Read attached article and answer the below questions

Is Formal Ethics Training Merely Cosmetic? A Study of Ethics Training and Ethical

Organizational Culture

Danielle E. Warren Rutgers, The State University of New Jersey

Joseph P. Gaspar Quinnipiac University

William S. Laufer University of Pennsylvania

ABSTRACT: U.S. Organizational Sentencing Guidelines provide firms with incentives to develop formal ethics programs to promote ethical organizational cultures and thereby decrease corporate offenses. Yet critics argue such programs are cosmetic. Here we studied bank employees before and after the introduction of formal ethics training—an important component of formal ethics programs—to examine the effects of training on ethical organizational culture. Two years after a single training session, we find sustained, positive effects on indicators of an ethical organizational culture (observed unethical behavior, intentions to behave ethically, perceptions of organizational efficacy in managing ethics, and the firm's normative structure). While espoused organizational values also rose in importance post-training, the boost dissipated after the second year which suggests perceptions of values are not driving sustained behavioral improvements. This finding conflicts with past theory which asserts that enduring behavioral improvements arise from the inculcation of organizational values. Implications for future research are discussed.

KEY WORDS: Formal ethics training, ethical organizational culture, values, ethical behavior, unethical behavior, U.S. Sentencing Guidelines

INTRODUCTION

T H E UNITED STATES SENTENCING GUIDELINES for Organizations were J . introduced in 1991 to provide guidance to federal judges in crafting sanctions

for a wide range of corporate crime cases (United States Sentencing Commission, 1991). Grounded in responsive regulatory theory, the U.S. Sentencing Guidelines provide organizations with distinct incentives to self-regulate by establishing effec- tive ethics and compliance programs that promote an ethical organizational culture

©2014 Business Ethics Quarterly 24:1 (January 2014). ISSN 1052-150X pp. 85-117 DOl: 10.5840/beq2014233

86 BUSINESS ETHICS QUARTERLY

(Hess, McWborter, & Eort, 2006; United States Sentencing Commission, 2004). Tbe importance of promoting an ethical organizational culture and tbe corresponding need for ethics programs are reflected in not only tbese guidelines, but also in tbe number, frequency, and scale of corporate scandals in tbe 1990s and early 2000s (Hess et al., 2006; Laufer, 2008). Yet, tbese scandals raise questions about a firm's ability to change its culture tbrougb tbe adoption of formal ethics programs. Cynics assume sucb programs are used to reduce fines and stave off prosecution and doubt tbe ability of formal etbics programs to affect aspects of an organization's ethical culture (Laufer, 2008; Laufer & Robertson, 1997; McKendall, DeMarr, & Jones- Rikkers, 2002; Stansbury & Barry, 2007; Weaver, Treviño, & Cocbran, 1999b).

Meanwbile, govemment prescriptions for organizational self-regulation bave motivated widespread adoption of formal etbics and compliance training programs in organizations (Treviño, Weaver, & Brown, 2008; Treviño, Weaver, & Reynolds, 2006) even thougb little causal evidence exists on tbe influence of formal ethics training programs on ethical culture (Kisb-Gephart, Harrison, & Treviño, 2010; Tenbrunsel & Smitb-Crowe, 2008; Treviño et al., 2006). One reason for the lack of empirical research is tbat data on the etbical perceptions, decisions, and behaviors of managers and otber employees are considered organizationally "sensitive" and therefore difficult to obtain once, let alone multiple times (Treviño et al., 2006). Most empirical studies on etbics programs link types of formal etbics programs (training as a component of a program) to otber organizational outcomes using cross-sectional data (Kaptein, 2009, 2011b; McKendall et al., 2002; Smitb-Crowe, Tenbrunsel, Cban-Serafin, Brief, Umphress, & Joseph, forthcoming; Treviño, Butterfield, & McCabe, 1998; Treviño & Weaver, 2001 ; Weaver & Treviño, 1999; Weaver, Treviño, &. Cocbran, 1999a, 1999c) or longitudinal panel studies tbat include different em- ployees in eacb wave (Kaptein, 2010). Tbese studies provide important insights into tbe possible positive effects of ethics programs but, at tbe same time, they offer only limited evidence of the direction of relationsbips. In cross-sectional studies, for instance, it is possible tbat tbose organizations witb etbically-oriented attitudes and behaviors adopt more elaborate etbics programs, such tbat etbics programs may not be tbe cause, but the outcome, of more ethical attitudes and behaviors. Eurther muddying our understanding of relationsbips, empirical evidence suggests training bas negative effects or no effect on etbical outcomes (Kaptein, 2011b; McKendall et al., 2002; O'Eallon & Butterfield, 2005; Sparks & Hunt, 1998). Eor instance, Kaptein's (201 lb) recent cross-sectional study of working adults suggests tbat tbe more frequently organizations engage in formal communication regarding the cor- porate codes of conduct, the more unethical bebavior is exhibited in organizations.

Motivated by an increasing number of regulatory incentives and conflicting findings in scbolarly research, we take an important step in advancing research by examining the effects of comprehensive etbics training, a form of training based upon regulatory prescriptions, on attributes of an ethical organizational culture. More specifically, we ask: Wben firms develop formal etbics training programs consistent with scholarly and governmental prescriptions, will tbe ethics training influence attributes of an organizational etbical culture over time? Here we make progress in testing tbe validity of claims tbat training is cosmetic, a criticism that

Is FORMAL ETHICS TRAINING MERELY COSMETIC? 87

conflicts with the main tenets of responsive regulatory theory, as well as theory on the effects of formal ethics programs. Furthermore, by adding a temporal component and examining attributes of organizational ethical culture separately, we are able to advance not only our understanding of causal relationships, but also the longev- ity of the differential effects of training on multiple components of organizational ethical culture.

To begin, we review the regulatory impetus promoting formal ethics programs and formal ethics training. From the regulatory and scholarly prescriptions, we present a comprehensive approach to ethics training, which theory suggests is most likely to influence organizational ethical culture. After introducing the concept of comprehensive ethics training, we rely on social cognitive theories to develop hy- potheses regarding comprehensive ethics training and specific indicators of ethical culture (e.g., behaviors, values, and beliefs). We follow with an empirical study that measures these indicators before and after training takes place in an organization. To understand the longevity of training, the indicators of an ethical organizational culture are studied initially after training (nine months after training) and at a later point (2.5 years after training). Finally, we conclude with a discussion on the im- plications of our research for extant theory and future research.

THEORETICAL BACKGROUND AND HYPOTHESES

Formal Ethics Programs

The U.S. Sentencing Guidelines and subsequent government prescriptions share a singular objective: to offer incentives and disincentives to ensure effective orga- nizational self-regulation of ethical conduct. This conceptual background of the Sentencing Guidelines is itself supported by the well-accepted theory of respon- sive regulation (Ayres & Braithwaite, 1992; Baldwin & Black, 2010). According to this theory, self-regulation is seen as optimal and only should be replaced or supplemented by additional layers of formal controls when there is evidence of firm failures. The lynchpin of the responsive regulation approach taken by the U.S. Sentencing Commission is that effective firm self-regulation will be accomplished through formal ethics programs.

Though these programs may vary in orientation across organizations, most pro- grams include such components as formal codes of ethics and business conduct, ethics cotnmittees, ethics officers, and ethics training programs (for a discussion, see Weaver et al., 1999b). The cornerstone of the more recent, amended U.S. Sentencing Guidelines is the requirement of ethics training for employees of all levels in the organization (Hess et al., 2006). Formal training programs provide a platform for organizations to communicate the standards and procedures of the organization, as well as other important aspects of the ethics and compliance programs.

The organizational outcome promoted by the prescriptions of the U.S. Sentenc- ing Commission is that of a firm culture that is expected to encourage ethical and discourage unethical behavior. The kind and character of due diligence expected by organizations, however, was left largely open. For this reason, "Organizations are now scurrying to figure out what 'culture' means" (Treviño et al., 2006: 979).

88 BUSINESS ETHICS QUARTERLY

In organizadonal researeh, eulture is refleeted in the shared beliefs, values, and behavior of the organizadon's members (Sehein, 1990, 1996). Consistent with this perspeetive, we expeet to find indieations of an ethieal eulture in both the organiza- tional members' unethieal and ethieal behaviors, as well as their beliefs about the organization and its values.

Comprehensive Ethics Training

The U.S. Sentencing Guidelines do not define the precise eontent for ethies programs and training. They do, however, make reeommendations that reasonably eorrespond to eertain eomponents of ethies training programs. For example, the U.S. Sentene- ing Guidelines suggest a dual foeus on "eomplianee and ethies" and eneourage the use of ethies training (United States Senteneing Commission, 2004). As Hess and eoUeagues (2006: 740) note, the intent of the U.S. Senteneing Commission is to push firms to abide by the "spirit of the law" and, thus, aim higher than the lowest possible standards of eomplianee or the "letter of the law."

While the U.S. Senteneing Guidelines push for formal ethies programs that em- phasize ethies and eomplianee and that utilize ehannels of eommunieation, left open for interpretation are the speeifie eharaeteristies of the formal training program and attributes of an ethieal organizational eulture. In order to make progress in examin- ing these preseriptions, a eoneeptualization of these ideals is needed. Drawing on the aeademie researeh on leaming and ethies, we review the leading thought on the orientadon of fonnal ethies programs and means for eommunieating ethies, and we integrate this researeh to develop the eoneept of comprehensive ethics training. This eoneept is meant to address govemmental preseriptions by applying insights from seholarly researeh to develop an understanding of what is thought to be the most ef- feetive means of promodng an ethieal organizational eulture. We begin by reviewing the soeial psyehologieal proeesses underpinning the use of formal ethies training.

Soeial Psyehologieal Foundation Mueh of the theory on eomprehensive ethies training, moral awareness, and ethieal behavior is grounded in soeial eognitive theories (O'Fallon & Butterfield, 2005; Treviño et al., 2006). These theories inelude Fiske and Taylor's (2008) soeial eogni- tion theory, whieh foeuses on the eognitive antecedents of ethieal behavior related to idendfying an ethieal dilemma, and Bandura's (1977) soeial learning theory, whieh foeuses on the eognitive and motivational faetors that explain an employee's deeision to exhibit ethieal or unethieal behavior. It is through these soeial eognitive perspeedves that we present theory on the role of eomprehensive ethies training in promoting an ethieal organizational eulture.

First and foremost, formal ethies training is meant to raise awareness, a eritieal eomponent of ethieal deeision making in organizadons (Jones, 1991; Rest, 1986). The work of Fiske and Taylor (2008) suggests that moral awareness is a funedon of the vividness and salienee of moral issues, as well as the eognitive aeeessibility of moral eoneepts (Jones, 1991; Reynolds, 2006). Researeh demonstrates that the aeeessibility of moral eoneepts is influeneed by not only individual differenees (Reynolds, 2008), but also organizational faetors (Butterfield, Treviño, & Weaver,

Is FORMAL ETHICS TRAINING MERELY COSMETIC? 89

2000). Similar to priming methods, which raise the accessibility of primed concepts (Fiske & Taylor, 2008), comprehensive ethics training may increase the accessibihty of morals and moral concepts. Because individuals attend more to stimuli associated with accessible concepts than inaccessible concepts (Fiske & Taylor, 2008; Reynolds, 2008), comprehensive ethics training may increase individuals' awareness of moral issues and, therefore, decrease their unethical behavior (Jones, 1991; Rest, 1986).

