஽ Academy of Management Review 2011, Vol. 36, No. 4, 663–685. http://dx.doi.org/10.5465/amr.2010.0128

MORAL MATURATION AND MORAL CONATION: A CAPACITY APPROACH TO EXPLAINING MORAL THOUGHT AND ACTION

SEAN T. HANNAH West Point–United States Military Academy

BRUCE J. AVOLIO University of Washington

DOUGLAS R. MAY University of Kansas

We set out to address a gap in the management literature by proposing a framework specifying the component capacities organizational actors require to think and act morally. We examine how moral maturation (i.e., moral identity, complexity, and metacognitive ability) and moral conation (i.e., moral courage, efficacy, and owner- ship) enhance an individual’s moral cognition and propensity to take ethical action. We offer propositions to guide future research and discuss the implications of the proposed model for management theory and practice.

There is growing recognition that organiza- velop the moral capacity of employees, the tim- tions are operating in increasingly more com- ing seems propitious to offer a theoretical plex and often global environments that inher- framework examining what constitutes the ently impose difficult moral challenges on moral capacity of individuals in the workplace organization members (e.g., Donaldson, 2003; and how that capacity drives the way individu- George, 2007; Hannah, Uhl-Bien, Avolio, & Ca- als respond to ethical challenges. Conse- varretta, 2009a). This increased complexity is quently, our primary motivation for writing this due to a number of factors, including greater article is to offer a comprehensive and testable scrutiny over individuals’ actions, more de- theoretical framework to serve as the basis for mands for transparency, the necessity for orga- guiding future research and practice concerning nizations to work across competing governmen- the moral capacities needed to process a moral tal and legal systems, expanded organizational challenge from recognition to action. stakeholders with competing interests, and the We suggest below that there is a need for a need to operate across different cultures that new and expanded theory of moral1 develop- constitute diverse sets of values. ment that better explains how individuals con- Paralleling a rise in the complexity of organi- sider and act on moral dilemmas and tempta- zational challenges, popular belief suggests tions. Yet we realize our position will not be that the scope and scale of greed and malfea- accepted without challenge, since some may be- sance in organizations are escalating (George, lieve the “gold standard” already exists in the 2007). In response to these trends, a growing well-known cognitive development models of number of organizations and governments are Jean Piaget (1965/1932), Lawrence Kohlberg establishing ethics offices and developing new (Kohlberg, 1981; Kohlberg & Candee, 1984), and ethical policies and mandatory ethics training James Rest and colleagues (Rest, Narvaez, Be- (Donaldson, 2003). As organizations strive to de- beau, & Thoma, 1999; see also Rest, 1986, 1994). We build on these earlier theoretical frame- works in our proposed model by starting our We thank Professor Jean-Philippe Bonardi and three anonymous reviewers for their guidance and numerous im- portant recommendations to improve this manuscript. We also thank Professor James Campbell (Jim) Quick, who pro- 1 In this article we treat the terms moral and ethical as vided valuable input on an early version of the paper. synonyms.

663 Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder’s express written permission. Users may print, download, or email articles for individual use only. 664 Academy of Management Review October discussion noting their strengths and limita- Second, previous theories have focused pri- tions. We then proceed by proposing a new the- marily on aspects of moral judgment while not oretical moral capacity framework that can be adequately explaining the capacities needed to used to examine how individuals process and carry those judgments through to intentions and respond to moral problems, as well as by focus- actions. For example, while Kohlberg (1981) fo- ing on how those capacities can be developed. cused exclusively on explaining the core as- First, in prior theory and research scholars pects of moral judgment, Rest et al.’s four- have focused on describing the processes component model suggests that moral whereby individuals handle moral incidents, judgments must also be preceded by moral but they have not adequately explained the un- awareness and followed by the formation of derlying capacities individuals require to effec- intentions to act and then, ultimately, action tively enact those processes. For example, Rest itself. Rest et al. conclude, however, that be- and colleagues (1999) made extensive contribu- sides some “forays into studying Components tions to the literature on moral psychology by 3 and 4,” little work has been done to explain developing their four-component model. This moral motivation or intentions and moral ac- model identifies four “inner psychological pro- tion and that “we believe that the overall prog- cesses [that] together give rise to outwardly ob- ress in the larger enterprise of moral psychol- servable behavior” (1999: 101)—moral sensitiv- ogy can be viewed in terms of how well ity, moral judgment, moral motivation, and research progresses in all four inner psycho- moral action. These four processes have served logical components leads to outwardly observ- as an important organizing framework and able behavior” (1999: 102). starting point for ethics research and practice. Third, prior research has highlighted the im- To understand our intended theoretical contri- portance of individual differences in ethical pro- bution, however, it is important to make several cessing, but that work has focused primarily on distinctions between Rest et al.’s as well as oth- a limited set of stable individual traits. Treviño ers’ approach to examining ethical processes (1986), for example, proposed that locus of con- and our approach to exploring ethical capaci- trol, ego strength, and field dependence would ties. Rest et al. state that the moral sensitivity moderate the linkage between ethical cognition process, for example, entails steps such as “in- and behavior. Although these are all important terpreting the situation, role-taking how various individual stable traits that can be used to po- actions would affect the parties concerned, tentially explain ethical processing in part, we imagining cause-effect chains of events, and be- focus on malleable individual capacities that ing aware that there is a moral problem when it can be developed to enhance one’s ethical cog- exists” (1999: 101). Research has shown that in- nition and behavior/actions. dividuals vary in their level of proficiency re- Fourth, a useful practical model of moral ca- garding each of these four processes (e.g., Be- pacities must explain both moral cognition and beau, 2002). Yet prior research has not moral conation or the impetus to act. We define sufficiently examined the individual capacities moral conation as the capacity to generate re- that explain the variance across individuals en- sponsibility and motivation to take moral action abling someone to