (a nonprofit insdtute dedicated to the study and promodon of organizational ethics), there are four key aspects to corporate ethics programs:

written standards of conduct, ethics training, ethics advice hnes/offices, and systems for anonymous reporting of misconduct. In the NBES, a survey of 1500 working individuals across the U.S., only 12 percent of respondents said that their employer had none of these four pracdces in place (Joseph, 2003). Weaver, Trevifio, and Co- chran (1999a) surveyed institudonal ethics pracdces among the Fortune 1000 firms in 1994, with items on the prevalence of ethics-oriented pohcy statements (including but not Umited to codes of conduct), the frequency with which they were revised, and the frequency with which employees were asked to confirm their receipt of and compliance with those statements; the formalizadon of management responsibili- des for ethics; the creadon of freestanding ethics offices with dedicated personnel; use of extemal evaluadons or benchmarking of ethical performance; creation of procedures for handling ethical incidents; maintenance of telephone repordng and advice systems; frequency and form of top-management attendon to ethics; and communicadon of and training in ethics standards. They found that although eth- ics-oriented policy statements were widespread, other institudonal supports for the promodon of ethical behavior were generally weak (Weaver et al., 1999a). In an early review of the literature on institutional ethics pracdces, Metzger, Dalton, and Hill (1993) noted the importance of codes of conduct that root their instmcdons in more compelling purposes than legal necessity or corporate reputadon, elaborate on employees' ethical obhgadons to consdtuencies extemal to the firm, and clearly prioridze the many goods that they espouse. Metzger, Dalton, and Hill (1993) also emphasized the necessity of leaders welcoming the bearers of bad news, contract- ing for outside audits of ethical performance in order to encourage candor, and ETHICS PROGRAMS AND THE PARADOX OF CONTROL 245 publicly rewarding strong performance on measures of safety, poUudon, and other social goods while publicly punishing weak performances on these metrics despite strong economic results.

Compliance-Based Versus Values-Based Programs Many ethics managers have been moving their activides toward what have been called values-based or integrity-based approaches (Weaver & Trevifio, 1999).

Whereas compliance-based approaches emphasize detection and disciphne of mle violations, values-based approaches emphasize the development of shared prin- ciples regarding ethical behavior (Weaver & Trevifio, 1999; Weaver et al., 1999b).

Paine notes:

Even in the best cases, legal compliance is unlikely to unleash much moral imagination or commitment. ... [But] an integrity strategy .. . requires an active effort to define the responsibilities and aspirations that constitute an organization's ethical compass. (Paine, 1994: 111) For example, a firm that publishes a concise statement of ethical aspiradons, together with anecdotes of exemplary employee behavior, takes a values-based approach.

The Johnson & Johnson Credo, originally penned by Robert Wood Johnson, is an example of such an aspiradonal statement. Its well-known apphcadon in the 1982 Tylenol poisoning crisis is a prime example of the values-based approach in practice (Paine, 1994), and Johnson & Johnson upholds the Tylenol crisis as the finest example of its values at work (Johnson & Johnson, 2004). Values-based approaches reinforce the sahence of ethical roles in the workplace, increase per- ceived organizadonal support among employees (Weaver & Trevifio, 1999), and enhance the sense of fair treatment by the organization among employees (Weaver & Trevifio, 2001). Importandy, these approaches are not mutually exclusive:

effec- tive ethics programs should combine accountability for violations of agreed-upon norms (i.e., compliance) with a strong values focus (Weaver & Trevifio, 1999).

This is necessary in part to maintain a sense of faimess among employees, who expect transgressions to be punished regardless of the status of the transgressor (Weaver & Trevifio, 2001).

The Efficacy of Ethics Programs Evidence of the impact of ethics programs on employee behavior is mixed, with most of the empirical studies to date having focused on the most widespread program element, codes of conduct. The NBES found that employees were twice as likely to claim to have reported observed violations of the law or of ethical principles in organizadons that had four major ethics program elements in place (written stan- dards of conduct, an anonymous helpline for ethical advice, anonymous reporting of misconduct, and training) than in organizations that had none (Joseph, 2003).

Employees in organizations with a code of conduct but no other program elements claimed to have reported observed misconduct 52 percent of the time, compared 246 BUSINESS ETHICS QUARTERLY with 67 percent in organizadons with a code and one or two of the other program elements hsted above, and 78 percent in organizadons with all four program elements (Joseph, 2003)."* However, in a survey of business students, respondents who read a sample code of conduct and then responded to seven ethical problems did not sig- nificantly differ in their answers from those who were told only that an unseen code of conduct would apply (Clark & Leonard, 1998).

This finding raises the possibility that general percepdons of an ethics program's broad orientadon towards goals and values matter more than specific program characterisdcs (Trevifio, Weaver, Gibson, & Toffler, 1999). Rather, leadership diat talks openly about ethics and visibly acts consistently with their rhetoric has a positive effect on such outcomes as awareness of ethical issues, willingness to ask for advice, and reporting of issues.

A culture of obedience or a perception that the compliance program exists to diffuse blame from top management each have a negative effect (Trevifio et al., 1999).

Empirical studies indicate that values-based approaches influence a number of organizational ethics outcomes. Employee judgments that an ethics program is val- ues-based are associated with a variety of beneficial outcomes, including employees' awareness of ethics at work, their integrity, their willingness to seek ethical advice, lower observed unethical conduct, commitment to the organization, employees' wilhngness to deliver bad news to supervisors, and percepdons that better decisions are made because of the program (Weaver & Trevifio, 1999). Similarly, researchers found that collegiate and corporate codes of conduct provide an explicit guide to a community's expectations of its members (McCabe, Trevifio, & Butterfield, 1996).