A related perspective complements Fiske and Taylor (2008) and focuses on the motivational dimensions of comprehensive ethics training. More specifically, Bandura's (1977) social learning perspective focuses on the importance of models of appropriate behavior and reinforcements—direct and vicarious—in understanding and predicting ethical and unethical behavior (Treviño & Youngblood, 1990). Ac- cording to Bandura's theory (1977), ethics training should entail relevant models of desired behavior and reinforcements to encourage these behaviors. Vicarious learn- ing through models, such as those described in ethics training sessions, infiuence employees' behaviors by defining their outcome expectancies (Bandura, 1977). A critical component of this theory is not only employees' attention to, but also their retention of, the behavior of the models.

Taken together, social psychological theories indicate that formal ethics training (as a form of social learning) should help employees identify moral dilemmas, offer appropriate models of ethical behavior, and emphasize reinforcements to promote the modeled behavior. With these theories in mind, we describe the latest research and best practices associated with raising moral awareness, providing ethical role models, communicating reinforcements, and promoting retention. This will serve as the basis for our description of comprehensive ethics training (for an in-depth discussion of variations in ethics training, see Sekerka, 2009).

Content Most studies related to the content of formal ethics programs make a distinction between values-oriented programs (also referred to as "ethics-oriented") and compliance-oriented programs (Treviño, Weaver, Gibson, & Toffler, 1999). In this literature, values-oriented programs are those that focus on the importance of values in the resolution of ethical dilemmas, and compliance-oriented programs are those that focus on the importance of rules and laws in the resolution of such dilemmas (Treviño et al., 1999). Importantly, it is generally accepted that values-oriented programs are more effective than compliance-oriented programs in promoting ethi- cal behavior. As Paine (1996: 480) explains, value-oriented programs focus on the adoption of "a substantive framework of general principles or values against which choices can be evaluated." By orienting employees towards specific values, such as integrity or honesty in formal ethics traitiing, employees respond to ethical dilemmas by choosing behaviors that would align with those ideals rather than simply choos- ing those behaviors that would be minimally required by law. In this view, aiming for more aspirational models of behavior is thought to be a better means to guiding employee behavior and instituting an ethical organizational culture (Paine, 1996).

However, compliance-oriented programs also offer important benefits (Weaver & Treviño, 1999). For example. Weaver and Treviño (1999) found that compliance-

90 BUSINESS ETHICS QUARTERLY

based programs predict many of the same outcomes as value-based programs, as well as interact to predict a particularly important attribute of an ethical culture: willingness to report unethical behavior to management. Furthermore, compliance programs focus on not only rules, but also reinforcements for non-compliance (Weaver & Treviño, 1999), and social learning theory suggests that realizing what punishments or rewards follow from specific behaviors is an important aspect of learning (Bandura, 1977). Compliance-oriented programs, which usually entail a review of previous legal action and sanctions, may not only deter non-compliance from employees, but also encourage employees to report non-compliance because they become aware of potential outcomes associated with such cases.

For these and related reasons, most firms rely on both orientations for their for- mal ethics programs (Weaver & Treviño, 1999; Weaver et al., 1999b). Given the promotion of a dual-emphasis on ethics and compliance in the scholarly literature and federal regulation, we include both of these dimensions in our description of a "comprehensive" approach to ethics training.

Communication Medium and Characteristics In addition to offering appropriate models and reinforcements, formal ethics training needs to address the learners' attention towards, and retention of, the communicated content (Bandura, 1977). Here we highlight two aspects of training thought to affect the learners' attention towards and retention of content: face-to-face communication and experiential learning methods.

In-person training is thought to be more effective because it is widely considered a more influential means of communicating than other mediums such as computer- mediated training. In a stream of social psychological research, face-to-face interactions are linked to interpersonal coordination and mutual positivity (Drolet & Morris, 2000), as well as cognitive and affective interpersonal trust (Rockmann & Northcraft, 2008). As McGinn and Croson (2004: 338) note, "the communication medium in which social interaction takes place affects not just the mechanical aspects of communication, but also the social aspects." In addition, the training literature more generally promotes the importance of face-to-face instruction in organizational contexts (for recent review, see Arthur, Bennett, Edens, & Bell, 2003). Thus, in the scholarly literature, face-to-face communication and discussion is regarded as par- ticularly effective for purposes of communicating ethics content (Sekerka, 2009).

One reason why face-to-face training is preferred is that it facilitates interper- sonal interaction; for this reason, the size of the training group matters as well. In face-to-face training of small groups, individuals are able to utilize experiential training approaches such as "role plays, simulations, film, skill practice, and field experiences" (Harrison, 1992: 952). Indeed, Ferrell, LeClair, and Ferrell (1998: 360) note that, "Most effective training sessions, although brief (two to four hours) focus on employee involvement and developing an interactive environment in which to discuss the issues."

Face-to-face training in small groups is thought to be particularly effective, as these sessions "provide opportunities for trainees to engage in specific behaviors, review the behaviors critically, abstract some useful insight from the analysis, and apply the

Is EoRMAL ETHICS TRAINING MERELY COSMETIC? 91

results in a practical situation" (Harrison, 1992: 952). Eurtber, tbese sessions provide opportunities for more interaction with "Uve models," which Bandura conceptual- izes as individuals demonstrating the desired behavior (Bandura & Mischel, 1965). Empirical research suggests "live models" are more effective for social learning tban simple verbal descriptions of appropriate bebavior (e.g., Bandura, Blancbard, & Ritter, 1969; Bandura & Miscbel, 1965; Wolf & Cheyne, 1972). Tbese interper- sonal, multi-bour sessions are also fundamental to perceptions tbat the program is more tban mere window-dressing or cosmetic (Laufer, 1999; Lauf er & Robertson, 1997; McCabe, Treviño, & Butterfield, 1996).

Taken togetber, prior research suggests that ethics training that includes values and compliance content (as opposed to one or tbe otber) and is delivered tbrougb face-to-face interaction (as opposed to computer-mediated) with experiential learning methods will be particularly effective in promoting an etbical organizational culture. In tbe remainder of tbis paper, we will refer to ethics training programs witb tbese characteristics as comprehensive ethics training programs. With tbis conceptualiza- tion developed, we focus on tbe theoretical relationship between comprehensive etbics training and indicators of organizational ethical culture.

Ethics Training and an Ethical Organizational Culture

Rigorous endeavors bave been made to capture a global assessment of organizational etbical culture (e.g., Kaptein, 2009, 2010; Treviño et al., 1998) and organizational ethical climate (e.g., Victor & CuUen, 1988), a more transient aspect of culture. Global assessments of ethical culture and climate are valuable for cross-sectional research of organizations because tbey allow comparisons of culture and climate across organizations. These approacbes, bowever, collapse aspects of climate and culture into an overall assessment and, tberefore, binder the development of specific linkages between tbe organizational context and dimensions of culture or climate such as specific values and behaviors. Eor tbis reason, some have questioned tbe usefulness of tbese broad assessments of ethical work climates in predicting spe- cific bebaviors (Treviño et al., 1998) and capturing important attributes of etbical culture (Key, 1999). Here, we consider the relationships between comprebensive etbics training and specific employee outcomes that comprise culture: bebaviors, perceptions, and values (Scbein, 1990, 1996). That is, rather than linking compre- hensive etbics training to a global categorization of the firm's culture, we develop specific theory on comprebensive ethics training and components of tbe culture, a focus tbat provides insight into tbe effects, if any, of training on specific attributes of etbical cultures.

Ethics Training and Behavioral Outcomes

Comprebensive ethics training may decrease unethical bebavior by raising aware- ness of moral issues and by providing models of normatively appropriate responses to etbical dilemmas as well as information on rules and laws. Providing models of appropriate behavior and information about industry rules, regulations, and laws is especially important to improving moral judgment, as research on cognitive moral

92 BUSINESS ETHICS QUARTERLY

development indicates that most adults reason at the conventional stage of devel- opment which means they rely on social conventions (social norms, rules, laws) when determining what is morally right (Kohlberg, 1981; Treviño et al., 2006). Comprehensive ethics training that provides models of ethical behavior by senior organizational members and compliance-oriented information regarding organiza- tional rules and laws, therefore, will appropriately address employees who reason at the conventional level of moral development. For those who reason at lower levels of cognitive moral development and focus on potential consequences when consider- ing appropriate behavior, ethics training further reinforces avoidance of unethical behavior by exposing employees to vicarious rewards and punishments (Ashkanasy, Windsor, & Treviño, 2006; Treviño &Youngblood, 1990). For instance, training often involves discussions of past violations, organizational outcomes, and descriptions of role models via the use of real and hypothetical ethical dilemmas (Gioia, 1992).

Thus, past research suggests that comprehensive ethics training wifl likely decrease unethical behavior by providing models of normatively appropriate behavior and reinforcing normatively appropriate responses to moral issues (Treviño et al., 2006; Treviño & Weaver, 2003; Weaver & Treviño, 1999).

Hypothesis 1: Levels of observed unethical behavior will be initially lower after comprehensive ethics training than before comprehensive ethics training.

Most comprehensive ethics training programs emphasize not only the prevention of unethical behavior (Hypothesis 1), but also the promotion of ethical behavior, which has been construed as behavior that exceeds some moral minimum (Treviño et al., 2006). Although ethical behavior in organizations may take many forms, most scholars argue that the reporting of unethical behavior (i.e., "whistle-blowing") is among the most important of ethical behaviors to organizations and societies (Dozier & Miceli, 1985; Kaptein, 2011a; Near & Miceli, 1995; Warren, 2003). Further, most scholars suggest that the reporting of unethical behavior is among the most desirable outcomes of comprehensive ethics training (Weaver & Treviño, 1999), as organizational members' willingness to report unethical behavior suggests that organizational members are aware that the behavior is unethical and believe that management will respond appropriately. Therefore, we are specifically focused on reporting unethical behaviors as a form of ethical behavior and as an attribute of an ethical organizational culture.

Comprehensive ethics training can influence employees' ethical behavior in several ways. First, such training enables employees to differentiate between ethical and unethical behavior by defining normatively appropriate responses to ethical issues (Weaver & Treviño, 1999). Indeed, classifying behavior as unethical is a necessary requisite to reporting such behavior. Second, comprehensive ethics training may encourage the integration of ethical behavior (such as the reporting of unethical behavior) into employees' models of role identities, such that employees perceive such behavior as an integral part of their organizational roles (Weaver & Treviño, 1999). A central precept of social learning theory is that individuals learn vicariously by attending to the behavior of a model; this model may appear in a hypothetical or real ethical dilemma presented to employees during a comprehen-

Is FORMAL ETHICS TRAINING MERELY COSMETIC? 93

sive ethies training session. The outeomes experieneed by this model vieariously shape individuals' outeome expeetaneies and therefore define roles (e.g., "I have to report it—it's part of my job.") for organizational members (Ashforth & Mael, 1989; Warren & Smith-Crowe, 2008; Weaver & Treviño, 1999). As developed in past researeh, aseribing the reporting of unethieal behavior into employees' role identides may inerease the likelihood that employees indeed report sueh behavior (Mieeli & Near, 2002; Treviño & Victor, 1992; Vadera, Aguilera, & Caza, 2009; Vietor, Treviño, & Shapiro, 1993). Notable whisde blowers, sueh as Cynthia Cooper (WorldCom), Sherron Watkins (Enron), and Coleen Rowley (FBI), never reported feeling like heroes beeause they pereeived that blowing the whistle was simply a part of their job (Near, Rehg, Seotter, & Mieeli, 2004).

Comprehensive ethies training also raises awareness of reporting proeedures, as the elear explieation of the process by whieh employees should report unethieal behavior is a fundamental eomponent of sueh training (Treviño & Weaver, 2003). Not surprisingly, researeh suggests that an understanding of reporting proeedures inereases the likelihood of whistle-blowing. Mieeli and Near (1984), for example, found that awareness of reporting ehannels was an important disdnetion between those federal employees who ehose to report wrongdoing and those employees who did not. In eomplementary researeh on employee voiee, Morrison and Milliken (2000: 713) highlight the existenee of eommunieadon ehannels in prevendng or- ganizadonal silenee and note "the laek of formal upward feedbaek meehanisms" in their presentation of theory on the stmetural features that will inhibit the sharing of threatening information.