effectively execute the steps in the face of adversity and persevere through or actions related to these four processes. Based challenges. The theories of Kohlberg, Rest et al., on Rest et al.’s description of moral sensitivity, and others have focused largely on the cognitive for example, we might ask, “What enables a processing of moral dilemmas—those complex person to execute the steps in interpreting a intellectual choices between right versus wrong, moral challenge and then estimating cause and or right versus right. Such models do not ad- effect chains better than another?” At the pres- dress how individuals process moral tempta- ent time prior theory and research do not ade- tions (Monin, Pizarro, & Beer, 2007), where they quately address this or similar questions. We know what is best but one personal value con- therefore set out here to identify the specific flicts with another. In this situation the individ- individual capacities that help account for the ual must have the adequate self-regulatory ca- level of variation across individuals in terms of pacity to resist one action in favor of another how they process, formulate judgments about, action. A person may be tempted, for example, to and respond to moral challenges. participate in unethical acts performed by his or 2011 Hannah, Avolio, and May 665 her group to achieve peer acceptance, requiring by Rest et al. (1999), that accounts for the capac- the conation to resist such temptation and to ities needed for ethical awareness and decision stand up and take action against those who making, as well as the conation to follow have done the wrong thing.2 through with action. Fifth, more recent theories have been devel- oped that attempt to explain those capacities TOWARD AN EXPANDED MODEL OF needed by moral actors at specific stages of MORAL CAPACITIES processing a moral issue, such as Reynolds’ (2008) theory of moral attentiveness. However, Measuring attitudes and judgments in gen- what is missing in the literature is a more inte- eral (Ajzen, 1991), as well as ethical judgments grated model of individual moral capacities that in particular (Bebeau, 2002; Blasi, 1980; Treviño & can fully explain how moral dilemmas are Youngblood, 1990), has generated relatively thought about and acted on across all four of weak relationships in predicting actual ethical Rest et al.’s processes. Supporting this need, behavior. Despite this fact, Reynolds (2006) notes Rest and colleagues note that “although most that ethics research has tended to focus predom- researchers would agree that there is much di- inantly on ethical judgment models versus eth- versity of constructs, processes, phenomena, ical behavior, which is a trend that is also evi- and starting points for the psychology of moral- dent in recent literature reviews (O’Fallon & ity, the greater challenge is to formulate how all Butterfield, 2005; Tenbrunsel & Smith-Crowe, these different parts fit together” (1999: 6). 2008; Treviño, Weaver, & Reynolds, 2006). This In sum, our principle goal is to provide re- trend has persisted even though Rest et al. (1999: searchers with a conceptual framework that in- 101) reported that the Defining Issues Test cludes the breadth of moral capacities required (DIT)—the most commonly used measure of cog- by moral actors to think about and act on a nitive moral judgment capacity—typically only moral dilemma. Specifically, our proposed explains approximately 20 percent of the vari- model seeks to achieve the following objectives: ance in actual ethical behavior. (1) to provide an organizing structure for an ex- Kohlberg (1981, 1986) based his model of cog- panded set of constructs required to be devel- nitive moral development on the work of Piaget oped in moral actors, while also explaining the (1965/1932) and proposed that moral develop- relationships and processes linking these con- ment is a maturation process that unfolds across structs; (2) to recognize both the processes moral the lifespan in six stages, whereby more com- actors must use and the underlying capacities plex “mental operations” related to the logic of they require to effectively enact those processes; morality are developed. Kohlberg originally at- and (3) to use what we have learned from the tempted to explain overall moral development first two objectives to facilitate assessment and as stemming from cognitive development but, in development by offering measurable and mal- response to critics (e.g., Gilligan, 1982), in later leable constructs for future research and prac- work narrowed the boundaries of his develop- tice in this area. mental theory to only apply to moral judgments In developing the proposed model, we exam- dealing with justice issues.3 ined relevant frameworks from clinical, social, In arguing for a “neo-Kohlbergian approach,” and developmental psychology and neurosci- Rest and colleagues (1999) suggested that be- ence, as well as leadership, organizational be- yond the limitations of solely focusing on jus- havior, and ethics. We assessed where each tice, Kohlberg’s theory and scoring system was framework’s boundary ended in contributing to too “macro-moral” in that it addressed abstract our understanding of moral capacity and where another framework with compatible logic be- gan. We believe that our contribution to moral/ 3 Kohlberg states that “the research programme of myself ethical theory and practice involves advancing and my Harvard colleagues has moved from restricting the a broader “enterprise” approach, as called for study of morality to the study of moral development to re- stricting it to the study of moral judgment...torestricting it to the form or cognitive-structural stage of moral judgment as embodied in judgments of justice”; he further states that 2 For parsimony we use “dilemma” from here on to repre- he does not imply that “these restrictions should guide all sent all moral problems, predicaments, and temptations. fruitful moral psychology research” (Kohlberg, 1986: 499). 666 Academy of Management Review October aspects of society (e.g., societal norms of justice) stand up and act in the face of adversity while versus the more commonly faced moral rela- another person of equal cognitive capacity will tions between individuals, friends, groups, and fail to act morally (Bandura, 1991). organizations. Rest et al. (1999) noted that theo- Therefore, we set out to extend this prior the- ries need to take these more relevant, frequent, ory and research by first identifying personal and typically more complex moral issues into capacities related to both moral thought and account. conduct. Next, we examine how they operate as Rest et al. (1999) also argued against the “hard determinants that influence an individual’s stages” associated with development recom- moral behavior. mended by Kohlberg, basing their model in- stead on the concept of cognitive schemas. The Overview of the Model’s Core Components authors noted that Kohlberg’s theory only ad- dressed component 2 (moral judgment) of their We provide an overview of the components, four-component model. Yet despite this criticism presented in Figure 1, constituting moral matu- of Kohlberg’s work, Rest and colleagues’ own ration and conation, and in subsequent sections work focused largely on component 2, develop- we