Codes of conduct lead to lower levels of self-reported unethical behavior when they reflect the informal norms of the organizadon, are clearly understood by employees, and guide everyday decision-making (McCabe et al., 1996). Finally, Makkai and Braithwaite (1994) found that nursing-home inspectors who set clear expectations not only for improvement but also for the operators' willingness and ability to make the needed improvements realized higher rates of compliance improvement than those who took an adversarial and punitive approach or those who downplayed the seriousness of the offenses; although this study pertained to audits of independent organizations' compliance with a govemment regulatory code, its parallel to inves- tigative funcdons within organizadons is clear.

Its findings also parallel Weaver and Trevifio's insistence that values-oriented programs should retain some compliance elements (Weaver & Trevifio, 1999, 2001).

To summarize, organizational ethics programs are often insdtuted out of concem for legal compliance, and provision for the detection and punishment of transgres- sions and the control of employee behavior is associated with higher levels of ethical awareness, repordng of misconduct, and employee integrity, and lower levels of observed misconduct and perceived role conflict (Weaver & Trevifio, 1999).

However, these effects are stronger for programs that also have a values-based ori- entadon, that is, that counsel employees, support employee goals and aspirations, encourage shared values, help individuals make decisions, and evaluate performance in light of company values (Weaver & Trevifio, 1999). ETHICS PROGRAMS AND THE PARADOX OF CONTROL 247 ///. Ethics Programs as Control Systems Organizational ethics programs are collections of control mechanisms, that is, ways of "standardizing employee behavior within the domains of ethics and legal compliance" (Weaver et al., 1999b: 42). These programs are often thought of by scholars and practitioners as coercing compliance with behavioral prescriptions or gamering commitment to shared values, which parallels the values- or integrity- orientation versus compliance orientation in the control literature (Weaver & Trevifio, 1999).

We also note that especially with regard to ethical decisions, enabhng control parallels the values-orientation because the context about which it informs the decision-maker is the set of values and goals that a given action should strive to meet. However, even values-based programs expect employees to comply with legal and ethical standards, and the combined effect of the two approaches parallels the combination of coercive and enabling forms of control in other contexts.

We believe that the control acdvities that comprise ethics programs can be use- fully ranged along the enabling/coercive continuum, and that a stronger values- or integrity-orientation generally corresponds to more enabling control. Importantly, the activities themselves can take on a more enabhng or coercive tone:

a telephone reporting and advice system may be labeled an "ethics helpline" or a "compliance hotline" (Weaver et al., 1999a), and even auditors may choose between a punitive approach that labels violators as deviant and a coaching tone that sets specific improvement goals (Makkai & Braithwaite, 1994). An insistence on the public punishment of offenders (Metzger et al., 1993) can be merely coercive if it illus- trates mainly the power of the enforcement authority, or reladvely more enabling if it reinforces a sense of faimess among employees (Weaver & Trevifio, 2001).

Reintegrative shaming (Braithwaite, 1989) is another form of enabling control in which the organizadon's norms are clarified and reinforced for other employees through vicarious leaming but the sancdoned employee is subsequently and fully reinstated to the community in good standing. However, although a values orientadon is necessary for enabling control, it is not sufficient: if the firm outlines its values without prioritizing them or otherwise providing guidance for acdng on them, then managers may be expected to favor those goods for which they are rewarded, such as contribudon margin or retum on assets (Metzger et al., 1993).

An ethics program can therefore be characterized according to its use of enabhng and coercive control mechanisms, as illustrated by Figure 1. Makkai and Braith- waite (1994) found that nursing-home inspectors who combined accountability for specific violadons with coaching and goal-setting for facility managers realized higher increases in compliance than those who imposed little accountability or offered litde coaching, suggesdng that the combinadon of enabling and coercive mechanisms influences overall control effectiveness. Weaver and Trevifio (1999) and Trevifio, Weaver, Gibson, and Toffler (1999) also found that a combination of the orientadons assessed at a program-wide level is more effecdve than either orientation alone. 248 BUSINESS ETHICS QUARTERLY Values-Based I 'c o U Low Anomic Low Both Values- and Compliance-Based Compliance-Oriented High Coercive Control Figure 1:

Typology of Control Environments In light of the constmctive organizational and personal outcomes that empirical researchers have idendfied, ethics programs clearly have salutary effects.

As coUec- dons of control mechanisms, however, they hold the potential for a set of serious, and frequently overlooked, negative outcomes:

the specter of indoctrination, the politicization of ethics, and the atrophy of competence.

We discuss each below.

The Specter of Indoctrination First, even ethics programs that rely on the adoption of shared values still un- avoidably legitimate and pursue those values that have been promulgated by their architects.

The prospect of employees uncridcally accepdng a value system by dint of the authority of their employers raises the specter of indoctrination, where the limits of permissible thought are set by authority rather than by logic (Tan, 2004).

Indoctrinadon has been characterized as a circumstance in which "the mind becomes closed on those issues which are fundamentally open, and the inability to imagine things as they are not imprisons the human being in the world of things as they are" (Laura, 1983: 45). We prefer a more concise working definition:

in- doctrination is the inculcation of a leamed unwillingness to consider the relative limitations of a system of thought, based on the authority of the teacher of that system. Laura's definition points to our key concem with the pracdce:

It stifles moral imaginadon (Werhane, 1999) and the prospect of cridcism of the system into which employees are indoctrinated. We share Werhane's insistence that there can be no fully objective position from which to formulate ethical critiques, so we disagree with critiques of indoctrinadon that decry the inculcation of values as an ETHICS PROGRAMS AND THE PARADOX OF CONTROL 249 intolerable violadon of moral autonomy, because effecdve ethical cridcism must be rooted in a self-conscious moral posidon. Rather, the loss of the opportunity to explore other value systems infringes upon the moral autonomy of the individual.

Therefore, indoctrination is dangerous not because it inculcates values, but because the closure of interpretadon and criticism infringes upon the moral autonomy of the individual (Habermas, 1990, 1993) and sdfles the potendal for collective change.

Reactance to indoctrination.