Thus, eomprehensive ethies training defines unethieal behavior, ineorporates ethieal behavior into employees' models of role identities and explieates the proeess of reporting unethieal behavior. We expeet this kind of training to be assoeiated with inereased ethieal behavior and, speeifieally, the willingness to report unethieal behavior in organizations.

Hypothesis 2a: Employees will initially intend to engage in more ethical behav- ior (i.e., report wrongdoing) after comprehensive ethics training than before comprehensive ethics training.

Hypothesis 2b: More employees will initially intend to engage in ethical behav- ior (i.e., report wrongdoing) after comprehensive ethics training than before comprehensive ethics training.

Ethics Training and Organizational Efficacy

A willingness to behave ethically is most often assoeiated with an expeetation that the organization will appropriately respond. Treviño and Weaver (2001) introdueed the eoneept of "ethies program follow-through" to eapture employees' pereeptions of the reasonableness of an organization's response to ethieal lapses. Past researeh has shown that a positive pereepdon of organizadonal effieaey in addressing ethieal behavior is assoeiated with sueh desirable organizadonal outeomes as future ethieal behavior and willingness to report unethieal behavior (Treviño & Weaver, 2001).

94 BUSINESS ETHICS QUARTERLY

Although past research has explored the consequences of employees' perceptions of organizational efficacy, such research has overlooked the antecedents of these perceptions; that is, the factors that influence employees' perceptions of organiza- tional efficacy. We propose that comprehensive ethics training is one such factor.

Comprehensive ethics training can influence employees' perceptions of organiza- tional efficacy in managing ethical violations in at least two ways. Most importantly, this kind of training provides organizations with an opportunity to describe not only the process by which employees can report observed unethical behaviors (see Hy- potheses 2a and 2b), but also the process by which the organization can subsequently attend to and resolve these reports (Treviño & Weaver, 2003). This is typically done by describing previous cases handled by the organization; these cases will influence the employee's knowledge of misconduct and organizational action in the past and shape their expectations for the future. In this perspective, comprehensive ethics train- ing clarifies employees' expectations. As past management research has shown, the dehneation of expectations by organizations is positively associated with employees' perceptions of decision making effectiveness (Tsui, Ashford, Clair, & Xin, 1995).

Comprehensive ethics training may also influence employees' perceptions of organizational efficacy in managing ethical violations by creating a psychological contract between the organization and its employees (e.g., "You report it, we'll deal with it"; Morrison & Robinson, 1997; Nicholson & Johns, 1985; Robinson, Kraatz, & Rousseau, 1994). Supported by research on social exchange theory (Blau, 1964), this contract creates a reciprocal obligation between employees (to report unethical behavior) and the organization (to follow-through on the report; for discussions, see Rousseau & Parks, 1993; Weaver & Treviño, 1999). As the contract is subsequently fulfilled by the organization, support for the process on the part of employees will strengthen (Weaver & Treviño, 1999). Taken together, therefore, extant theory and research suggest that comprehensive ethics training will strengthen employees' perceptions of orgatiizational efficacy in managing ethical matters.

Hypothesis 3: Employees' perceptions of organizational efficacy in managing ethical violations will initially be more favorable after comprehensive ethics training than before comprehensive ethics training.

Ethics Training and Normative Structure

Ethical culture has been conceptualized as the multidimensional interplay between an organization's formal and informal systems that promote ethical or unethical behavior (Smith-Crowe et al., forthcoming; Tenbrunsel, Smith-Crowe, & Umphress, 2003; Treviño, 1990; Treviño et al., 1998). In a seminal paper, Treviño (1986) argues that a strong culture has a correspondingly strong normative structure that leads to more agreement among organizational members about what is and is not appropriate behavior (see Treviño, 1986: Proposition 11). Although past research explores the effects of informal and formal components of ethical cultures on (un) ethical behavior (Stnith-Crowe et al., forthcoming; Tenbrunsel et al., 2003; Treviño, 1990; Treviño et al., 1998), there remains a need to examine the effects of the more formal components of culture on the more informal components of culture. That is.

Is FORMAL ETHICS TRAINING MERELY COSMETIC? 95

we still know little about what Treviño (1990) describes as the interplay between the formal and informal components of ethical cultures.

As noted earlier, formal ethics programs involving comprehensive ethics training may contribute significantly to the ethical culture of an organization. Thus, we expect comprehensive ethics training to infiuence an important, informal component of an organization's informal ethics culture: the organization's normative structure. Fol- lowing prior research, we define an organization's normative structure as the norms about what is ethical and unethical behavior (Treviño et al., 1998).

Comprehensive ethics training may shape an organization's normative structure by denning values (the value-based dimension of ethics training) and rules (the compliance-based dimension) that refiect the aspirations and expectations of the organization (Weaver et al., 1999b). As Weaver and Treviño (1999: 319) note, "Widespread attention to shared values . . . helps create expectations or norms for appropriate behavior within an organization." Comprehensive ethics training may further reinforce such values and rules by detailing real or hypothetical dilemmas that expose employees to vicarious rewards and punishments (Ashkanasy et a l , 2006; Treviño & Youngblood, 1990).

Our arguments suggest that ethics training (a formal component of an organi- zation's ethical culture) will influence an organization's normative structure (an informal component of an organization's ethical culture), such that comprehensive ethics training will promote agreement among organizational members in the identification and perceptions of unethical behavior and, therefore, strengthen the normative structure underlying the organization's culture (Treviño, 1986).

Hypothesis 4: The normative structure of the organization will initially demonstrate more convergence in perceptions of unethical behavior after com- prehensive ethics training than before comprehensive ethics training.

Ethics Training and Organizational Values

Organizational values are a central feature of ethical cultures of an organization, and comprehensive ethics training may influence perceptions of these values in at least two ways. First, comprehensive ethics training provides an opportunity for organizations to formally introduce employees to the values of the organization; this exposure, research suggests, is likely to strengthen employees' perceptions of these values. Indeed, research on automatic social cognitive theory (Greenwald & Banaji, 1995) suggests that mere exposure to an organization's values may lead to the adoption of implicit associations between the organization and these values (Reynolds, 2006). Related laboratory research demonstrates that implicit associations are not only adopted by organizational members, but also infiuence the perceptions and behaviors of these members (Reynolds, Leavitt, & DeCelles, 2010).

Second, comprehensive ethics training provides employees with an opportunity to vicariously observe the values of the organization (Ashkanasy et al., 2006; Treviño & Youngblood, 1990). For instance, employees should be more likely to perceive that an organization possesses a value of discretion if exposed to policies regard- ing reporting procedures and information about the handling of previous cases. In

96 BUSINESS ETHICS QUARTERLY

tbeir study of environmentalism and hotels, for example, Goldstein, Cialdini, and Griskevicius (2008) found tbat statements tbat described tbe hotel-inspired conser- vationism efforts of otber guests increased tbe level of conservationism exhibited by study participants. This study demonstrated tbat merely describing tbe bebavior of others strengthened participants' perceptions of the hotel's values and increased tbeir subsequent compliance. Tbese findings, as well tbe research on automatic social cognition, suggest tbat exposure to organizational values will, directly and indirectly, positively influence employees' perceptions of tbese values.

Hypothesis 5a: Employees will initially perceive firm values as more important to their coworkers afier comprehensive ethics training than before comprehen- sive ethics training.

Hypothesis 5b: Employees will initially perceive firm values as more important to their supervisor afier comprehensive ethics training than before comprehen- sive ethics training.

Hypothesis 5c: Employees will initially perceive firm values as more important to senior managers afier comprehensive ethics training than before compre- hensive ethics training.

Training Effects over Time

Prior researcb has overlooked tbe longitudinal effects of formal etbics training, and tbeory in tbe behavioral sciences on these effects is contradictory, suggesting tbat formal ethics training will influence tbe organization in two possible ways. One tbeoretical perspective is consistent witb researcb on training and indicates tbat tbe effects of training will dissipate over the long-term (Ricbards, 1999; Weber, 1990). A second tbeoretical perspective, however, is grounded in Bandura's reciprocal determinism (1977, 1978) and indicates tbat changes associated witb training will become self-reinforcing and the effects sbould strengtben. We present tbeory related to botb and derive competing bypotbeses.

Conventional wisdom and general training research suggest a need to retrain employees periodically because the effects are expected to dissipate over time (Ricbards, 1999; Weber, 1990). Eurthermore, recent researcb documents individu- als' pervasive tendencies to rationalize their unetbical bebavior (Detert, Treviño, & Sweitzer, 2008; Sbu, Gino, & Bazerman, 2011 ) and "strategically forget" moral rules and principles (Shu et al., 2011; Sbu & Gino, 2012). Tbis experimental evidence, coupled with the persistence of corporate scandals, has led many to question tbe long-term effectiveness of formal ethics programs (Bazerman & Tenbrunsel, 2011; Tenbrunsel, Diekmann, Wade-Benzoni, & Bazerman, 2010). Eor instance, Tenbrunsel and colleagues note tbat tbe self-deceptive psychological process of organizational members "substantially negates any systematic efforts at the organization level" (Tenbrunsel et al., 2010: 169).

In contrast, Bandura's (1977, 1978) conceptualization of reciprocal determin- ism as a foundational component of social learning supports the long-term effects of comprebensive etbics training. Reciprocal determinism is a process whereby

Is FORMAL ETHICS TRAINING MERELY COSMETIC? 97

"behavior, interpersonal factors, and environmental influences all operate as inter- locking determinants of each other" (Bandura, 1978: 346). The complex, interactive natures of these determinants are fundamental. As Bandura (1977: 198) explains, "Studies of dyadic exchanges document how the behavior of one member activates particular responses from the repertoire of the other member which, in turn, prompt reciprocal counteractions." While reciprocal determinism has not been studied in the context of ethics training, empirical studies in organizational behavior offer insight into the process of reciprocal determinism in firms. We expect the introduction of comprehensive ethics training to influence the behavior of employees towards organizationally desired behaviors, such that once employee behavior is initially influenced, the behavior becomes self-perpetuating.

Supportive organizational research suggests that employees align with the behavior of in-group members, such that the behavior of in-group members is contagious and mirroring (Gino, Ayal, & Ariely, 2009; Gino, Gu, & Zhong, 2009). This effect may be explained by the saHence of moral norms, shifts in the underlying normative structure of the organization that define appropriate and inappropriate responses to ethical dilemmas, or employees' observations of the rewards and punishments experienced by what Bandura (1977) refers to as "live models" (that is, other or- ganizational members).

Other research indicates that as organizations "follow-through" on their ethical commitments (i.e., in post-training periods) and take "action to deal with ethical issues employees raise and violations of the company's formal ethics policies," em- ployees engage in more ethical and less unethical behavior (Treviño & Weaver, 2001 : 651). Collectively, a wide range of theory-driven research supports this perspective of reciprocal determinism whereby the effects of comprehensive ethics training on unethical behavior (see Hypothesis 1), ethical behavior (see Hypotheses 2a and 2b), and the normative structure of the organization (Hypothesis 4) will reinforce each other and persist with the passage of time.