provide more in-depth analysis of each com- ing and validating the DIT, which purports to ponent. At the base of the model is a depiction of measure levels of moral schema development the four psychological processes as proposed by using hypothetical moral judgment exercises. Rest et al. (1999). Moral sensitivity includes pro- Thus, while the models of Kohlberg and Rest et cesses related to being aware of a moral prob- al. foremost attempt to address one aspect lem, interpreting the situation, and identifying (schema development) of the cognitive capaci- various options to address the problem. Moral ties that individuals need to recognize and judgment concerns processes taken to deter- judge moral issues, they do not help explain the mine what action is the most proper to pursue. self-regulatory capacities that promote how an Moral motivation entails processes geared to- individual engages his or her full cognitive ca- ward gaining commitment to a given action and pacities in a given moral dilemma. Nor do those the weight assigned to specific moral values models attempt to explain why one person will over other values. Finally, moral action involves

FIGURE 1 Framework for Moral Maturation and Moral Conation

Experience, reflection, and feedback

Moral maturation capacities Moral conation capacities Meta- Moral Moral Moral Moral Moral cognitive complexity identity ownership efficacy courage ability

P1 P2 P3 P5 P6 P7 P8 P4 P5

Moral cognition processes Moral conation processes

Moral Moral Moral Moral sensitivity judgment motivation action 2011 Hannah, Avolio, and May 667 persistence in a moral task, overcoming fatigue lier defined moral conation as the capacity to and various temptations and challenges to take generate responsibility and motivation to take the appropriate action. moral action in the face of adversity and perse- While some of these four component labels vere through challenges. read akin to developed capacities, Rest et al. have defined them as psychological processes related to sequencing or addressing a moral Dynamics of the Model dilemma. Indeed, the four components were first developed to categorize ethics theories and Before proceeding, it is important to clarify serve as a “heuristic tool in conceptualizing the three aspects of our theoretical framework. First, psychology of morality as a whole” (Rest et al., we propose that each capacity is necessary but 1999: 101). Yet while serving as a valuable heu- not sufficient for moral decision making and ristic, we believe this framework suffers from behavior. High levels of moral complexity, for some important gaps. The moral motivation example, must be accompanied by sufficient component, for example, denotes a psychologi- metacognitive ability to process complex moral cal process where commitment to action is gen- knowledge, along with a moral identity guided erated but does not specify what individual dif- by self-regulation and processing that achieves ferences drive that process or the level of a desired moral outcome. commitment needed to behave in a moral man- Second, our propositions suggest that moral ner. We therefore seek to advance Rest’s work maturation capacities will primarily drive moral (1986, 1994) by articulating the moral capacities cognitive processes, while moral conation ca- displayed in the upper half of Figure 1 that can pacities will primarily drive moral motivational serve to explain why individuals are more or or conative processes. Yet we also consider in less inclined and able to effectively execute our discussion of the proposed model that these those four processes. capacities may also have some influence on To organize our proposed model, we group other stages of Rest et al.’s four-component Rest et al.’s components of moral sensitivity and model of moral decision making. Most notably, moral judgment into moral cognition processes as shown in Figure 1, we explicitly propose that since they both entail the awareness and pro- moral identity is unique in that it will drive both cessing of information pertaining to moral is- moral cognition and one’s motivation or sues. Further, we group the moral motivation conation to act. For example, an individual who and moral action components into what we term has a very highly developed sense of moral moral conation processes since they both entail identity will more likely exhibit ethical behav- the tendency for and the practice of moral be- iors that are more in line with his or her moral havior. identity, thus enhancing the individual’s moral Similarly, we use the category labels moral conation to act (e.g., Aquino & Reed, 2002; Blasi, maturation and moral conation to group the six 1993; Weaver, 2006). However, we also note that moral capacities of our model for ease of con- the judgments made and actions taken may also ceptualization and description. We are not, how- vary across different moral issues and domains, ever, proposing higher-order latent constructs. even where the individual has a highly devel- Accordingly, we first define each capacity and oped moral identity. then offer individual propositions related to Finally, as suggested by the shaded dotted each of the six capacities in our model. line “Experience, reflection, and feedback” in As shown in the upper portion of Figure 1, we Figure 1, we propose that these six capacities suggest that the three constructs labeled moral are all open to development, which goes beyond maturation capacities are critical in driving earlier literature primarily focusing on identify- moral cognition processes. We define moral ing stable traits and attributes that may influ- maturation as the capacity to elaborate and ef- ence moral judgments and actions. Conse- fectively attend to, store, retrieve, process, and quently, we provide in our discussion some make meaning of morally relevant information. preliminary suggestions on how these capaci- Next, we suggest that the three constructs la- ties could best be measured and developed to beled moral conation capacities are critical in help initiate and guide future research in this driving moral motivational processes. We ear- area. 668 Academy of Management Review October

In order to provide sufficient space to fully knowledge domain, depending on their level of define the six capacities included in our model, development. Greater complexity in a given do- we do not spend a great deal of time focusing on main is composed of highly differentiated and contextual factors that might influence these re- richly connected mental representations that spective capacities. However, we fully recognize the individual can call on to allow him or her to that the context, including the characteristics, process information in greater depth and with culture, and climate of the organization more elaboration (Rafaeli-Mor & Steinberg, 2002; (Treviño, 1986; Treviño & Youngblood, 1990), and Streufert & Nogami, 1989). For example, Be- other factors, such as leadership (Brown, beau’s (2002) work in the dental profession sug- Treviño, & Harrison, 2005; Mayer, Kuenzi, Green- gests that dentists will have varying levels of baum, Bardes, & Salvador, 2009), are also impor- complexity associated with dental ethics. Fur- tant elements that could significantly influence ther, a dentist may have low