As with many specters, the reactions of the fearful are often more genuine than the threat against which they react.

The values promoted by the ethics program may not always be uncritically accepted. They may rather be uncridcally rejected, if employees perceive that their exisdng freedom of acdon or even of thought is being consd-ained.

We draw here on the classic nodon of reactance in social psychology. Brehm and Brehm (1981) found that individuals generally expect to be able to engage in various behaviors that may be more or less uniquely effecdve in instmmentally sadsfying motives of greater or lesser strength; the mo- dvation to preserve the freedom to engage in these behaviors is called reactance.

Importantly, reactance does not result if the individuals in question do not perceive both a freedom to act and their own competence to do so.

For example, instmcting customer-service employees to refer press inquiries to a corporate communications function should not be controversial if the employees do not feel competent to field such inquiries themselves. Reactance varies in its strength proportionally to the unique instmmental effecdveness of the constrained behavior and the value of the modve that it sadsfies (Brehm & Brehm, 1981):

Dropping a health-insurance carrier from the employee benefits package in a given city may prompt minor or major resistance from employees, depending on whether the city's best (or only) ferdlity specialist accepts the other insurance carriers. Freedom of action also pertains to the avoidance of negative outcomes (Brehm & Brehm, 1981; Wicklund, 1974); the stipulation that students who witness an honor code violation report it to the honor council, with its consequent administradve inconvenience and social awkwardness, is unpopular (Pershing, 2003).

A number of empirical studies have also found that persuasive statements demanding compliance from their hsteners shift listeners' at- dtudes away from the advocated posidon (Brehm, 1966; Snyder & Wicklund, 1976; Worchel & Brehm, 1970).

More authoritative inffuencers (Brehm, 1966) and more formidable barriers (Brehm & Weinraub, 1977) result in more reactance. Finally, in an analysis of workplace influence attempts that resulted in commitment, com- pliance, and resistance, Falbe and Yukl (1992) found that "hard" influence tactics like pressure, coalitions, and legidmadng (explaining why a request is legitimate) were less successful alone or in combination than "soft" tactics like consultation, inspiradonal appeals, or ingratiation.

The implications of reactance and related forms of resistance to inffuence for ethics programs are several. Coercive control mechanisms can be expected to elicit substandal reactance, especially for sdpulations that entail a persistent threat to the exercise of valued freedoms (Brehm & Brehm, 1981). For example, a one-time managerial request to report a thief may gamer greater comphance than a sdpuladon 250 BUSINESS ETHICS QUARTERLY that employees always report coworker theft, because the latter implies a long-term demand that workers bear the social awkwardness of snitching on each other.

Enabling control mechanisms may also arouse reactance. If the values that the ethics program promotes are similar to those that employees already hold, then the perceived loss of freedom may be minimal. If they are controversial (as racial and gender equality still are in some quarters) then reactance may be greater, because of the perceived intmsion upon employees' freedom of thought. Values whose impact pertains to the work environment alone are likely to ehcit less reactance than those that also impact employees' hves outside work: insisdng that concem for employees' health extends to smoking or drinking alcohol outside work may occasion substantial reactance.

The possibility of reactance may or may not become actual resistance, which in tum may or may not deter the program's designers.

Some amount of resistance may be expected, for instance, when a firm implements grievance procedures for vicdms of harassment. Assuming that the firm remains committed to its goal of reducing harassment, it may also reduce reactance by reiterating that it is unconcemed with employees' behavior outside the workplace (thereby reducing the perceived threat to freedom of action), or by framing the program as grandng freedom to stem objecdonable behavior rather than preaching about the evils of sexism (thereby reducing the threat to employees' freedom of thought).

Note also that the actions we cite as potendally reducing reactance can be constmed as enabling control: by clarifying the circumscribed intent and extent of a grievance mle, a firm enables employees to understand its scope of applicability.

Indoctrination and the threat to moral imagination. Having explored the potendal for employees to react against potential threats to their freedom of thought or action, we can now consider the possibility that employees may adopt the val- ues promoted by an ethics program. These values may pertain to an insdtudonal context that has no parallel in employees' existing value systems; for instance, the avoidance of conflicts of interest may seem both novel and reasonable for many people who have never considered the implications of working for one firm while invesdng in others. These values may also supersede an earlier set within the same domain: gender equality in the workplace may supplant prior understandings of the "proper" roles for women and men. Both of these examples seem uncontro- versially beneficial; fortunately, practical examples of the theoretical problem of successful indoctrination by an ethics program are relatively rare, for reasons that we will explain below.

Theoretically, organizations have "thick moralities," or pattems of moral inter- pretadon that are shared by organizadon members and that govem responses to problems that are relatively common in the domain of the organization (Sonenshein, 2005; Walzer, 1994). These are generally not products of systematic ethical rea- soning, but rather accme through the organizations' own histories and vemaculars (Sonenshein, 2005). "Thin moralities" based on ethical principles independent of the organization can be used to inform ethical decision-making (Sonenshein, 2005; ETHICS PROGRAMS AND THE PARADOX OF CONTROL 251 Werhane, 1999), and ethics programs often represent an importadon or institution- alization of thin principles that can inform thick moralities.

Thick moralities often leave much to be desired:

Hamilton and Berken (2005) document Exxon's decision to legally dump noxious oilfield waste at Grand Bois, Louisiana, in 1994, when Exxon's cost-consciousness combined with an unprincipled wilhngness to ignore the known harms imposed on the residents of Grand Bois, and conclude that the Exxon managers who ordered and executed the dumping acted in a manner inconsistent with social expectations but consistent with those in use at Exxon. Intemal social criticism (Sonenshein, 2005), in which a member of an organization challenges inconsistencies between the organizadon's actions and its own thick morality, allows endogenous ethical reform. However, it may also be necessary at times to introduce moral standards exogenous to the organization, such as the notion that the legal disposal of hazardous waste may sdll impose unacceptable harms on others.