For similar reasons, the effects of comprehensive ethics training should further hold for organizational efficacy and values. As organizations fulfill their psycho- logical contractual obligations, they will develop employees' trust—trust that is theoretically and empirically associated with support for the organization (Dirks & Ferrin, 2001; Robinson, 1996). Further, as organizations follow-through on their commitments to ethics, employees' outcome expectancies, perceptions of proce- dural and distributive justice, and perceptions of values are reinforced (Treviño & Ball, 1992; Treviño & Weaver, 2001). Collectively, the reinforcement of outcome expectancies, the positive experience of justice and trust, and the confirmation of perceptions of values are expected to confirm employees' earlier perceptions of organizational efficacy and values.

Hypothesis 6a: Indicators of an ethical organizational culture will shift towards ethics training content over time.

Hypothesis 6b: Indicators of an ethical organizational culture will shift away from ethics training content over time.

98 BUSINESS ETHICS QUARTERLY

METHODS

Study Setting

Motivated by the increasing regulatory ineentives to adopt formal ethies programs, a large, muldnadonal bank headquartered in the U.S. instituted its first fonnal eth- ies training initiadve. We were granted aeeess to the employees of the bank both before (Time 1) and after (Time 2: nine months later; Time 3: two-and-a-half years later) the organizadon adopted a eomprehensive ethies training program. The pro- gram, developed with the involvement of the first author, ineluded small, four-hour sessions eondueted faee-to-faee and delivered in the loeal language by one of two eonsultants who worked alongside management.

Ethies training sessions were the same for all employees. Eaeh session began with a foeus on the bank's values and prineiples. The sessions then eontinued with ease analyses of several ethieal dilemmas. In these diseussions, employees were eneouraged to identify unethieal behavior in the dilemmas, to eonsider ways that the behavior eould be avoided or prevented, to exereise their voiee in reporting wrong- doing, and to diseuss appropriate responses by the bank. Sessions eoneluded with an emphasis on the bank's mies, policies, and proeedures, as well as a diseussion of previous ethieal lapses and the bank's responses.

The bank had not previously developed and administered ethies and eomplianee training to its employees. Further, the Code of Conduet that ineluded the bank's Guiding Prineiples and Values was distributed prior to the ethies training. The only ethics inifiative during the period of this study was that of ethies training; the bank did not repeat the training during the time period of our study.

Participants: Time 1 and Time 2

The data used to test the first set of hypotheses (Hypotheses l-5e) were eolleeted from employees of the bank immediately before (Time 1) and nine months after (Time 2) the introduedon of the eomprehensive ethies training. The organizadon distributed an online survey to all 2,204 employees. The response rates for the Time 1 and Time 2 surveys were 47% (1,031 of 2,204 employees) and 69% (1,518 of 2,204), respeedvely. Three hundred and ninety-two employees (57% between the ages of 40 and 55'; 56% male; 18% of the organizadon) responded to both surveys and thus serve as our primary, matched data set. For robustness, we also tested our hypotheses on an unmatehed data set that ineludes those who responded to the Time 1 or the Time 2 survey (but not both). Through our unmatehed and matehed samples for Times 1 and 2, we eapture no less than 47% of the organizadon.^

Participants: Time 3

We retumed to the multinadonal bank two-and-a-half years after the organization adopted a eomprehensive ethies training program and administered the same survey used for the first two surveys. In this final data eolleedon effort, the organizadon again distributed the online surveys to all employees. The response rate for the Time 3 survey was about 52%. Five hundred and thirty-four employees (60% between the

Is FORMAL ETHICS TRAINING MERELY COSMETIC? 99

ages of 40 and 55^; 51% male, 24% of the organization) responded to both (Time 2 and Time 3) surveys and thus served as our primary, matched data set to test the extended hypotheses. For robustness, we again test our hypotheses on an unmatched data set that includes those who responded to the Time 2 or the Time 3 survey (but not both). Through our unmatched and matched samples for Times 2 and 3, we capture no less than 52% of the organization.

Dependent Variables

Unethical Behavior We captured observed unethical behavior with an eight-item scale adapted from Weaver and Treviño (1999). The scale asked employees to indicate the frequency (1 = "have never seen" to 6 = "more than monthly") with which they observed eight unethical behaviors. Sample items include, "How often have you seen an employee sharing confidential information?" and "How often have you seen an employee claiming unjustified expenses?" Following Weaver and Treviño ( 1999) and Treviño et al. (1998), we measured observed behavior, rather than self-reported behavior, to minimize concerns about social desirability bias. As Treviño and colleagues (1998: 456) note, "respondents are more likely to report that they observed others' unethical behavior than that they were unethical."

Ethical Behavior We captured ethical behavior through a measure of willingness to report unethical behavior using the same eight items from the observed unethical behavior scale above (adapted from Treviño et al., 1998, and Weaver & Treviño, 1999). The scale asked employees to indicate whether or not (dichotomous: "yes" or "no") they would report each of the eight unethical behaviors to the Batik's Ethics Committee or Chief Ethics Officer. This measure provides a particularly conservative test of employees' willingness to report unethical behavior, as many employees may be hesitant to report observed unethical behavior directly to the Ethics Committee or Chief Ethics Officer.

Perception of Organizational Efficacy We adopted a more comprehensive construct than that used in prior research to capture perceptions of organizational efficacy in managing ethical violations (cf. Treviño & Weaver, 2001). In contrast to past measures, which focus broadly on "follow-through" and "discipline," our seven-item measure captured more specific characteristics of these components, as well as other characteristics that reflect policies developed by the organization in the ethics training sessions (see Treviño & Weaver, 2001). The scale items include "Cases of misconduct are managed in a timely manner"; "Senior management does not influence the decisions of miscon- duct cases"; "Identities are protected in cases of misconduct"; "All information is kept confidential in cases of misconduct"; "All employees are treated as equals in the judgment of misconduct cases"; "Cases of ethics are handled in a fair manner"; and "Cases of ethics are effectively handled." All items were assessed on a six-point Likert-style scale (1 = "Strongly Disagree" to 6 = "Strongly Agree"). The scale also

100 BUSINESS ETHICS QUARTERLY

included a "Don't Know" option, wbicb allowed us to exclude tbose employees wbo were unfamiliar with tbe handling of etbics cases and thus to maintain tbe validity of tbe construct. Tbe scale demonstrated very bigb reliability across all time peri- ods (a = .90 to .91 for Time 1 and 2 Matched Data; a = .89 to .91 for Time 1 and 2 Unmatched Data; a = .92 for Time 2 and 3 Matcbed Data; a = .90 to .92 for Time 2 and 3 Unmatcbed Data).

Normative Structure We captured tbe organization's normative stmcture, conceptualized by Treviño (1986) as agreement among organizational members in tbe identification of unetbical bebavior, in a manner tbat is consistent witb prior researcb on the measurement of agreement in organizational researcb (for discussion of tbis measure, see Scbnei- der, Salvaggio, & Subirats, 2002). As sucb, normative stmcture was measured as tbe standard deviation of employees' responses to tbe unetbical bebavior constmct (Schneider et al., 2002).

Perceived Values: Coworkers, Supervisors, and Senior Managers Six items assessed employees' perceptions of tbe importance (Likert-scale; 1 = "Extremely Unimportant" to 7 = "Extremely Important") of moral values to tbree segments of tbe organization: coworkers, supervisors, and senior managers. The six values (integrity, loyalty, equity, tolerance, impartiality, and discretion) not only reflect tbe organization's specific values, but also align with the fundamental moral values of corporate cultures (for discussion, see Jones, Eelps, & Bigley, 2007). Subscales (simple averages of tbe items) were created for eacb of three segments (coworkers, supervisors, and senior managers). All subscales demonstrated very bigh reliability (a = .91 to .95 for Time 1 and 2 Matcbed Data; a = .92 to .96 for Time 1 and 2 Unmatcbed Data; a = .91 to .94 for Time 2 and 3 Matcbed Data; a = .92 to .94 for Time 2 and 3 Unmatched Data).

RESULTS

Preliminary Analyses

We performed exploratory and confirmatory factor analyses (EEA and CEA, respec- tively) on tbe items for the four latent constmcts in the study. Tbe four constmcts are perceptions of coworker values, perceptions of supervisor values, perceptions of senior manager values, and perceptions of organizational efficacy. Eirst, we performed an EEA. Tbe analysis indicated tbat tbe measured items loaded on four factors and tbat none of tbe measured items loaded on more than one factor. Tbat is, tbe items for perceptions of coworker values loaded on one factor, tbe items for perceptions- of supervisor manager values loaded on a second factor, tbe items for perceptions of senior manager values loaded on third factor, and tbe items for per- ceptions of organizational efficacy loaded on a fourth factor. The EEA, tberefore, revealed a factor structure consistent witb tbe proposed four-factor model.

Second, we performed a CEA. Tbe purpose of the CEA was to determine wbetber tbe data fit tbe tbeoretical, four-factor stmcture. In tbis analysis, we focused on

Is FORMAL ETHICS TRAINING MERELY COSMETIC? 101

the most common indices of fit in management and ethics research. These indices include the normed chi-squared (x^idf), the comparative fit index (CFI), and the standardized root mean square residual (SRMSR). The fit of the data to the proposed model is typically described as acceptable if the x /̂df ratio falls below 5 (Wheaton, Muthen, Alwin, & Summers, 1977) or 3 (Carmines & Mclver, 1981), the CFI is approximately 0.9 (Jaccard & Wan, 1996), and the SRMSR does not exceed .08 (Hu & Bentler, 1999).

The results of the CFA were consistent with the results of the EFA and indicated that the four-factor model fit the data. Indeed, the x'/df ratio was less than 3.0 or 5.0 in all time periods, the CFI was between 0.86 and 0.91 in all time periods, and the SRMSR was less than or equal to 0.08 in all time periods for the matched and unmatched data. (Table 1 includes the results of the CFA for the matched and un- matched data.) The CFA, therefore, provided strong support for the validity of the four-factor model.

Table I:'' Results of Confirmatory Factor Analyses for the Four Latent Constructs: Perceptions of Coworker, Supervisor, and Senior Manager Values, and Perceptions of Organizational Efficacy

X̂ Degrees of Normed x̂ Comparative Fit Standardized Root Time Period freedom (df) (xVdf) Index (CFI) Mean Square Re-

sidual (SRMSR)

Time 1 and 2 Matched

Time 1 and 2 Unmatched

Time 2 and 3 Matched

Time 2 and 3 Unmatched

Time 1

Time 2

Time 1

Time 2

Time 2

Time 3

Time 2

Time 3

733.73

656.41

752.92

945.19

898.37

683.95

970.42

588.14

269

269

269

269

269

269

269

269

2.73

2.44

2.80

3.51

3.34

2.54

3.61

2.19

0.89

0.89

0.88

0.91

0.88

0.90

0.92

0.86

0.06

0.06

0.06

0.05

0.06

0.06

0.05

0.08

In the development of and for the duration of the study, we took several steps to minimize concerns regarding common method bias. For instance, we followed many of the recommendations made by Podsakoff, MacKenzie, Lee, and Podsakoff (2003). More specifically, we measured our constructs across multiple time periods, we reminded participants that their responses were anonymous, we carefully devel- oped, pretested, and revised scale items, and we included many different types of scales in the survey (for a detailed discussion, see Podsakoff et al., 2003).

In the analysis of the data, we took further steps to minimize the concern for common method bias. For instance, we replicated our analyses on unmatched and matched samples of respondents, and we performed Harman's single-factor test to statistically assess the presence of common method bias (see prior discussion). As noted by Podsakoff and Organ (1986), common method bias is a concern only if (a) one factor emerges in an EFA or (b) one factor in an EFA captures the majority of the covariance among measures. The results of Harman's single factor test revealed multiple factors accounting for substantial variance across all data sets and therefore further alleviated any concerns of common method bias.^

102 BUSINESS ETHICS QUARTERLY

Time I and 2 Matched Data

To test Hypothesis 1, we performed a planned, repeated-measure eomparison. As predieted, the analysis revealed that unethieal behavior was lower in the time period after ethies training (Time 2; mean =1.19) than in the time period before ethies training (Time 1; mean = 1.29; t [391] = -5.07, p< .001).