levels of moral individual moral processing, motivation, and complexity in other domains, such as account- behaviors. We highlight some of these contex- ing ethics or, more generally, biomedical ethics, tual factors that should be included in future lacking adequate knowledge in those areas of research in the discussion section. ethics to process and apply information to a given dilemma with depth and elaboration. Moral complexity is a critical moral capacity MORAL MATURATION because the distinctive dimensions individuals Understanding what constitutes moral matu- use to organize and make meaning of the world ration requires greater refinement, beyond Kohl- strongly influence how they make decisions and berg’s (1981) specific stage model of cognitive behave within a specific domain (Rafaeli-Mor & moral development and Rest et al.’s three levels Steinberg, 2002; Streufert & Nogami, 1989). All of cognitive development (i.e., personal interest, other things being equal, more cognitively com- maintaining norms, and postconventional). Rest plex individuals process information more thor- et al. recognized the coarseness of their ap- oughly because they have more categories to proach, noting, “We recognize that the cognitive discriminate among information received in structures we talk about are somewhere be- their environment and are more able to see com- tween cognitive development stages and social monalities and connections among those cate- schemas” (1999: 185). To refine Rest et al.’s gories (Schroder et al., 1967; Streufert & Nogami, stages we incorporate theories of cognitive com- 1989). This led Hannah, Lester, and Vogelgesang plexity (e.g., Streufert & Nogami, 1989), thereby (2005) to propose that rich moral representations decomposing individuals’ mental representa- will help individuals achieve greater coherence tions of moral knowledge into more refined as- when processing complex moral dilemmas. pects of content and structure to help better ex- Greater moral complexity provides a larger and plain what constitutes cognitive moral more developed set of prototypes with which to maturation (e.g., Street, Douglas, Geiger, & Mar- process moral information, and these prototypes tinko, 2001). are drawn on during either controlled or auto- matic processing of ethical challenges/incidents (see Reynolds, 2006; Sonenshein, 2007). There- Complexity of Moral Representations fore, the concept of moral complexity underpins We know that individuals are more or less what Werhane (1999) calls “moral imagination,” complex in their mental representations of var- or the ability to understand the various dimen- ious domains of knowledge based on their sions of moral dilemmas and develop various breadth of experience and learning across the moral “realities” to consider and with which to life span (Bandura, 1991; Schroder, Driver, & create imaginative solutions. This occurs be- Streufert, 1967; Streufert & Nogami, 1989), includ- cause more cognitively complex individuals are ing in different moral domains (Narvaez, 2010; better able to acquire and make sense of com- Swanson & Hill, 1993). Here we use the term peting information, while spending more time domain to mean a specific area of ethics (e.g., interpreting a broader range of information to accounting ethics, medical ethics, military eth- help resolve dilemmas (e.g., Bower & Hilgard, ics, or parenting ethics). Individuals can have 1981; Dollinger, 1984). Indeed, in their work with more or less complex representations of each adolescents, Swanson and Hill (1993) found that 2011 Hannah, Avolio, and May 669 richer moral knowledge predicted higher moral complex knowledge (Margolis & Phillips, 1999; reasoning. Narvaez (2010) suggests that more Shweder, 1991). In sum, the extent to which indi- complex moral actors will also have greater viduals possess moral complexity in a given “negative expertise” in that they will know what domain will enhance their level of moral sensi- actions not to take when facing a moral di- tivity and ability to differentiate and integrate lemma. (i.e., elaborate) moral information in that do- Beyond enhancing the processing of moral main, particularly since people tend to use the judgments, moral complexity should also en- highest stages of cognitive development avail- hance moral sensitivity (i.e., stage 1 of Rest et able to them when processing moral dilemmas, al.’s four-stage model). Reynolds notes that indi- if ample motivation exists (Treviño, 1992). This viduals differ from one another in moral atten- leads to our first proposition. tiveness, which he proposes enables greater sensitivity to moral issues, and he calls for fu- Proposition 1: Higher (lower) levels of ture research to determine the “origins of moral complexity in a specific domain(s) of attentiveness” (2008: 1039). In response to Reyn- moral knowledge will be associated olds call, we suggest that moral complexity is a with higher (lower) levels of (a) moral central antecedent to moral attentiveness be- sensitivity and (b) elaborative moral cause individuals have a heightened propensity judgments in that domain(s). to attend to information that is consistent with their mental representations, while discounting incongruent information (Dutton & Jackson, Metacognitive Ability 1987). This suggests that the more distinct inter- nal dimensions or prototypes people possess to The next moral maturation capacity in our perceive “moral cues,” the more likely and able model is metacognitive ability. We include this they will be to perceive and attend to moral capacity to explain the variance in individual indicators, when present. ability to use moral complexity to enhance Moral judgments are also inherently contex- moral cognition (Metcalfe & Shimamura, 1994; tualized. Groups, organizations, and societies Narvaez, 2010). We propose that moral complex- seek to function on a shared set of values (Victor ity provides a deeper understanding of what & Cullen, 1988), and yet those values vary, per- constitutes moral maturation, thus going be- haps extensively, across different collectives yond the general conceptualizations of Kohl- (Margolis & Phillips, 1999; Shweder, 1991). Even berg’s or Rest et al.’s stages. A high level of when individuals agree on sets of values or complexity is, however, like fuel without an en- moral standards, the meanings and applica- gine to process that fuel. Street et al. (2001) argue tions of those standards are often contentious that individuals also need the capacity or “en- (Sonenshein, 2007). Given this equivocality, we gine” to deeply process complex moral knowl- do not offer prescriptions as to any specific edge. As with other constructs in our proposed knowledge content comprising moral complex- model, we focus on metacognitive ability as a ity as being more or less “moral”—only that developed capacity underlying such depth of greater complexity in a given moral domain(s) processing (Petty & Cacioppo, 1986), and we sug- will drive more elaborate moral judgments con- gest some techniques to measure and develop cerning that domain. We define elaborate moral this construct. judgment as the extent to which an individual Metacognitive ability is composed of monitor- differentiates and integrates moral information. ing and regulating cognitive processes, thus