The risk with an ethics program is that even as it applies a principle or set of principles to the organizadon, it may also seal the organization off from exogenous principles. For instance, Friedman's often cited argument that the social responsibility of business is to make a profit, constrained only by the law and by ethical custom (Friedman, 1970), is problematic because of its very simplicity; an ethics program explicitly endorsing such a thick morality is closed to principled cridcism, and intemal critics arguing against an acdon like the Grand Bois dumping would be hard pressed to explain why such a legal action should be avoided.

Practically, although the prospect of an ethics program sealing off an orga- nizadon's thick morality from external criticism is interesting, ethics programs generally represent attempts to introduce a greater measure of principle into the thick morality of the organizadon. Because codes of conduct in particular are often very public statements of an organizadon's moral aspirations that are valuable in part as public-relations tools (Frankel, 1989), it seems unlikely that an organization would establish an ethics program with a self-contained and exclusionary feel, lest it become an easy target for principled critics outside the organizadon. Howsoever closed and in need of cridque an organization's thick morality may be, to the extent that the ethics program engages that thick morality rather than simply legitimadng it to outsiders, it holds out promise for being part of the solution rather than part of the problem.

The Politicization of Ethics Ethics programs have the potential to become not only the subjects of political conflict, but also the instmments of it, thereby inviting their own subversion. A Boeing employee told then-CEO Harry Stonecipher in 2004 "it really aggravates the hell out of me that we're all signing this code of conduct all around here when all the double is at the top of the company" (Lunsford & Pasztor, 2004:

Al).

Adler and Borys found that coercive formalization was sometimes used to displace culpability for failures onto the party govemed by the mles; one senior manager in their study intervened to dismpt such a trend, saying "We have to ensure that these procedures 252 BUSINESS ETHICS QUARTERLY are used as tools, not weapons" (Adler & Borys, 1996: 76). A survey of 10,000 employees of six firms found that when ethics programs are seen by employees as mechanisms for deflecting blame from top management, then employee commit- ment, intention to report misconduct, and awareness of ethical issues all decreased, while frequency of observed misconduct increased (Trevifio et al., 1999: 139).

There are two accounts of pohdcs in die literature. One account ascribes a func- donal role to polidcs, which is characterized as a system of informal social influences (Drory & Romm, 1988) that equilibrate interests within the organizadon (Pfeffer, 1981), align power with pardes who control cridcal resources when formal channels have all been brought under the control of a vested interest (Salancik & Pfeffer, 1977), and somedmes hasten the death or repurposing of an obsolete or purposeless organization (Mintzberg, 1985).

However, pohdcs have no inherent commitment to normadvely praiseworthy purposes; they may arise simply because one actor seeks a new deal (Mintzberg, 1985).

Another account characterizes polidcs as a percepdon that illegidmate and self-interested influence is dysfuncdonally exercised to steer decisions in the interests of the powerful (Cropanzano, Howes, Grandey, & Toth, 1997); this perception is held mostly by employees lacking in power (Ferris & Kac- mar, 1992; Parker, Dipboye, & Jackson, 1995), and resuldng mosdy in diminished job satisfacdon (Ferris & Kacmar, 1992) and withdrawal behaviors (Cropanzano et al., 1997) including curiailed innovation (Parker et al., 1995).

It is important to incorporate both characterizations of politics in a discussion of their implications for ethics programs. If there is disagreement over the pur- pose and balance of power in the firm, then the ethics program is both important symbolic high ground for those who would reframe the firm's ethos and a vantage point from which to assail the opposition. The ethics program therefore cannot theoredcally be set aside as sacred ground: If the political process funcdonally enables the organizadon to adapt to changing purposes and differing conceptions of the good when agreement can no longer be reached among an organization's members (Pfeffer, 1981; Mintzberg, 1985), then the ethics program must itself be adaptable. Mintzberg suggests that organizations experiencing extended political conffict eventually will either realize a resolution or cease to exist, driven out of business by the expense of the conflict and their failure to identify an acceptable purpose or ethos to justify condnued operadon.

However, if the ethics program is to become a political staging area itself, then the consequences of its use as coercive versus enabling control are important. Adler and Borys (1996) contend that conflict and power asymmetry in organizadons not only make coercive formahzadon more likely, they also make enabhng formalizadon seem naive. However, because coercion is likely to increase the perceived power- lessness of its subjects and the perception of politics within the organization, it is also hkely to catalyze reactance, withdrawal, or even resistance behaviors. Such behaviors may be legitimated by the percepdon that the ethics program narrowly serves the interests of the organizadon's formal authorities (Morrill, Zald, & Rao, 2003).

Consider, for instance, the institudon of coercive control without enabling ETHICS PROGRAMS AND THE PARADOX OF CONTROL 253 control to create a compUance-oriented ethics program, such as the introducdon of a comphance reporting hotline without an ethics advice helpline.

This would seem to enact Mintzberg's "counterinsurgency" game (Mintzberg, 1985), in which sancdons are apphed to organizadonal deviants acdng to undermine legidmate authorides, thereby inviting the condnuadon of an insurgency against the interests of the formal audiorides (Mintzberg, 1985; Morrill et al., 2003). Enabling cond-ol, on the odier hand, reduces ambiguity and therefore the potendal to perceive political conflict.

Although a radical disjuncture of interests may somedmes make coercive control and even the insurgency/counterinsurgency games between the formal authorides and another base of power inevitable, enabling control may often offer the oppor- tunity to avoid a damaging subversion of the program's goals.

The Atrophy of Competence We describe here a potendal (tragic) irony of ethics programs viewed through the lens of control with respect to an employee's moral agency: the risk that highly coercive programs will diminish rather than enhance the individual's ability to man- age ethical ambiguity. This risk seems minimal, or at least less salient, in contexts where ethical hazards are relatively stadc and prototypical, and hence predictable.