To test Hypotheses 2a and 2b, we eondueted two separate analyses. In the first analysis, we summed responses aeross the observed unethieal behavior items to ereate a eonstmet that eaptured the number of unethical behaviors employees were willing to report. As predieted by Hypothesis 2a, a planned, repeated-measure eom- parison demonstrated that the average number of unethieal behaviors employees were willing to report was higher in the dme period affer ethies training (Time 2; mean = 2.33) than in the dme period before training (Time 1 ; mean = 1.00; t [391] = 10.07, p < .001), sueh that employees were willing to report on average more than twiee as many behaviors in the time period after ethies training than in the fime period before ethies training. In the seeond analysis, we eompared the number of employees willing to report eaeh of the eight unethieal behaviors (that formed the eonstmet) in the dme period before and after ethies training. The results of the MeNemar tests, whieh analyze frequeney data from paired samples, eonfirmed Hypothesis 2b: more employees were willing to report eaeh of the eight unethieal behaviors in the dme period after ethies training than in the time period before ethies training. See Table 2 for the results of the MeNemar tests.

Hypothesis 3 foeused on employees' pereepdons of organizational effieaey in managing ethieal violations. As predieted, a planned, repeated-measure eomparison demonstrated that pereepdons of organizadonal effieaey improved in the dme period

Table 2: Results of MeNemar Tests for Time 1 and Time 2 Matched Data

Violation

Accepting a gift in excess of $100

Claiming an unjustified expense

Sharing confidential information

Accepting a kickback

Engaging in nepotism

Performing personal business on company time

Stating discriminatory remarks

Stating inappropriate sexual comments

WUIing to Report: Time 1

52

38

47

92

69

16

34

44

Wiiiing to Report: Time 2

119

91

114

172

145

39

113

122

Test Statistic

45.34 ***

33.84***

42.75***

47.06***

48.94***

11.76***

59.44***

52.45***

* p < .05 * * p < . 0 1 ***i><.001

Is FORMAL ETHICS TRAINING MERELY COSMETIC? 103

after ethics training (Time 2; mean = 3.30) when compared to the time period before ethics training (Time 1; mean = 3.07; t [261] ^ 3.73, p < .001).

To test Hypothesis 4, which focused on the normative structure underlying the organization's ethical culture, we performed a Pittman-Morgan test of variance, which tests for the equality of two dependent variances. As expected, the normative structure strengthened, such that there was significantly more agreement (which is reflected statistically by lower variance) about appropriate behavior in the time period after ethics training (Time 2; variance = 0.10) than in the time period before ethics training (Time 1; variance = 0.21; î [390] = -8.60, p < .001).

Our last predictions (Hypotheses 5a-5c) for Time 1 and Time 2 concerned percep- tions of values across three segments of the organization: coworkers, supervisors, and senior managers. As predicted, planned repeated-measure comparisons demonstrated that employees perceived the firm's values as more important to coworkers in the time period after ethics training (Time 2; mean - 5.63) than in the time period before ethics training (Time 1; mean = 5.53; t [391] = 1.93, p < .05), and that employees perceived the firm's values as more important to senior managers in the time period after ethics training (Time 2; mean - 5.37) than in the time period before ethics training (mean = 5.24; í [391] = 2.39, p < .05). However, there were no differences in employees' perceptions of the firm's values to supervisors in the time period after ethics traitiing (mean = 5.79) and the time period before ethics training (mean = 5.75; r [391] = 0.61, p = n.s.).

Time 1 and 2 Unmatched Data

For robustness, we repeated the above analysis on the unmatched data, which includes those who responded to the Time 1 or Time 2 survey (but not both). To test Hypothesis 1, we performed a planned comparison.^ As predicted, the analysis revealed that unethical behavior was lower in the time period after ethics training (Time 2; mean = 1.21) than in the time period before ethics training (Time 1; mean = 1.34; Í [471.13] = -4.39, p < .001).

To test Hypothesis 2a, we again summed responses across the observed unethical behaviors to create a construct that captured the number of behaviors employees were willing to report. A planned comparison confirmed that the number of unethi- cal behaviors employees were willing to report was higher in the time period after ethics training (Time 2; mean = 2.06) than in the time period before training (Time 1; mean = 0.91; t [793.81] = 8.27, p < .001), such that employees were willing to report more than twice as many behaviors in the time period after ethics training than in the time period before ethics training.^

Hypothesis 3 focused on employees' perceptions of organizational efficacy in managing ethical violations. As expected, a planned comparison confirmed that perceptions of organizational efficacy improved in the time period after ethics train- ing (Time 2; mean = 3.42) when compared to the time period before ethics training (Time 1; mean = 3.01; t [423.95] = 6.20, p < .001).

To test Hypothesis 4, which focused on the normative structure underlying the organization's ethical culture, we tested for the equality of the two (independent)

104 BUSINESS ETHICS QUARTERLY

variances. As expected, the normative structure strengthened, such that there was more agreement about appropriate behavior in the time period after ethics training (Time 2; variance - 0.14) than in the time period before ethics training (Time 1; variance = 0.24; F [315, 794] = 1.71, p < .001).

Our last predictions (Hypotheses 5a-5c) for the Time 1 and Time 2 unmatched data concerned perceptions of the firm's values across three segments of the or- ganization: coworkers, supervisors, and senior managers. As expected, a series of planned comparisons showed that employees perceived the firm's values as more important to coworkers in the time period after ethics training (mean = 5.65) than in the time period before ethics training (mean = 5.31; í [482.84] = 4.34, p < .001), more important to supervisors in the time period after ethics (mean = 5.90) than in the time period before ethics training (mean = 5.66; t [482.20] = 2.87,p < .01), and more important to senior managers in the time period after ethics training (mean = 5.53) than in the time period before ethics training (mean = 5.12; t [527.58] = 4.93, p < . 0 0 1 ) .

Time 2 and 3 Matched Data

To test Hypotheses 6a and 6b, we performed planned, repeated-measure compari- sons to test statistically significant shifts in means of the attributes of organizational culture. In support of Hypothesis 6b, we found unethical behavior was lower in the time period two-and-a-half years after ethics training (Time 3; mean =1.17) than in the time period nine months after ethics training (Time 2; mean = 1.20; í [531] = -2.12,p<.05).

To test shifts in ethical behavior, we again conducted two separate analyses. In the first analysis, we summed responses across the observed unethical behavior items to create a construct that captured the number of unethical behaviors employees were willing to report in the time period before and the time period after the ethics training. A planned, repeated-measure comparison demonstrated that the number of unethical behaviors employees were willing to report in Time 3 (mean = 2.20) remained at Time 2 levels (mean = 2.16; í [533] = 0.30, p = n.s.), suggesting that

Table 3: Results of McNemar Tests for Time 2 and Time 3 Matched Data

Violation

Accepting a gift in excess of $100

Claiming an unjustified expense

Sharing confidential information

Accepting a kickback

Engaging in nepotism

Performing personal business on company time

Stating discriminatory remarks

Stating inappropriate sexual comments

Willing to Report: Time 2

167

121

148

224

192

52

131

141

Willing to Report: Time 3

174

125

142

217

181

51

131

134

Test Statistic

0.29

0.11

0.23

0.29

0.71

0.02

0.00

0.31

t / 7 < . 1 0 *p<.05 **p<a\ ***;>< .001

Is EoRMAL ETHICS TRAINING MERELY COSMETIC? 105

tbe effect of comprebensive ethics training persisted for two-and-a-balf years after tbe comprebensive etbics training sessions. In tbe second analysis, we compared tbe number of employees willing to report eacb of the eigbt unetbical bebaviors (tbat formed tbe constmct) in tbe time period before and after etbics training. Tbe results of tbe McNemar tests (see Table 3) suggest that as many employees were willing to report eacb of tbe eigbt unetbical bebaviors in botb Time 3 and Time 2.

In order to examine sbifts in tbe normative stmcture underlying tbe organiza- tion's etbical culture, we once again performed a Pittman-Morgan test. Tbis test indicated that tbe normative stmcture furtber strengtbened in Time 3, such tbat tbere was more agreement (wbicb is statistically reflected by lower variance) about appropriate bebavior in Time 3 (variance = 0.08) tban in Time 2 (variance = 0.10; Í [530] = 2.84,p< .01).

No evidence was found of sbifts in employees' perceptions of organizational efficacy in managing etbics. A planned, repeated-measure comparison sbowed no evidence of a sbift in perceptions of organizational efficacy from Time 2 (mean = 3.31) to Time 3 (mean = 3.30; t [388] = 0.55, p = n.s.).

In support of Hypothesis 6a, planned, repeated-measure comparisons sbowed statistically significant reductions in employees' perceptions of tbe importance of tbe firm's values to employees, supervisors, and senior managers. Employees per- ceived tbe firm's values as less important to coworkers in Time 3 (mean = 5.47) tban in Time 2 (mean = 5.65; í [530] = -3.43, p < .001), less important to supervisors in Time 3 (mean = 5. 75) tban in Time 2 (mean = 5.86; t [531] = -2.09, p< .05), and less important to senior managers in Time 3 (mean = 5.22) tban in Time 2 (mean = 5.41; Í [529] =-3.76, p < . 0 0 1 ) .

Thus, we found mixed support for Hypotheses 6a and 6b in tbe matcbed data. Sbifts in perceptions of values supported H6a; sbifts in observed unetbical bebavior and tbe normative stmcture supported H6b. Tbe lack of statistically significant sbifts for intentions to bebave etbically and perceptions of organizational efficacy suggest tbese attributes of culture remained similar between Time 2 and Time 3.

Time 2 and 3 Unmatched Data

To furtber test Hypotheses 6a and 6b, we performed planned comparisons to capture statistically significant sbifts in means of tbe attributes of etbical organizational culture. A planned comparison demonstrated that unetbical bebavior remained at tbe Time 2 level (mean = 1.20) in Time 3 (mean = 1.22; t [1242] = 1.11, p = n.s.).

To examine sbifts in tbe levels of observed etbical bebaviors, we again summed responses across tbe observed unetbical bebavior items to capture the number of behaviors employees were willing to report in tbe time period before and tbe time period after tbe etbics training. A planned comparison demonstrated that the number of unethical bebaviors employees were willing to report in Time 3 (mean = 2.02) resembled those in Time 2 (mean = 2.05; t [1242] = -0.21, p = n.s.), sucb tbat employees were willing to report the same number of unetbical bebaviors in Time 3 as in Time 2.

106 BUSINESS ETHICS QUARTERLY

To examine shifts in the normative structure underlying the organization's ethi- cal culture, we tested the equality of the two independent variances. The analysis indicated that the strength of the normative structure persisted, such that agreement about appropriate behavior in Time 3 (variance = 0.37) was as strong as agreement in Time 2 (variance = 0.37; F [931, 311] = 1.00,/? = n.s.).

We also examined shifts in employees' perceptions of organizational efficacy. A planned comparison suggests that perceptions of organizational efficacy in Time 3 (mean - 3.31) were similar to Time 2 (mean = 3.38; t [969] = -0.94, p = n.s.).