Moral complexity, therefore, incorporates rich serving both self-referential and executive con- knowledge content representative of the moral- trol functions (Dunlosky & Metcalfe, 2009; Met- ity of a given culture or social group. With this in calfe & Shimamura, 1994), which are critical for mind, we suggest that our proposed model can moral cognition. Metacognitive ability has been be generalized across cultures and organiza- shown to be related to yet distinct from general tions yet needs to be further specified within intelligence or cognitive ability (see Dunlosky & any given culture by incorporating the virtues, Metcalfe, 2009, and Veenman & Elshout, 1999). norms, and mores of that culture to determine Recent neuroscience research has shown that what constitutes moral, immoral, or even amoral these general and more specific cognitive abil- 670 Academy of Management Review October ities each draw from different neural capacities in the use of moral capacity when judging moral (e.g., Jausovec, 2008). issues (Thoma, Rest, & Davidson, 1991). This dis- Overall intelligence or cognitive ability refers tinction may also be particularly useful in ex- to the general capacity to reason and solve plaining how individuals process ethical issues problems, which is distinct from the ability to that force them to address multiple competing regulate and control cognition as these reason- values. For example, higher levels of metacog- ing processes unfold (Dunlosky & Metcalfe, 2009; nitive ability would enhance the monitoring and Metcalfe & Shimamura, 1994). Indeed, research control over what information is being used in has shown that as an individual gains expertise making a judgment, assessments of that infor- in a certain domain, the individual’s metacogni- mation’s accuracy, influences of emotions or tive ability gets progressively more tailored to competing values on judgment, and whether all processing information in that domain. Conse- possible aspects of a moral dilemma have been quently, over time, the direct influence of gen- considered. This leads to our next proposition. eral intellectual ability diminishes as an indi- Proposition 2: Higher (lower) levels of vidual acquires expertise in a domain, while the moral metacognitive ability will be impact of metacognition increases on how that associated with higher (lower) levels individual processes information (Veenman & of (a) moral sensitivity and (b) elabo- Elshout, 1999). This suggests that at higher lev- rative moral judgments. els of development, moral cognitive functioning will be guided less by some general moral in- telligence and more by metacognitive ability Moral Identity that has been tailored to various areas of prior moral experiences—with that metacognitive Another limitation of Rest’s (1986, 1994) and ability drawing on underlying complex moral Kohlberg’s (1981) models is that they focus on knowledge developed within specific domains moral domain knowledge in determining what (i.e., moral complexity). constitutes cognitive moral development (justice We suggest that by acquiring higher levels of concepts in the case of Kohlberg, as noted ear- moral metacognitive ability, individuals will lier). Yet research has shown that in addition to also enhance their moral sensitivity and moral knowledge of concepts of morality, individuals’ judgment. This is because complex moral dilem- knowledge about themselves as moral actors mas require the capacity to select from, access, (i.e., moral identity) is also critical in driving and modify moral knowledge and to apply elab- both moral cognition and moral conation (e.g., orative reasoning to the specific moral dilemma Aquino, Freeman, Reed, Lim, & Felps, 2009). This being confronted in order for an individual to is because self-identity consists of the most acces- achieve a sense of logical coherence. Metacog- sible and elaborate knowledge structures individ- nitive ability provides the executive control uals hold (Hannah, Woolfolk, & Lord, 2009b; Kihl- functions over these processes, determining strom, Beer, & Klein, 2003) and, thus, imposes a what is attended to and recalled by an individ- strong influence on how individuals regulate ual (Metcalfe & Shimamura, 1994), along with thought and control behavior (Carver & Scheier, the selection and employment of mental proto- 1998; Lord & Brown, 2004). Therefore, identity can- types (Reynolds, 2008; Sonenshein, 2007) used not be separated from moral processing, particu- during ethical decision making. Prior research larly since moral and immoral actions all influ- indicates that these metacognitive processes ence one’s self-evaluations (e.g., “Am I a good can become so tailored and habituated that they person?”) and sense of self-consistency (e.g., become automated and triggered by cues in the “What ethical action is most in line with my be- context (e.g., Gollwitzer & Schaal, 1998; Metcalfe liefs about myself?”). & Shimamura, 1994). Where these metacognitive The self-regulatory functions provided by abilities have become habituated, we expect to moral identity are critical in rounding out and see a greater use of automatic moral processing more fully explaining what constitutes moral of ethical dilemmas. maturation. Moral complexity and metacogni- Separating moral complexity from an individ- tive ability combined provide a richer elabora- ual’s ability to process that complexity via meta- tion of moral knowledge, yet these processes cognition helps explain individual differences must be guided by self-regulatory standards 2011 Hannah, Avolio, and May 671 used when processing moral incidents/events global identity. This approach has explicitly as- (Carver & Scheier, 1998; Lord & Brown, 2004). For sumed that (1) individuals possess a unified example, highly complex individuals might moral identity that is distinct from other identi- come up with ways to justify moral disengage- ties they hold, (2) this moral identity can be more ment if such rich processing is not guided by or less central to individuals’ overall identity, self-standards. and (3) situations influence how accessible or We are certainly not the first to propose that salient this moral identity is at any point in time moral identity is critical to moral processing (see Aquino et al., 2009). and behavior (e.g., Aquino & Reed, 2002; Blasi, As we suggested above, we believe that there 1993; Weaver, 2006). For example, Reynolds and are important alternative ways of conceptualiz- Ceranic (2007) demonstrated that moral identity ing moral identity that move beyond viewing it is an important construct in helping to explain as a unified concept, and we use self-complex- the link between moral judgments and behavior. ity theory to support our position (e.g., Hannah Here we suggest that prior work on moral iden- et al., 2009b; Linville, 1987). Consequently, we tity can be expanded in three specific ways. suggest that moral identity is not a separate First, instead of conceptualizing moral identity intact identity, any more than “sociable” is a in isolation, we integrate moral identity with separate identity. Instead, self-complexity the- other interrelated capacities of moral matura- ory suggests that identity content (e.g., moral tion and link moral identity to Rest et al.’s four- or sociable) is