This risk is authendc, however, in setdngs where ethical decision making emerges from complex and uncertain organizadonal processes.

Braverman (1974) suggests that cond-ol that removes the details of the labor process from the discredon of the employee results in a twofold atrophy of skill: first, in the separadon of thinking from execudon, and second, in the division of skilled tasks into unskilled subtasks.

Both reduce the employee's abihty to adapt the labor process without the explicit coordinadon of management. Adler and Borys (1996) suggest that coercive pracdces diminish the ability of employees to respond to novel situadons, whereas enabhng pracdces help employees to form mental models of the process that they can apply when it breaks down.

In the interest of simplicity, many codes of conduct outline a simple commit- ment to the fair treatment of employees (for example) and instmct employees to contact the legal department if they have concems about the ethicality of a pardcular practice. Such an approach is appeahng because it reassigns decisions, and account- ability for them, to a relatively small number of the firm's resident experts (Simon, 1947/1997). However, such a division of labor may be unrealisdc.

Many employees may be reluctant to refer a decision to or seek input from a party outside of their own group because of the effort and delay involved, as well as the paper trail that may attract further unwanted attendon; we suspect that entreades to refer difficult quesdons to legal counsel are roudnely ignored. Conversely, employees who rehably refer difficult moral problems to higher audiorides for resoludon may not develop their own faculdes of moral judgment. That atrophy of competence may leave them unprepared to solve problems for which there is no dme for a referral. These effects may be partially remedied by creadng anonymous helphnes that employees can call for advice (Joseph, 2003), or instmcdng the legal department or ediics office 254 BUSINESS ETHICS QUARTERLY to advise decision makers on the moral implications of their acdons rather than relieving them of the issue.

Both opdons develop employees' own moral discredon widiout incurring substandal delays or unwanted attendon.

A separate problem is the moral hazard created by tossing "hard cases" to experts for resoludon. If difficult problems are the responsibihty of senior or specialized managers, then some managers may perceive an opportunity to ignore a given prob- lem rather than take responsibility for its resoludon, in the expectadon that when it becomes severe enough to demand attention it will also be difficult enough to warrant treatment by specialists. Jackall's industrial ethnography provides several powerful examples of what he characterizes as a determinadon not to make difficult decisions, because doing so exposes the decision-maker to blame.

In one case, nec- essary but expensive maintenance for an industrial battery was delayed for several years undl it predictably failed, creadng massive producdon dismpdons, exposing the firm to environmental liabihty, and requiring cosdy repairs (Jackall, 1988).

However, the necessary expenditures after the fact were uncontroversial decisions for the firm's general counsel, and environmental health and safety staff, whereas taking producdon offline beforehand would have been a controversial decision for plant managers who hoped to be promoted or transferred before the consequences of their negligence became apparent (Jackall, 1988).

Finally, if the organization formalizes ethical decision-making in such a way that the identity of the decision-maker is irrelevant for the content of the decision (Weber, 1946), whether through a division of labor that removes decisions from the local context or through coercive formalizadon, then the prospect of moral imagi- nadon is eliminated, because a decision maker cannot bring his or her "thin self (Werhane, 1999) to bear on a problem for which the self is officially irrelevant. If moral sensidvity requires empathy and the ability to imagine altemadve courses of acdon (Rest, 1994), both of which are products of moral imaginadon, dien die loss of moral imaginadon may be fatal to the development of moral sensidvity beyond the requirements specified in the bureaucradc role.

Although it could be argued that such a loss is a simplification necessary for efficient operadons (which is, after all, the purpose of depersonahzed bureaucracy), ethical pracdce within the organizadon may not be able to adapt to changing operadonal problems or social expectadons without moral imaginadon or intemal social cridcism. Instead, the locus of ediics policy and administradon (such as a central ethics funcdon) would have to andci- pate or idendfy every relevant trend and formalize a response to it; this would be a formidable problem of environmental scanning when activist movements or changes in employee demographics or govemment reguladons happen at a local level.

We contend that the formalizadon that an ethics program may introduce should, in- sofar as possible, be designed to enable managers to make appropriate ethical decisions with reladve autonomy.

Although opportunides to obtain advice should be available when necessary, we would caudon that encouraging employees to refer controversial or difficult decisions to experts rather than enabling them to make appropriate choices themselves can lead to an atrophy of competence in ethical decision-making. ETHICS PROGRAMS AND THE PARADOX OF CONTROL 255 IV. Conclusion Enabling and coercive control mechanisms both have important effects on the three outcomes that we have discussed above. Reactance to the specter of indoctri- nadon increases with the applicadon of either enabling or coercive controls: only reductions in the scope of control reduce reactance.

Reactance to polidcs increases with coercive controls but decreases with enabling controls, which commends the use of the latter mechanism alone or in combinadon with coercive controls.

It should be noted that reactance mediates between control and actual resistance; we do not argue that control leads directly and inevitably to resistance on the part of employees. Finally, coercive controls increase the atrophy of competence but enabling controls reduce it. We summarize these reladonships in the form of a causal schematic in Figure 2.

Conflicting Interests between Management d l / Enabling Control Management \ + and Employees V Reactance to Indoctrination Reactance to Politics Resistance Coercive Control —^^^^ Atrophy of Competence Figure 2:

Effects of Enabling and Coercive Controls Theoretical Implications In discussing potendal drawbacks to the imposidon of ethical control, we have emphasized that the risks are more serious for coercive than for enabling forms of control, and for environments with greater degrees of ambiguity. Because ethics programs are important for reducing misconduct in organizadons, we believe that it is important to mininnize the reactance or closure of consideradon that the imposi- don of values may generate, prevent these programs' subversion as instmments of polidcal intrigue, and avoid the atrophy of competence that they may cause.

We also believe that the importadon of the constmct of reactance into the hterature on ethics programs is important for future inquiries, as is the application of the concepts of moral imaginadon and intemal social cridcism to ethics programs.