Our last analyses concerned perceptions of the firm's values across three segments of the organization: coworkers, supervisors, and senior managers. Our planned com- parisons revealed statistically significant shifts in perceptions of the firm's values in all three segments. Employees perceived the firm's values as less important to coworkers in Time 3 (mean = 5.45) than in Time 2 (mean = 5.65; t [1242] = -2.96, p < .01), less important to supervisors in Time 3 (mean = 5. 67) than in Time 2 (mean = 5.86; t [1240] = -2.75, p < .01), and less important to senior managers in Time 3 (mean = 5.22) than in Time 2 (mean = 5.48; / [1237] = -3.32, p < .001)

Similar to the matched data set, we find partial support for Hypothesis 6a in the shifts of perceptions of values. Unlike the matched data set, observed unethical behavior and the normative structure did not shift in a statistically significant man- ner, which suggests these attributes of culture, like intentions to behave ethically and perceptions of organizational efficacy, remained similar between Time 2 and 3.

DISCUSSION

Though the U.S. Organizational Sentencing Guidelines provide firms with strong incentives to develop formal ethics training programs, promote ethical organiza- tional cultures, and thereby decrease corporate offenses, much of the past research has overlooked the characteristics of these training programs. In this study, we introduced a theoretical foundation for comprehensive ethics training programs and the influence of these programs in organizations. We also reported the results of a longitudinal study that focused on the influence of comprehensive ethics training on organizations and the perceptions and decisions of their members. Though prior research on ethics training programs suggests that formal ethics training effects are momentary (Richards, 1999), non-existent (Laufer, 1999; Laufer & Robertson, 1997; McCabe, Treviño, & Butterfield, 1996; Stansbury & Barry, 2007), or negative (Kaptein, 201 lb; McKendall et al., 2002; Sparks & Hunt, 1998), our study indicates that comprehensive ethics training is positively related to attributes of an ethical organizational culture and that many of these effects persist in the years after the introduction of the training program. This study provides not only evidence that formal ethics training programs are not merely cosmetic but also provides important descriptive—and prescriptive—insight into the development of effective formal ethics training programs in organizations.

Most importantly, through our longitudinal study design, we were able to cap- ture both the convergence of initial effects on values and behavior within a year of training and the divergence of eftects after the second year. More specifically, all

Is FORMAL ETHICS TRAINING MERELY COSMETIC? 107

of our indieators of an ethieal organizational eulture shifted towards the eontent of the ethies training within the first year after training, and many remained the same or improved with dme; pereepdons of the importanee of organizadonal values, however, did not remain elevated after two years. Eor instanee, after the seeond year, a typieal employee observed fewer aets of unethieal behavior, but also pereeived organizadonal values as less important to their eoworkers, supervisors, and upper management. Even in our unmatehed samples of study partieipants, values deereased in importanee after the seeond year, but behaviors remained the same. Given the sustained or improved behavioral measures and pereeptions of organizational ef- fieaey in managing ethies, the laek of eorresponding long-term improvement in pereepdons of values is somewhat surprising, as scholars theorize that sustained behavioral improvements refleet the ineuleation of organizadonal values (Kaptein, 201 la; Paine, 1994; Tyler & Blader, 2005). From this theoredeal perspeetive, ethi- eal behavior is best aehieved through the intemalization of shared organizational values, suggesting that the intemalization of values mediates the relationship be- tween formal ethies training and behavioral outeomes. While the nature of our study design prevents us from testing the mediating role of values in shifting behavioral outeomes, observations of unethieal behavior deereased between Times 2 and 3, an effeet that suggests the two may not be as dghtly eoupled as past theory suggests. In all, we find that changes to pereepdons of values shift independendy of unethieal and ethieal behaviors, such that an unfavorable shift in pereeptions of values does not parallel a similar shift in unethieal behaviors.

The departure between past theory on the importanee of values and our findings ean be interpreted in a variety of ways. First, the behavioral outeomes of the training may have been inidally driven by pereeptions of the importanee of organizadonal values, but onee behavioral outeomes improved, the behavior may have beeome self-reinforeing, sueh that future deereases in pereepdons of values had no effeet on the behavioral outeomes. This eoneept aligns with Bandura's (1977) eoneept of reeiproeal determinism, whereby one individual's behavior affeets a seeond indi- vidual's behavior and the seeond individual's behavior, in tum, infiuenees the first. An altemative explanation suggests that prior to training the bank had already met a minimal level of shared organizational values needed for the positive behavioral and pereeptual outeomes related to fonnal ethies training. Given that bank employee pereeptions of organizational values were already eonsidered important before train- ing, the bank may have already possessed a foundation of shared organizational values needed for formal ethies training to be effeefive. Therefore, the retum to pre-training pereeptions of organizational values had no impaet on the behavioral and pereeptual variables. Last, it is possible that reduction in unethieal behavior may have been driven by the eomplianee aspeet of the training while shifts in the pereeptions of values were driven by the ethies-oriented aspeet of training, and these eomponents fare differently over time. Future researeh should disentangle training effeets on values from other important behavioral outeome variables, sueh as a will- ingness to report wrongdoing, with the goal of understanding the meaningfulness of misalignment between attributes of an ethieal organizadonal eulture. Knowing whieh attributes are likely to shift undesirably over time furthers our understanding

108 BUSINESS ETHICS QUARTERLY

of how culture evolves and can ultimately contribute to theory on promoting and maintaining an ethical organizational culture.

Regardless of the underiying mechanisms explaining the differential effects in the second year, our study findings suggest that the positive behavioral effects of formal ethics training can last more than two years after training, which contradicts the findings of some recent empirical studies (Kaptein, 2011b; McKendall et al., 2002; O'Fallon & Butterfield, 2005; Sparks & Hunt, 1998). Past inconsistencies in research on formal ethics training effectiveness may reflect differences in types of training. Here, we studied comprehensive ethics training, a specific form of for- mal ethics training that we described as driven by prescriptions found in judicial guidelines and framed and supported by a stream of research from highly regarded social cognition theories. While past researchers studied formal ethics training, the specific attributes of the formal ethics training—beyond frequency of communica- tion (Kaptein, 2011b)—has been largely unexamined in field studies on unethical behavior. Though some studies consider the content for business ethics courses (e.g.. Palmer & Zakhem, 2001; Pamental, 1991) and ethics programs (e.g., Reynolds & Bowie, 2004), few consider and study the content of formal ethics training (for an exception, see Sekerka, 2009) or relate it to specific employee outcomes. Research- ers who found a negative relationship between formal ethics training and ethical outcomes did not differentiate types of training based upon content (Kaptein, 201 lb; McKendafl et al., 2002; O'Fallon & Butterfield, 2005; Sparks & Hunt, 1998). For instance, while Sparks and Hunt (1998) consider both participants' formal ethics training in their college education and in their careers, the two forms of ethics education are added to create a singular variable of ethics training. Presumably, the training received at work is more oriented towards the types of dilemmas faced in a particular work organization than the training received in college; with combined variables, however, the distinct effects of workplace ethics training are unknown.

The many renditions of training content and mechanisms of delivery thus com- plicate generalizations of all forms of ethics training, ours included. We view this as a fruitful direction for future research, as organizational literature would benefit from a more detailed understanding of how and when formal ethics traitung affects ethical outcomes. For instance, future studies could vary the mode of delivery, length of training, or content of the formal ethics training to better understand the effects on employee behavior and perceptions. In particular, a deeper understanding of the need for face-to-face training is particularly important given the prevalence of computer-mediated education, in general, and in ethics, in particular (Sekerka, 2009). While computer-mediated training appears, at first, to be cost-effective, organizations must consider the intermediate and long-term costs if computer-mediated ethics training has little to no effect on attributes of an ethical organizational culture. In such situations, less frequent face-to-face training may be a more effective approach to promoting an ethical organizational culture. Specifying how different kinds of ethics training fare over time would not only help further theory on formal ethics programs, but also assist in shaping policy on the requirements of such training.

Is FORMAL ETHICS TRAINING MERELY COSMETIC? 109

Limitations

This study is not without shortcomings. Eirst, as with much data supporting research in behavioral ethics, we were limited by our field site. These constraints did not permit an exploration of the underiying theoretical mechanisms through which comprehensive ethics training inñuenced the outcome variables. In order to obtain such highly sensitive data, we agreed to the research site's restrictions on research questions and study design. Though a difficult undertaking, future research could empirically examine these mechanisms by directly measuring, for example, em- ployees' outcome expectancies and moral awareness.

Second, the lack of a control sample necessarily limits any causal claims. This data limitation is quite common in pre-post quasi-experimental designs and restricts our causal inferences. Eor this reason, our study adopted the more conservative matched and unmatched pre-post analysis approach recommended by Hall and Mansfield ( 1971 ) to minimize concerns over internal validity. This approach allowed us to exclude most of the "rival hypotheses" described in a classic treatment by Campbell and Stanley (1963), except for history and maturation (for discussion, see Hall & Mansfield, 1971). At the same time, the results of the study are difficult to reconcile with these alterna- tive hypotheses. Eor instance, if the results of this study reflect maturation effects, the measures should increase or decrease through the duration of the study (see Campbell & Stanley, 1963). In contrast, the results of the present study demonstrated increases and decreases in constructs across time periods (e.g., values). Eurther, if the results of the study reflect historical effects, the measures should increase or decrease in paral- lel within time periods. Yet, the results demonstrated striking divergence in measures within time periods (e.g., values and behaviors). These inconsistencies, in combina- tion with the results of prior cross-sectional research (Treviño et al., 1998; Treviño & Weaver, 2001; Weaver & Treviño, 1999; Weaver et al., 1999a, 1999c), strongly support our findings about the effects of comprehensive ethics training programs.

CONCLUSION

Despite the importance of formal ethics training programs to the U.S. Organizational Sentencing Guidelines, much of the past research has overlooked the characteristics and the influence of these programs. This research, however, is necessary to under- stand the kinds of programmatic efforts that successfully translate the theoretical constructs of responsive regulation. Erom the perspective of regulatory policy, the success of corporate oversight depends, in large part, on what makes for efficient and effective self-regulation—e.g., the kind of organizational due diligence reflected in comprehensive ethics training.

At the foundation of responsive regulatory theory is the premise that organiza- tions can successfully self-regulate when certain incentives and disincentives are in place. Here we take an important step in examining the introduction of formal ethics training as a mechanism of firm self-regulation, with the goal of instilling an ethical organizational culture. The findings suggest that such mechanisms are not merely cosmetic and have important—and complex—influences on the perceptions and decisions of organizational members.

l i o BUSINESS ETHICS QUARTERLY

NOTES

We are grateful for insightful comments on earlier drafts from Kristin Smith-Crowe and valuable guidance from our associate editor, Bmce Barry, and three anonymous reviewers. We also thank the Zicklin Center for Business Ethics Research for generously funding this research project.

1. At the time of the Time 2 survey. 2. The respondents in the matched Time 1 data were older (p < .001) and consisted of a higher pro-

portion of males (p < .01) than the Time 1 respondents in unmatched data. However, the respondents were similar in the number of years the respondents worked in their current position (p = n.s.) and for the bank (p = n.s.). Further, the Time 1 respondents in the matched data were older (p < .001), consisted of a higher proportion of males (p < .01), and had more years of experience in their current position (p < .001) and for the bank (p < .001.) than the Time 2 respondents in the unmatched data. This indicates that the analyses on the matched data are particularly important, as the use of the same participants provides for a control of differences in age, gender, and experience.

3. At the time of the Time 3 survey, one was missing. 4. Acceptable model fit is typically inferred when the •¿Idfxa.úo falls below 5 (Wheaton, Muthen,

Alwin, & Summers, 1977) or 3 (Carmines & Mclver, 1981), when the CFI is approximately 0.9 (Jaccard & Wan, 1996), and when the SRMSR does not exceed .08 (Hu & Bentler, 1999).

5. The results of the exploratory factor analyses are available from the authors. 6. We adjusted all Time 1 vs. Time 2 unmatched planned comparisons for unequal variances. This

adjustment was necessary, as preliminary analyses suggested that the assumption of homoscedasticity was not satisfied for the Time 1 and Time 2 unmatched data. Note, however, that significance levels remain unchanged compared to the non-adjusted (standard) planned comparison.