instead structured across the component model. various subidentities that make up a person’s Second, we approach moral identity as com- multifaceted identity. These subidentities are posed of more than self-descriptive moral traits. largely based on social roles, such as parent Aquino and Reed’s (2002) construct of moral or team leader, and are developed as actors identity, for example, includes nine Kantian-like perform these roles over time (Markus & Wurf, moral traits, including caring, compassionate, 1987). This suggests that each social role will fair, friendly, generous, helpful, hardworking, be composed of different forms and levels of honest, and kind. Identity, however, is thought to moral identity content. Specifically, we pro- include not only traits but other dynamic struc- pose that the content of moral identity consists tures, including roles, goals and motivation, af- of self-knowledge components (e.g., “What do I fect, and autobiographical narratives, along stand for?” or “What are my core beliefs?”) and with other components (Hill & Roberts, 2010; evaluative components (e.g., “AmIamoral Lord, Hannah, & Jennings, 2011). We therefore person?” or “How well do I stand up for my conceptualize moral identity as a more complex beliefs?”) and their associated sets of goals, structure. affect, self-regulatory plans, etc., whereas Third, we propose that moral identity is not a structure refers to how such content is orga- singular identity structure but is multifaceted nized and categorized across the social roles and represented across various identities (Han- composing one’s overall identity (Woolfolk et nah et al., 2009b). We therefore incorporate self- al., 2004). complexity theory (e.g., Linville, 1987; Woolfolk, One’s overall self-identity is thus elaborate Gara, Allen, & Beaver, 2004) with moral identity and differentiated, actually being more of an theory to help explain the multifaceted nature of assemblage of selves rather than a unified moral identity. We believe that this third contri- whole (Markus & Wurf, 1987). For example, one bution is critical if we are to understand why can see oneself as highly truthful in one’s role moral behavior varies across situations in the as a team leader, but less so in one’s secondary workplace, which is a question that largely has role as a company media spokesperson. Moral been unanswered in the literature (Hardy & identity can thus be defined and measured as Carlo, 2005). more or less complex based on how rich and Prior theories of moral identity have focused differentiated it is as structured across individ- on the content associated with an individual’s uals’ self-identities (Hannah et al., 2009b; Lord et self-concept (i.e., how people see themselves as al., 2011). For instance, in a self-complexity a moral being/actor) and have operationalized study Woolfolk et al. (2004) demonstrated that moral identity by asking participants to respond ethics-related attributes, such as honest, selfish, to the extent various traits apply to their overall scornful, admirable, bad, dependable, and dis- 672 Academy of Management Review October honest, were represented to greater and lesser measures to moral identity will allow research- extents across participants’ various subidenti- ers to assess both differentiation and integra- ties. Further, emerging research has demon- tion/unity, as well as the concomitant effects of strated that individuals’ preferences and values each on the moral behaviors enacted. change when separate subidentities are primed Aspects of moral identity held with higher lev- (LeBoeuf, Shafir, & Bayuk, 2010). For example, els of unity are especially critical in driving Reicher and Levine (1994) showed that when a moral cognition and conation, in that we know scientist versus a student identity was primed, that individuals have higher self-awareness for individuals had more favorable attitudes to- more core and salient aspects of the self (Setter- ward practices such as animal vivisection. To- lund & Niedenthal, 1993). Higher salience also gether, this research suggests that a self- increases moral sensitivity and attentiveness complexity approach will offer a more refined and the rapid processing of moral dilemmas understanding of the multifaceted moral iden- that match patterns or prototypes familiar to an tity, and thereby variance in moral thought and individual (Reynolds, 2008). We also know from behavior across situations and social roles. research on self-verification (Swann, 1983) and Moral identity and the unity of the self. A cybernetic self-regulatory processes (e.g., complex, differentiated moral identity, however, Carver & Scheier, 1998; Lord & Brown, 2004) that does not suggest individuals are devoid of core people are motivated to behave in ways consis- values or consistency in their self-identity. As tent with salient core self-attributes. This sort of discussed earlier, individuals’ level of complex- consistency allows them to maintain the integ- ity is represented both by their differentiation rity of their self-concept and, thus, creates and integration of knowledge—in this case self- conation for self-congruent behavior (Verplan- knowledge. Taking this position helps to ex- ken & Holland, 2002). Consistent with this re- plain how one can differ as a moral actor across search, Stahlberg, Peterson, and Dauenheimer various roles (i.e., differentiation) yet at the (1999) demonstrated that in areas where people same time understand one’s consistency or lack have high self-unity, they tend toward self- thereof across those roles based on certain core verifying motives to confirm their existing self- attributes (i.e., integration). Concerning integra- beliefs in order to establish stability in “who tion, greater moral identity complexity would they are,” whereas in areas of lower unity, they therefore be associated with what has been are more likely to lean toward self-enhancing called “self-unity,” defined as “the extent to motives. which self-beliefs (e.g. perceived personal attri- Because ethical dilemmas typically present butes) are clearly and confidently defined, inter- individuals with competing values and choices, nally consistent, and stable” (Campbell et al., this combined research suggests that in areas 1996: 141). where the self is less “invested” (i.e., low unity), Using justice as an example, a person may individuals may be more likely to swing to other identify him/herself as being a fair individual values and perhaps away from taking moral across a broad range of social roles. These roles action. This suggests that organizations may could then “roll up” and contribute to a more want to develop high levels of unity across their integrated or generalized aspect of self-identity members’ respective subidentities for those crit- as a just person (Hannah et al., 2009b; Hill & ical core organizational values where they want Roberts, 2010). This same person, however, may to ensure that their members will make “the be more differentiated across subidentities right” decision when facing difficult choices. based on other aspects of moral identity. The Moral identity and self-complexity. Aquino et concept of unity thus provides an underlying al. (2009) conceptualized individuals as possess- basis for what Aquino et