There is, however, one important counterargument to our approach that deserves attendon. Enabling control supposes that employees can be helped to understand the system within which they operate, and that they will accordingly act with dis- cemment. However, as Perrow (1986) notes, some decisions are sufficiently routine that the exercise of discemment is wasteful, and some employees lack the abihty or motivadon to exercise discemment. Govemments and religious organizadons sometimes coercively impose moral values on their constituents, often with appar- ent success.' Moreover, Barker's (1993) study of self-managing work teams in an 256 BUSINESS ETHICS QUARTERLY electronics factory found that group values originally subject to egalitarian interpre- tadon became formal mles that conferred power on die employees who were there at the mles' creation; the mles eventually supplanted discussion and interpretation, and a formal hierarchy formed among the workers to support the mle system. The coercive element of control that was imposed to supplement values whose meaning was unclear eventually overwhelmed die enabling elements of the system. By diis line of reasoning, coercive control may be not just necessary, but inevitable.

We disagree.

Values-oriented ethics programs that incorporate compliance provi- sions are prime examples of enabling control, in that they provide opportunides for employees to glean an understanding of the principles behind the control.

Employees who cannot or will not engage the program's values must necessarily respond to its compliance provisions. Furthermore, even environments with apparendy roudne ethical requirements may benefit from enabhng control, in that changes to social expectadons (e.g., polludon standards or racial equality) or localized crises (e.g., a power outage or hurricane) may dismpt the system's roudne before formal ad- justments can be made to procedures. Retuming to Perrow's (1986) formuladon, maintaining enabling cond-ol entails retaining premise control instead of bureau- cradc or direct control wherever possible. The factory in Barker's (1993) example transferred veteran employees' shared understanding of their workplace values to new employees by creadng mles, a pracdce that itself may have been a product of prior experience with being managed rather than an inevitability. It seems possible to insist on teaching organizadon members the premises that govem the workplace, instead of creadng a burgeoning system of mles and exceptions; such training is more time-consuming than mlemaking, but preserves the enabling capacity of the system. Although govemments may need to place a greater emphasis on coercive control than private organizadons because they must govem people and organizadons whose values may be irreconcilable, we beheve that where a working agreement on values is possible, enabling control is preferred.

Overall, our concem is that ethics programs, like bureaucracy in general, tend to undermine their own effectiveness because they insdtutionalize control and thereby risk polidcizadon, indoctrination, and an atrophy of competence. By preserving as much individual discredon as possible through enabling control, realized through a balance of values- and compliance-oriented program elements, ethics programs may inhibit their own subversion and maximize their relevance to changing circumstances.

Implications for Practice Ethics programs are useful for reinforcing and elevadng standards of business conduct. However, the forms of control they apply may hmit their own effecdveness.

Accordingly, we see pracdcal imphcadons for designers and administrators of ethics programs that will avoid some of these programs' more serious limitadons.

First, because both enabhng and coercive controls entail reactance to the specter of indoctrinadon, and potendally actual resistance, program administrators should ETHICS PROGRAMS AND THE PARADOX OF CONTROL 257 be careful to circumscribe the scope of their control mechanisms, and especially the scope of their rhetoric.

Although high-minded declaradons of right and purpose may put pardcular pohcies or sdpuladons into a normadve context, and may be inspiring to many employees, the more controversial those declaradons and the more far-reaching those pohcies, the more reactance they risk ehcidng fi'om other employees.

In some cases, program administrators would be pmdent to clarify the Hmits of controversial controls, and to expect and manage reactance to ambidous control efforts.

Moreover, although self-contained moral jusdficadons like Friedman's argument for the social responsibility of firms to their shareholders make elegant foundadons for ethics programs, they should be avoided.

Attempts to build a cult-like organizadonal culture that are advocated in the popular management literature (e.g., Colhns & Por- ras, 2002) are similarly problemadc.

Rather, in order to fulfill their public legidmadon funcdon (Frankel, 1989) and to remain adaptable to novel problems and changing expectadons, ethics programs should be open to regular criticism from employees and the public. We recommend that ethics programs be reviewed and revised at regular intervals by a panel of experts intemal and extemal to the firm, incorporating anonymous feedback and interviews with employees and other stakeholders in order to capture these contributors' moral imaginadon (Werhane, 1999).

Second, ethics programs cannot reahsdcally be declared, or assumed to be, inviolable and off-hmits to management pohdcs, because they reflect the purposes and values of the firm which may themselves change. However, a coercive orienta- tion that places responsibility upon employees without enabling them to understand and effectively carry out their responsibilides risks becoming not only an object but also an instmment of political intrigue.

Imposing mles upon employees without clarifying the purposes of those mles, or their priority in the event of a conflict, may invite rebelhon against the program or its provisions. We recommend that ethics programs promulgate a set of foundadonal principles (akin to a "credo") describing the firm's purposes and values along with a code of conduct, and that provisions within the code of conduct be exphcitly linked to these principles so that they may be contextuahzed and even critiqued within a set of overarching values.

The credo, of course, must also prioridze the purposes and values that the firm seeks; a credo that does not distinguish between the protecdon of animal rights and the preservadon of customer safety offers litde guidance to product-safety testers (for example), and one that does not mendon profitability or shareholder value may be dismissed by many employees as mere window-dressing. Furthermore, the regular cridques and revisions recommended above may also be used to idendfy inconsistencies in the code's provisions, or contradictions in the values espoused by the credo and those enacted by top management, thereby reducing the risk that the program is seen as a top-management protection program.

Finally, the architects of ethics programs should not assume that difficult ethical issues will be reliably identified and referred to experts for resolution.