7. McNemar's test requires matched (i.e., paired) data; thus, our Time 1 and Time 2 urmiatched data do not permit an exploration of Hypothesis 2b.

REFERENCES

Arthur, W., Bennett, W., Edens, P. S., & Bell, S. T. 2003. Effectiveness of training in or- ganizations: A meta-analysis of design and evaluation features. Journal of Applied Psychology, 88: 234-45. http://dx.doi.Org/10.1037/0021-9010.88.2.234

Ashforth, B., & Mael, F. 1989. Social identity theory and the organization. Academy of Management Review, 14: 20-39.

Ashkanasy, N. M., Windsor, C. A., & Treviño, L. K. 2006. Bad apples in bad barrels revisit- ed: Cognitive moral development, just world beliefs, rewards, and ethical decision- making. Business Ethics Quarterly, 16: 449-73. http://dx.doi.org/10.5840/beq200616447

Ayres, I., & Braithwaite, J. 1992. Responsive regulation: Transcending the deregulation debate. New York: Oxford University Press.

Baldwin, R., & Black, J. 2010. Really responsive risk-based regulation. Law and Policy,

32: 181-213. http://dx.doi.Org/10.llll/j.1467-9930.2010.00318.x

Bandura, A. 1977. Social learning theory. Englewood Cliffs, NJ: Prentice Hall.

1978. The self-system in reciprocal determinism. American Psychologist, 33: 344-58. http://dx.doi.org/10.1037/0003-066X.33.4.344

Bandura, A., Blanchard, E. B., & Ritter, B. 1969. Relative efficacy of desensitization and modeling approaches for inducing behavioral, affective, and attitudinal changes. Journal of Personality and Social Psychology, 13: 173-99. http://dx.doi.org/10.1037/h0028276

Is FORMAL E T H I C S TRAINING M E R E L Y COSMETIC? I l l

Bandura, A., & Mischel, W. 1965. Modifications of self-imposed delay of reward through exposure to live and symbolic models. Journal of Personality and Social Psychol- ogy, 2: 698-705. http://dx.doi.org/10.1037/h0022655

Bazerman, M. H., & Tenbrunsel, A. E. 2011. Blind spots: Why we fail to do what's right and what to do about it. Princeton, NJ: Princeton University Press.

Blau, P. M. 1964. Exchange and power in social life. New York: Wiley.

Butterfield, K. D., Treviño, L. K., & Weaver, G. R. 2000. Moral awareness in business organizations: Influences of issue-related and social context factors. Human Rela- ftons, 53:981-1018.

Campbell, D. T., & Stanley, J. C. 1963. Experimental and quasi-experimental designs for research. Chicago: Rand McNally College Publishing Company.

Carmines, E. G., & Mclver, J. D. 1981. Analyzing models with unobserved variables: Analysis of covariance structures. In G. W. Bohinstedt & E. F. Borgatta (Eds.), So- cial measurement: Current issues: 65-115. Beverly Hills, CA: Sage.

Detert, J. R., Treviño, L. K., & Sweitzer, V. L. 2008. Moral disengagement in ethical deci- sion making: A study of antecedents and outcomes. Journal of Applied Psychol- ogy, 93: 374-91. http://dx.doi.Org/10.1037/0021-9010.93.2.374

Dirks, K. T., & Ferrin, D. L. 2001. The role of trust in organizational settings. Organiza- tional Science, 12: 450-67. http://dx.doi.Org/10.1287/orsc.12.4.450.10640

Dozier, J. B., & Miceli, M. P. 1985. Potential predictors of whistle-blowing: A prosocial hehaviorperspcctive.Academy of Management Review, 10: 823-36.

Drolet, A. L., & Morris, M. W. 2000. Rapport in conflict resolution: Accounting for how face-to-face contact fosters mutual cooperation in mixed-motive conflicts. Journal of Experimental Social Psychology, 36: 26-50. http://dx.doi.org/10.1006/jesp.1999.1395

Ferrell, O. C , LeClair, D. T., & Ferrell, L. 1998. The federal sentencing guidehnes for or- ganizations: A framework for ethical compliance. Journal of Business Ethics, 17: 353-63. http://dx.doi.Org/10.1023/A:1005786809479

Fiske, S. T., & Taylor, S. E. 2008. Social cognition: From brains to culture. New York: McGraw-Hill.

Gino, F., Ayal, S., & Ariely, D. 2009. Contagion and differentiation in unethical behav- ior: The effect of one bad apple on the barrel. Psychological Science, 20: 393-98. http://dx.doi.Org/10.llll/j.1467-9280.2009.02306.x

Gino, F., Gu, J., & Zhong, C. B. 2009. Contagion or restitution? When bad apples ean moti- vate ethical behavior. Journal of Experimental Social Psychology, 45: 1299-1302. http://dx.doi.Org/10.1016/j.jesp.2009.07.014

Gioia, D. A. 1992. Pinto fires and personal ethics: A script analysis of missed opportunities. Journal of Business Ethics, 11: 379-89. http://dx.doi.org/10.1007/BF00870550

Goldstein, N. J., Cialdini, R. B., & Griskevicius, V. 2008. A room with a viewpoint: Using social norms to motivate environmental conservation in hotels. Journal of Con- sumer Research, 35: 472-82. http://dx.doi.org/10.1086/586910

112 BUSINESS E T H I C S QUARTERLY

Greenwald, A. G., & Banaji, M. R. 1995. Implicit social cognition: Attitudes, self-esteem, and stereotypes. Psychological Review, 102: 4-27. http://dx.doi.org/10.1037/0033-295X. 102.1.4

Hall, D. T., & Mansfield, R. 1971. Organizafional and individual response to extemal re- sponse. Administrative Science Quarterly, 16: 5 3 3 ^ 7 . http://dx.doi.org/10.2307/2391771

Harrison, J. K. 1992. Individual and combined effects of behavior modeling and the cul- tural assimilator in cross-cultural management training. Journal of Applied Psy- chology, 77: 952-62. http://dx.doi.Org/10.1037/0021-9010.77.6.952

Hess, D., McWhorter, R., & Fort, T. L. 2006. The 2004 amendments to the Federal Sen- tencing Guidelines and their implicit call for a symbiotic integrafion of business ethics. Fordham Journal of Corporate & Financial Law, 11: 725-64.

Hu, L., & Bentler, P. M. 1999. Cutoff criteria for fit indexes in covariance stmcture analy- sis: Conventional criteria versus new altematives. Structural Equation Modeling, 6: 1-55. http://dx.doi.org/10.1080/10705519909540118

Jaccard, J., & Wan, C. K. 1996. LISREL approaches to interaction effects in multiple regression. Thousand Oaks, CA: Sage.

Jones, T. M. 1991. Ethical decision making by individuals in organizafions: An issue- contingent model. Academy of Management Review, 16: 366-95.

Jones, T. M., Eelps, W., & Bigley, G. A. 2007. Ethical theory and stakeholder-related de- cisions: The role of stakeholder culture. Academy of Management Review, 32: 137-55. http://dx.doi.org/10.5465/AMR.2007.23463924

Kaptein, M. 2009. Ethics program and ethical culture: A next step in unraveling their multi- faceted relationship. Journal of Business Ethics, 89: 261-81. http://dx.doi.org/10.1007/sl0551-008-9998-3

2010. The ethics of organizations: A longitudinal study of the U. S. working population. Journal of Business Ethics, 92: 601-18. http://dx.doi.org/10.1007/sl0551-009-0175-0

201 la. From inaction to extemal whistleblowing: The influence of the ethical culture. Journal of Business Ethics, 98: 513-30. http://dx.doi.org/10.1007/sl0551-010-0591-l

2011b. Toward effective codes: Testing the reladonship with unethical behavior. Journal of Business Ethics, 99: 233-51.

Key, S. 1999. Organizational ethical culture: Real or imsiginedl Journal of Business Ethics, 20: 217-25. http://dx.doi.Org/10.1023/A:1006047421834

Kish-Gephart, J. J., Harrison, D. A., & Treviño, L. K. 2010. Bad apples, bad cases, and bad barrels: Meta-analytic evidence about sources of unethical decisions at work. Jour- nal of Applied Psychology, 95: 1-31. http://dx.doi.org/10.1037/a0017103

Kohlberg, L. 1981. The philosophy of moral development. San Francisco: Harper and Row.

Laufer, W. S. 1999. Corporate liability, risk shifting, and the paradox of compliance. Vanderbilt Law Review, 54: 1343.

Is FORMAL ETHICS TRAINING MERELY COSMETIC? 113

2008. Corporate bodies and guilty minds: The failure of corporate criminal liability. Chicago: University of Chicago Press.

Laufer, W. S., & Robertson, D. C. 1997. Corporate ethics initiatives as social control. Jour- nal of Business Ethics, 16: 1029-48. http://dx.doi.Org/10.1023/A:1017965820673

McCabe, D. L., Treviño, L. K., & Butterfield, K. D. 1996. The influence of collegiate and corporate codes of conduct on ethics-related behavior in the workplace. Business Ethics Quarterly, 6: 461-76. http://dx.doi.org/10.2307/3857499

McGinn, K. L., & Croson, R. 2004. What do communication media mean for negotiators? A question of social awareness. In M. J. Gelfand & J. M. Brett (Eds.), The hand- book of negotiation and culture: 334-49. Stanford, CA: Stanford University Press.

McKendall, M., DeMarr, B., & Jones-Rikkers, C. 2002. Ethical compliance programs and corporate illegality: Testing the assumptions of the corporate sentencing guidelines. Journal of Business Ethics, 'il: 367-83. http://dx.doi.Org/10.1023/A:1015287823807

Miceli, M. P., & Near, J. P. 1984. The relationships among beliefs, organizational posi- tion and whistle-blowing status: A discriminant analysis. Acode/wj of Management Journal, 27: 687-705. http://dx.doi.org/10.2307/255873

2002. What makes whistle-blowers effective? Three field studies. Human Re- lations, 55: 455-79. http://dx.doi.org/10.1177/0018726702055004463

Morrison, E. W., & Milliken, F. J. 2000. Organizational silence: A barrier to change and development in a pluralistic world. Academy of Management Review, 25: 706-25.

Morrison, E. W., & Robinson, S. L. 1997. When employees feel betrayed: A model of how psychological contract violation develops. Academy of Management Review, 22: 226-56.

Near, J. P., & Miceli, M. P. 1995. Effective whistle-blowing. Academy of Management Review, 20: 679-708.

Near, J. P., Rehg, M. T., Scotter, J. R. V., & Miceli, M. P 2004. Does type of wrongdoing affect the whistle-blowing process? Business Ethics Quarterly, 14: 2 1 9 ^ 2 . http://dx.doi.org/10.5840/beq200414210

Nicholson, N., & Johns, G. 1985. The absence culture and the psychological contract: Who's in control of absence? Acade/nj of Management Review, 10: 3 9 7 ^ 0 7 .

O'Fallon, M. J., & Butterfield, K. D. 2005. A review of the empirical ethical decision- making Uterature: 1996-2003. Journal of Business Ethics, 59: 3 7 5 ^ 1 3 .

Paine, L. S. 1994. Managing for organizational integrity. Harvard Business Review, 11: 106-17.