al. (2009) and others ing a single global moral identity more or less have called the “centrality of moral identity.” central to their overall identity, and that various We suggest that unity is not a separate identity situations influence how accessible this moral as proposed by Aquino et al. but, rather, is the identity is at any point in time. Building on this the extent to which select moral content is rep- work, a self-complexity approach is, we believe, resented centrally across aspects of one’s self- a more refined approach that considers both identity, and therefore highly salient. Impor- moral content and structure, proposing that in- tantly, applying self-complexity and related dividuals possess both integration (i.e., unity) 2011 Hannah, Avolio, and May 673 and differentiation of moral content across the defined as more or less complex, and (2) the self (see LeBoeuf et al., 2010; Reicher & Levine, makeup of that identity will influence moral 1994; Woolfolk et al., 2004). With this theoretical cognition and conation through the activation of framework, the influence of moral identity on various working selves across contexts. We now moral cognition and conation would be based suggest that (3) a more complex moral identity on (1) what subset of an individual’s identity will provide a broader base of moral content (e.g., a social role) is activated in a given situa- that individuals may draw from. Based on the tion and (2) what moral content that subset con- discussion of cognitive differentiation presented tains. Instead of asking, “How active is one’s earlier, more complex individuals can better tai- global moral identity?” as suggested by prior lor their working self across a broader range of research (e.g., Aquino et al., 2009), we ask, situations (Hannah et al., 2009b; Lord et al., 2011; “Which identity is activated and what is the Rafaeli-Mor & Steinberg, 2002). This aligns with unique moral makeup of that identity?” Yearley’s discussion of virtue ethics: “I do not This approach may better explain cross- act benevolently in order to be benevolent or to situational variability in moral thought and ac- be seen as benevolent by myself or others....I tion through assessing the activation of specific act benevolently because the situation I face fits moral content in the working self-concept a description of a situation that elicits my be- (Markus & Wurf, 1987). Because the self-concept nevolence” (1990: 14). In sum, individuals’ moral is a vast and manifold structure, only a subset of content will vary in levels of complexity—that identity is activated at any one point in time. is, differentiation and integration/unity across This portion—the working self-concept—in- different social roles. The structure and content cludes aspects of identity and associated cogni- of moral identity will therefore influence its ac- tive, affective, and motivational components tivation and extent of influence in the working that are activated in a specific situation and self, and thereby drive moral cognition and that drive thoughts and behaviors in that situa- conation. In summarizing this research on moral tion (Lord et al., 2011). For example, the identity identity, we propose the following. of “self as parent” may have highly salient and Proposition 3: Higher (lower) overall accessible content associated with role model- levels of moral identity complexity ing and discipline, which guides a parent to will be associated with higher (lower) closely regulate his or her own behavior (e.g., levels of (a) moral sensitivity and (b) control the use of profanity) and to quickly dis- elaborative moral judgments across cipline a child for even the slight use of profan- situations. ity. Conversely, the individual’s identity as “self as coworker” may include less emphasis on be- Proposition 4: Higher (lower) levels of ing a moral disciplinarian. When the coworker moral complexity in the activated role is active during the workday, the individual working self will be associated with may be much less sensitive to moral issues re- higher (lower) levels of (a) moral sen- lated to coworker behavior. This difference in sitivity and (b) elaborative moral judg- moral sensitivity may impact not only moral ments within that situation. judgments at work but also the motivation to act Proposition 5: Dimensions of moral once a judgment is reached. The net result is a identity with higher (lower) levels of coworker who might allow his or her own and unity will be related to higher (lower) other coworkers’ behavior to degenerate below levels of (a) moral sensitivity, (b) elab- a level that would not be tolerated as a parent. orative moral judgments, (c) moral Consequently, examining both the content and motivation, and (d) moral action con- structure of moral identity can provide a deeper cordant with those core dimensions. understanding of moral thought and behavior within and between the various roles that indi- viduals maintain as part of their personal and MORAL CONATION work identities. This discussion of moral identity and self- As noted by Rest et al. (1999) and others (e.g., complexity suggests that (1) the assemblage Thoma et al., 1991), the processes of moral moti- (content and structure) of moral identity can be vation and moral action, and more so the capac- 674 Academy of Management Review October ities that produce such conation, have not been process through an internally driven orientation adequately developed in the literature. This for moral action. Pierce et al. (2003) propose that lack of attention is surprising, given research individuals hold varying levels of psychological demonstrating relatively weak relationships be- ownership for various aspects of their environ- tween ethical judgments and ethical behaviors ment, or what they call “targets.” They define (Bebeau, 2002; Blasi, 1980; Treviño & Young- psychological ownership as “the state in which blood, 1990). This gap points to the importance of individuals feel as though the target of owner- the moral conation capacities depicted in Fig- ship or a piece of that target is theirs” (2003: 86). ure 1. Advancing a model of moral capacity for While not yet applied to ethics, we put forth in use in dynamic organizations where individuals our model a specific form of moral ownership we face competing values will require identifying define as the extent to which members feel a and developing the factors that underlie the sense of psychological responsibility over the transference of moral judgments into action. We ethical nature of their own actions, those of oth- build the construct of moral conation drawing ers around them, their organization, or another from literature on (1) human agency, psycholog- collective. This “other collective” could be a ical ownership, and engagement (e.g., Bandura, group, a club, or even a society. 