Rather, advi- sory mechanisms like advice hnes (i.e., anonymous helplines) or training sessions should coach employees on the moral implications of the decisions they face and 258 BUSINESS ETHICS QUARTERLY inform them of the values that apply, so that individuals can retain effecdve respon- sibility for their decisions. Moreover, employees should be encouraged to bring their moral imaginadon to decisions, so that they are not constrained by artificial imagined constraints on the moral scope of their role (Bird & Waters, 1989).

To conclude, ethics programs that rely on coercive control pose serious threats to their own effecdveness. Enabling control mechanisms offer a less perilous approach to expanding ethical awareness, reporting of misconduct, and employee integrity.

Clear linkages between a firm's stated values and its compliance pohcies, suitable consultadve mechanisms, and regular program cridcism and revision can help to mitigate problems of overreach, closure, and polidcal intrigue.

Notes We benefited from the comments we received at the 2005 Academy of Management meeting, and would like to thank Gary Weaver, Bart Victor, and three anonymous reviewers for their thoughtful and insightful comments and suggestions on earlier drafts of this manuscript.

The term "atrophy of competence" was coined by Harry Braverman (1974) to describe the loss of capacity for home production in a rich consumer economy. It also aptly describes the possibility of reduced aptitude for moral decision-making.

For instance, a public accountant faced with an inventory count that is substantially higher than a firm's records suggest may look for several ways to reconcile the discrepancy, but would be unlikely to even consider ordering the excess inventory destroyed, because doing so violates the premises that she has been hired to accurately report the inventory that exists, and that her actions should benefit the firm.

A United States Supreme Court decision in January 2005 rendered the guidelines advisory rather than mandatory. However, pending further legislation, many corporations are being advised by their legal counsel to continue abiding by the provisions of the FSGO (Fields, 2005).

No test of statistical significance was provided in the NBES report.

We are grateful to an anonymous reviewer for pointing this out.

References Adler, P. S., & Borys, B. 1996. Two Types of Bureaucracy: Enabling and Coercive.

Administrative Science Quarterly, 41:

61-89.

Barker, J. R.

1993.

Tightening the iron cage:

Concertive control in self-managing teams.

Administrative Science Quarterly, 38: 408-37.

Bamard, C. 1938/2001. The functions of the executive, 30th anniversary edidon.

Cambridge MA: Harvard University Press.

BellSouth.

2004.

About ethics, http://ethics.bellsouth.comyaboutetliics.html. Accessed December 13, 2004.

Bird, F. B., & Waters, J. A. 1989. The moral muteness of managers. Califomia Man- agement Review, 32(1): 73-89. ETHICS PROGRAMS AND THE PARADOX OF CONTROL 259 Braithwaite, J. 1989. Crime, shame, and reintegration. New York: Cambridge Uni- versity Press.

Braverman, H. 1974. Labor and monopoly capital:

The degradation of work in the twentieth century. New York:

Monthly Review Press.

Brehm, J. W. 1966. A theory of psychological reactance. New York: Academic Press.

Brehm, S. S., & Brehm, J. W. 1981. Psychological reactance: A theory of freedom and control. New York:

Academic Press.

Brehm, S. S., & Weinraub, M. 1977. Physical barriers and psychological reactance:

2-year-olds' responses to threats to freedom. Joumal of Personality and Social Psychology, 35: 830-36.

Clark, M.

A., & Leonard, S. L. 1998. Can corporate codes of ethics influence behavior?

Joumal of Business Ethics, 17:

619-31.

Collins, J., & Porras, J.

1.2002.

Built to last:

Successful habits of visionary companies.

New York: HarperBusiness Essendals.

Cropanzano, R., Howes, J. C, Grandey, A. A., & Toth, P. 1997. The reladonship of organizadonal polidcs and support to work behaviors, atdtudes, and stress.

Joumal of Organizational Behavior, 18(2): 159-80.

Desio, P.

2004.

An overview of the organizational guidelines. Washington DC:

United States Sentencing Commission. http://www.ussc.gov/TRAINING/corpover.PDF.

Accessed December 15, 2004.

Drory, A., & Ronun, T. 1988. Polidcs in organizadon and its percepdon within the organizadon. Organization Studies, 9(2): 165-79.

Etzioni, A. 1990. The moral dimension: Toward a new economics. New York: Free Press.

Ezzamel, M. & Willmott, H. 1998. Accounting for teamwork: A critical study of group-based systems of organizadonal control. Administrative Science Quarterly, 43:

358-96.

Falbe, C. M., & Yukl, G. 1992. Consequences for managers of using single infiuence tac- dcs and combinadons of tacdcs.

Academy of Management Joumal, 35: 638-52.

Ferris, G., & Kacmar, K. M. 1992. Percepdons of organizational polidcs. Joumal of Management, 18(1): 93-116.

Fields, G. 2005. Ruling on sentencing guidelines may also affect corporate crime. The Wall Street Joumal (January 17): A4.

Frankel, M. S. 1989. Professional codes: Why, how, and with what impacti Joumal of Business Ethics, 8(2,3): 109-15.

Friedman, M. 1970. The social responsibility of business is to increase its profits. The New York Times Magazine (September 13).

Garrison, R. H., & Noreen, E. W. 1997. Managerial accounting (8th ed.). Chicago:

Irwin.

Gullapalli, D. 2004. Lapses at Fannie vindicate the case for the tough mles on account- abihty.

Wall Street Joumal (Octoher 4):

Cl.

Habermas, J.

1990.

Moral consciousness and communicative action, trans.

C. Lenhardt & S. W. Nicholson. Cambridge MA: MIT Press. 260 BUSINESS ETHICS QUARTERLY _. 1993. Justification and application: Remarks on discourse ethics, trans. C.

Cronin. Cambridge MA: MIT Press.

Hamilton, J.

B., HI, & Berken, E. J.

2005.

Exxon at Grand Bois, Louisiana: A three-level analysis of management decision making and corporate conduct. Business Ethics Quarterly, 15: 385-408.

Jackall, R. 1988. Moral mazes. New York: Oxford University Press.