1996. Moral thinking in management: An essential capability. Business Ethics Quarterly, 6: 477-92. http://dx.doi.org/10.2307/3857500

Palmer, D. E., & Zakhem, A. 2001. Bridging the gap between theory and practice: Using the 1991 federal sentencing guidelines as a paradigm for ethics training. Journal of Business Ethics, 29: 77-84. http://dx.doi.Org/10.1023/A:1006471731947

Pamental, G. 1991. The course in business ethics: Why don't the philosophers give busi- ness students what they need? Business Ethics Quarterly, 1: 385-93. http://dx.doi.org/10.2307/3857604

114 BUSINESS E T H I C S QUARTERLY

Podsakoff, P. M., MacKenzie, S. M., Lee, J., & Podsakoff, N. P 2003. Common method variance in behavioral research: A critical review of the literature and recommended Kmedi&s. Journal of Applied Psychology, 88: 879-903. http://dx.doi.Org/10.1037/0021-9010.88.5.879

Podsakoff, P. M., & Organ, D. W. 1986. Self-reports in organizational research: Problems and prospects. Journal of Management, 12: 69-82. http://dx.doi.org/10.1177/014920638601200408

Rest, J. R. 1986. Moral development: Advances in research and theory. New York: Praeger.

Reynolds, S. J. 2006. A neurocognitive model of the ethical decision-making process: Implications for study and practice. Journal of Applied Psychology, 91: 7 3 7 ^ 8 . http://dx.doi.Org/10.1037/0021-9010.91.4.737

2008. Moral attentiveness: Who pays attention to the moral aspects of life? Journal of Applied Psychology, 92: 1027-41. http://dx.doi.Org/10.1037/0021-9010.93.5.1027

Reynolds, S. J., & Bowie, N. E. 2004. A Kantian perspective on the characteristics of ethics programs. Business Ethics Quarterly, 14: 275-92. http://dx.doi.org/10.5840/beq200414214

Reynolds, S. J., Leavitt, K., & DeCelles, K. A. 2010. Automatic ethics: The effects of implicit assumptions and contextual cues on moral behavior. Journal of Applied Psychology, 95: 752-60. http://dx.doi.org/10.1037/a0019411

Richards, C. 1999. The transient effects of limited ethics training. Journal of Education for Business, 74: 332-34. http://dx.doi.org/10.1080/08832329909601706

Robinson, S. L. 1996. Trust and breach of the psychological contract. Administrative Sci- ence Quarterly, 41: 574-99. http://dx.doi.org/10.2307/2393868

Robinson, S. L., Kraatz, M. S., & Rousseau, D. M. 1994. Changing obligations and the psychological contract: A longitudinal study. Academy of Management Journal, 37: 137-52. http://dx.doi.org/10.2307/256773

Rockmann, K. W., & Northcraft, G. B. 2008. To be or not to be trusted: The influence of media richness on defection and deception. Organizational Behavior and Human Decision Processes, 107: 106-22. http://dx.doi.Org/10.1016/j.obhdp.2008.02.002

Rousseau, D. M., & Parks, J. M. 1993. The contracts of individuals and organizations. Research in Organizational Behavior, 15: 1-43.

Schein, E. H. 1990. Organizational culture. A/nmcan Psychologist, 45: 109-19. http://dx.doi.Org/10.1037/0003-066X.45.2.109

1996. Culture: The missing concept in organization studies. Administrative Science Quarterly, 41: 229-40. http://dx.doi.org/10.2307/2393715

Schneider, B., Salvaggio, A. N., & Subirats, M. 2002. Service climate: A new direction for climate research. Journal of Applied Psychology, 87: 220-29. http://dx.doi.Org/10.1037/0021-9010.87.2.220

Sekerka, L. E. 2009. Organizational ethics education and training: a review of best prac- tices and their application. International Journal of Training and Development, 13: 77-95. http://dx.doi.Org/10.llll/j.1468-2419.2009.00319.x

Is EoRMAL E T H I C S TRAINING M E R E L Y COSMETIC? 115

Shu, L. L., & Gino, F. 2012. Sweeping dishonesty under the rug: How unethical actions lead to forgetting of moral rules. Journal of Personality and Social Psychology, 102: 1164-77. http://dx.doi.org/10.1037/a0028381

Shu, L. L., Gino, F., & Bazerman, M. H. 2011. Dishonest deed, clear conscience: When cheat- ing leads to moral disengagement and motivated forgetting. Personality and Social Psychology Bulletin, 37: 330-49. http://dx.doi.org/10.1177/0146167211398138

Smith-Crowe, K., Tenbrunsel, A. E., Chan-Serafin, S., Brief, A. P., Umphress, E. E., & Joseph, J. Forthcoming. The ethics "fix": When formal systems make a difference. Journal of Business Ethics.

Sparks, J. R., &. Hunt, S. D. 1998. Marketing researcher ethical sensitivity: Conceptu- alization, measurement, and exploratory investigation. Journal of Marketing, 62: 92-109. http://dx.doi.org/10.2307/1252163

Stansbury, J., & Barry, B. 2007. Ethics programs and the paradox of control. Business Eth- ics Quarterly, 17: 239-61. http://dx.doi.org/10.5840/beq200717229

Tenbrunsel, A. E., Diekmann, K. A., Wade-Benzoni, K. A., & Bazerman, M. H. 2010. The ethical mirage: A temporal explanation as to why we aren't as ethical as we think we are. Research in Organizational Behavior, 30: 153-73. http://dx.doi.Org/10.1016/j.riob.2010.08.004

Tenbrunsel, A. E., & Smith-Crowe, K. 2008. Ethical decision making: Where we've been and where we're going. Academy of Management Annals, 2: 545-607.

http://dx.doi.org/10.1080/19416520802211677

Tenbrunsel, A. E., Smith-Crowe, K., & Umphress, E. E. 2003. Building houses on rocks: The role of the ethical infrastructure in organizations. Social Justice Research, 16: 185-307. http://dx.doi.Org/10.1023/A:1025992813613

Trevifio, L. K. 1986. Ethical decision making in organizations: A person-situation interac- tionist model. Academy of Management Review, 11: 601-17.

1990. A cultural perspective on changing and developing organizational eth- ics. In R. Woodman & W. Passmore (Eds.), Research in organizational change and development: 195-230. Greenwich, CT: JAI Press.

Treviño, L. K., & Ball, G. A. 1992. The social implications of punishing unethical be- havior: Observers' cognitive and affective reactions. Journal of Management, 18: 751-68. http://ax.doi.0rg/lO.l 177/014920639201800409

Treviño, L. K., Butterfield, K. D., & McCabe, D. L. 1998. The ethical context in organiza- tions: Influences on employee attitudes and behaviors. Business Ethics Quarterly, 8: 447-76. http://dx.doi.org/10.2307/3857431

Treviño, L. K., & Victor, B. 1992. Peer reporting of unethical behavior: A social context perspective. Academy of Management Journal, 35: 38-64. http://dx.doi.org/10.2307/256472

Treviño, L. K., & Weaver, G. R. 2001. Organizational justice and ethics program "follow- through": Influences on employees' harmful and helpful behavior. Business Ethics Quarterly, 11: 651-71. http://dx.doi.org/10.2307/3857765

2003. Managing ethics in business organizations: Social scientific perspec- tives. Stanford, CA: Stanford University Press.

116 BUSINESS E T H I C S QUARTERLY

Treviño, L. K., Weaver, G. R., & Brown, M. E. 2008. It's lovely at the top: Hierarchical lev- els, identities, and perceptions of organizational ethics. Business Ethics Quarterly, 18: 233-52. http://dx.doi.org/10.5840/beq200818217

Treviño, L. K., Weaver, G. R., Gibson, D. G., & Toffler, B. L. 1999. Managing ethics and legal compliance: What works and what hurts. California Management Review, 41: 131-51. http://dx.doi.org/10.2307/41165990

Treviño, L. K., Weaver, G. R., & Reynolds, S. J. 2006. Behavioral ethics in organizations: A review. Journal of Management, 32: 951-90. http://dx.doi.org/10.1177/0149206306294258

Treviño, L. K., & Youngblood, S. A. 1990. Bad apples in bad barrels: A causal analysis of ethical decision-making behavior. Journal of Applied Psychology, 75: 378-85. http://dx.doi.Org/10.1037/0021-9010.75.4.378

Tsui, A. S., Ashford, S. J., Clair, L., & Xin, K. R. 1995. Dealing with discrepant expecta- tions: Response strategies and managerial effectiveness. Academy of Management Journal, 38: 1 5 1 5 ^ 3 . http://dx.doi.org/10.2307/256842

Tyler, T. R., & Blader, S. L. 2005. Can business effectively regulate employee conduct? The antecedents of rule following in work settings. Academy of Management Jour- nal, 6: 1143-58. http://dx.doi.org/10.5465/AMJ.2005.19573114

United States Sentencing Commission. 1991. United States sentencing commission: Guide- Hnes manual. http://www.ussc.gov/Guidelines/1991_guidelines/Manual_PDF/1991 _Guidelines_Manual_Full.pdf.

2004. United States sentencing commission: Guidelines manual. http://www .ussc.gov/Guidelines/2004_guidelines/Manual/gl2004.pdf.

Vadera, A. K., Aguilera, R. V., & Caza, B. B. 2009. Making sense of whistle-blowing's an- tecedents: Learning from research on identity and ethics programs. Business Ethics Quarterly, 19: 553-86. http://dx.doi.org/10.5840/beq200919432

Victor, B., & CuUen, J. N. 1988. The organizational bases of ethical work climates. Admin- istrative Science Quarterly, 33: 101-25. http://dx.doi.org/10.2307/2392857

Victor, B., Treviño, L. K., & Shapiro, D. L. 1993. Peer reporting of unethical behavior: The influence of justice evaluations and social context factors. Journal of Business Ethics, 12: 253-63. http://dx.doi.org/10.1007/BF01666528

Warren, D. E. 2003. Constructive and destructive deviance in organizations. Academy of Management Review, 28: 622-32.

Warren, D. E., & Smith-Crowe, K. 2008. Deciding what's right: The role of external sanc- tions and embarrassment in shaping moral judgment in the workplace. Research in OrgflMizaftona/ßeÄavior, 28: 81-105. http://dx.doi.Org/10.1016/j.riob.2008.04.004

Weaver, G. R., & Treviño, L. K. 1999. Compliance and values oriented ethics programs: Influences on employees' attitudes and behavior. Business Ethics Quarterly, 9: 315-35. http://dx.doi.org/10.2307/3857477

Weaver, G. R., Treviño, L. K., & Cochran, P. L. 1999a. Corporate ethics practices in the niid-1990's: An empirical study of the Fortune 1000. Journal of Business Ethics, 18: 283-94. http://dx.doi.Org/10.1023/A:1005726901050

Is FORMAL E T H I C S T R A I N I N G M E R E L Y C O S M E T I C ? 117

1999b. Corporate ethics programs as control systems: Influences of executive commitment and environmental factors. Academy of Management Journal, 42: 41-57. http://dx.doi.org/10.2307/256873

i- I 1999c. Integrated and decoupled corporate social performance: Management

commitments, extemal pressures, and corporate ethics practices. Academy of Man- agement Journal, 42: 539-52. http://dx.doi.org/10.2307/256975

Weber, J. A. 1990. Measuring the impact of teaching ethics to future managers: A review, assessment, and recommendations./oMrna/o/ßHsi/iess £i/iics, 9: 183-90. http://dx.doi.org/10.1007/BF00382643

Wheaton, B., Muthen, B., Alwin, D., & Summers, G. F. 1977. Assessing the rehability and stability in panel models. In D. R. Heise (Ed.), Sociological methodology: 84-136. San Francisco: Jossey-Bass.

Wolf, T. M., & Cheyne, A. 1972. Persistence of effects of live behavioral, televised behav- ioral, and live verbal models on resistance to deviation. Child Development, 43: 1429-36. http://dx.doi.Org/10.2307/l 127530

Copyright of Business Ethics Quarterly is the property of Philosophy Documentation Center and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.