Johnson & Johnson. 2004. Our credo history, http://www.jnj.com/our_company/our_ credo_history/index.htm. Accessed December 15, 2004.

Joseph, J.

2003.

National business ethics survey: How employees view ethics in their organizations. Washington DC: Ethics Resource Center.

Kelman, H. C. 1958. Compliance, identification, and intemalization: Three processes of atdtude change. Joumal of Confiict Resolution, 2:

51-60.

Laura, R. S.

1983.

To educate or to indoctrinate: That is sdll the quesdon. Educational Philosophy and Theory, 15: 43-55.

Lawler, E. L., in, & Rhode, J. G. 1976. Information and control in organizations.

Pacific Palisades CA: Goodyear.

Lunsford, J. L., & Pasztor, A. 2004. Higher plane: New boss stmggles to lift Boeing above military scandals. The Wall Street Joumal (July 14), Al.

Makkai, T., & Braithwaite, J. 1994. Reintegrative shaming and compliance with regula- tory standards. Criminology, 32: 361-86.

McCabe, D. L., Trevifio, L. K., & Butterfield, K. D. 1996. The influence of collegiate and corporate codes of conduct on ethics-related behavior in the workplace. Busi- ness Ethics Quarterly, 6: 461-76.

Metzger, M., Dalton, D. R., & Hill, J. W.

1993.

The organization of ethics and the eth- ics of organizations: The case for expanded organizational ethics audits. Business Ethics Quarterly, 3:

27^3.

Mintzberg, H. 1985. The organization as polidcal arena. Joumal of Management Studies, 22(2): 133-54.

Morrill, C, Zald, M. N., & Rao, H.

2003.

Covert pohdcal conflict in organizations:

Challenges from below. Annual Review of Sociology, 29: 391^15.

Paine, L. S. 1994. Managing for organizadonal integrity. Harvard Business Review, 72(2):

106-19.

Parker, C. P., Dipboye, R. L., & Jackson, S. L. 1995. Percepdons of organizadonal politics: An investigation of antecedents and consequences. Joumal of Manage- ment, 21(5): 891-912.

Perrow, C. 1986. Complex organizations: A critical essay (3rd ed.). New York: Ran- dom House.

Pershing, J. L.

2003.

To snitch or not to snitch? Applying the concept of neutralization techniques to the enforcement of occupational misconduct. Sociological Perspec- tives, 46(2):

149-78.

Pfeffer, J.

1981.

Power in organizations. Marshfield MA: Pitman.

Rest, J. R. 1994. Theory and research. In J. R. Rest and D. Narvaez (Eds.), Moral development in the professions: Psychology and applied ethics. Hillsdale NJ:

Lawrence Erlbaum Associates. ETHICS PROGRAMS AND THE PARADOX OF CONTROL 261 Salancik, G. R., & Pfeffer, J. 1977. Who gets power—and how they hold on to it: A strategic-condngency model of power. Organizational Dynamics (Winter): 3-21.

Simon, H. A.

1947/1997.

Administrative behavior (4th ed.).

New York:

The Free Press.

Snyder, M.

L., & Wicklund, R.

1976.

Prior exercise of freedom and reactance.

Joumal of Experimental Social Psychology, 12: 120-30.

Sonenshein, S. 2005. Business ethics and intemal social criticism. Business Ethics Quarterly, 15(3): 475-98.

Tan, C. 2004. Indoctrinadon, imagination, and moral educadon. Proceedings of the 2nd Intemationai Conference on Imagination and Education: Imagination in Educational Philosophy and Research, 1-13. http://www.ierg.net/confs/2004/Pro- ceedings/Tan_Charlene.pdf.

Accessed December 14, 2005.

Trevifio, L. K., Weaver, G. R., Gibson, D. G., & Toffler, B. L. 1999. Managing ethics and legal compliance: What works and what hurts.

Califomia Management Review, 4(2):

131-51.

United States Sentencing Commission (USSC).

2004.

Federal sentencing guidelines for organizations, chapter 8, part B—remedying harm from criminal conduct, and ef- fective compliance and ethics program. http://www.ussc.gov/2004guid/8b2_l.htm.

Accessed December 15, 2004.

Vaughan, D.

1990.

Autonomy, interdependence, and social control:

NASA and the Space Shutde Challenger. Administrative Science Quarterly, 35: 225-57.

Walzer, M. 1994. Thick and thin: Moral argument at home and abroad.

Notre Dame IN: University of Notre Dame Press.

Warhurst, C.

1998.

Recognizing the possible:

The organizadon and control of a socialist labor process. Administrative Science Quarterly, 43: 470-97.

Weaver, G. R., & Trevifio, L.

1999.

Comphance and values oriented ethics programs:

Influences on employees' attitudes and behavior. Business Ethics Quarterly, 9:

315-35.

2001.

The role of human resources in ethics/comphance management: A faimess perspective. Human Resource Management Review, 11: 113-43.

Weaver, G. R., Trevifio, L. K., & Cochran, P. L. 1999a. Corporate ethics practices in the mid-1990s: An empirical study of the Fortune 1000. Joumal of Business Eth- ics, 18(3): 283-94.

1999b. Corporate ethics programs as control systems: Influences of executive commitment and environmental factors. Academy of Management Joumal, 42:

41-57.

Weber, M. 1946. "Bureaucracy." In H. H. Gerth and C. W. Mills (Eds.), Erom Max Weber:

Essays in Sociology. New York:

Oxford University Press.

Werhane, P.

1999. Moral imagination and management decision making.

New York:

Oxford University Press.

Wicklund, R. A. 1974. Ereedom and reactance. Potomac MD: Lawrence Erlbaum Associates.

Worchel, S., & Brehm, J.

1970. Effect of threats to atdtudinal freedom as a funcdon of agreement with the communicator. Joumal of Personality and Social Psychol- ogy, 14: 18-22.