1991, 1999; May, Gilson, & Harter, 2004; Pierce, Like the other moral conation constructs, we Kostova, & Dirks, 2003); (2) self- and means effi- suggest moral ownership is state-like, varying cacy (e.g., Bandura, 1997; Eden, 2001); and (3) across “targets” and contexts. Unlike general courage (e.g., Gould, 2005; Kidder, 2003). We pro- beliefs about the extent to which individuals pose that these three constructs are distinct yet can control the general events in their life, as support each other. Affirming our position, Oss- reflected in concepts like locus of control wald, Greitemeyer, Fischer, and Frey state that (Treviño, 1986), individuals will vary in the ex- “before a person can act with moral courage, tent they want to or feel a sense of responsibility s/he has to perceive an incident as a situation of to take ethical action in a given situation and moral courage, s/he has to take responsibility not others (Bandura, 1991, 1999). [i.e., moral ownership] and has to feel competent Bandura’s (1991, 1999) theory of moral agency [i.e., moral efficacy] to act” (2009: 98). is helpful in identifying the underlying psycho- logical mechanisms that create (or reduce) a sense of moral ownership across contexts. Ban- Moral Ownership dura defines agency as the capacity to exercise Kohlberg and Candee (1984) proposed that a control over the nature and quality of one’s life. sense of responsibility must first be formed be- As the central organizing principle of social cog- fore people will initiate dedicated moral action. nitive theory, agency reflects individuals’ en- Jones and Ryan’s (1997, 1998) moral approbation gagement in their experiences. Bandura argues model suggests that individuals attribute their that interactions occur between the person (e.g., level of responsibility for taking moral action us- cognition, identity, and affect), his/her behavior, ing their various referent groups as standards. and his/her environment. Through these interac- This attribution is based on assessments of the tions, people can be producers as well as prod- severity of the consequences, moral certainty, ucts of their environments. However, Bandura degree of complicity, and extent of pressure (1991, 1999) suggests that individuals may use from organizational factors. Individuals will “moral disengagement” techniques, such as dif- then act if their level of anticipated approbation fusing responsibility to or attributing blame to from their referent group exceeds their desired others, or discounting the extent of harm asso- level of approbation. This suggests that, unlike ciated with their behavior, in order to psycho- the influence of stable traits, such as locus of logically disengage from their actions and control on moral processing (Treviño, 1986), thereby maintain a positive self-image. Tenbrun- moral approbation will vary across situations as sel and Messick (2004) make a similar argument, one’s reference group and other factors change. noting that individuals may employ strategies To represent this sense of ownership, we pro- of self-deception during the omission of ethical pose a construct of moral ownership, which, we or commission of unethical acts so as to deceive suggest, will create unique causations or con- themselves into believing that their ethical prin- tingencies in this psychological approbation ciples are still being upheld. We suggest that 2011 Hannah, Avolio, and May 675 higher levels of moral ownership will combat dent but still remain inactive because of a lack these tendencies. of confidence, thus lowering moral conation. For Bandura (1991) suggests that people vary in example, an individual may feel that he or she how much they practice four aspects of agency: should confront someone who has behaved un- (1) intentionality—the extent to which acts of ethically, but then not feel capable of doing so agency are done intentionally, (2) forethought— effectively. Moral efficacy has been recognized the extent to which agents anticipate likely con- as an important factor in potentially addressing sequences of actions and select courses of ac- this gap in whether individuals will act ethi- tion that produce desired outcomes and avoid cally (Hannah et al., 2005; May, Chan, Hodges, & detrimental ones, (3) self-reactiveness—the level Avolio, 2003). We define moral efficacy as an of ability to self-motivate and self-regulate to individual’s belief in his or her capabilities to achieve goals, and (4) self-reflectiveness—the organize and mobilize the motivation, cognitive level of ability to reflect on the adequacy of resources, means, and courses of action needed one’s thoughts and actions. While these theoret- to attain moral performance, within a given ical processes have not been adequately stud- moral domain, while persisting in the face of ied empirically, they do provide a basis for un- moral adversity. This definition draws on com- derstanding the underlying psychological ponents of both means efficacy (Eden, 2001) and processes that may drive moral ownership as self-efficacy (Bandura, 1997), which are both we have defined it in this article. We suggest malleable and open to development. that moral ownership represents individuals’ Bandura and Eden both argue that self- (Ban- sense of responsibility for and impetus to stand dura) and means (Eden) efficacy are contextual- up and act to influence morality in their current ized and, thus, domain specific, which is what environment. A lack of moral ownership would distinguishes these constructs from the global likely predict various negative outcomes, such and decontextualized construct of general effi- as a failure to act when moral action is needed, cacy (Chen, Gully, & Eden, 2001). In this case a or more subtle forms of poor organizational cit- domain merely represents a defined area for izenship, such as social loafing. which individuals may have greater or lesser In sum, moral ownership will prompt conation levels of confidence, such as leader efficacy to act since those with higher levels of owner- (Hannah, Avolio, Luthans, & Harms, 2008). Moral ship are simply less able to turn a blind eye. efficacy specifies the domain of moral action, This is consistent with the findings of Janoff- Bulman, Sheikh, and Hepp (2009), who showed which may be quite distinct from a person’s ef- that some individuals demonstrate a proscrip- ficacy for other domains, such as leading others tive morality (avoid doing bad) while others or public speaking. have a prescriptive focus (seek to do good). In a Metaanalytical research has demonstrated similar line of research, Ryan and Riordan (2000) the influential role of self-efficacy in driving hu- showed that some people seek to avoid blame man intentions and behaviors (see Stajkovic & while others seek to earn praise through taking Luthans, 1998). Applicable to one’s capacity for moral action. Moral ownership may help ex- moral conation, social and empathetic efficacy plain these motivational profiles, since we promote prosocial behavior, as evidenced by would expect high levels of ownership to spon- higher levels of helpfulness, cooperation, and sor prescriptive motivation and action. This sharing (Bandura, 1991). Further, based on find- leads to our next proposition. ings that perceived behavioral control (a form of efficacy beliefs) helps to explain the intention- Proposition 6: Higher (lower) levels of behavior linkage, we propose that moral effi- moral ownership will be associated cacy will similarly create conation to act ethi- with higher (lower) levels of (a) moral cally in line with intentions (Ajzen, 1991). This is motivation and (b) moral action. consistent with Treviño’s (1986) suggestion that ego strength is a critical link between ethical judgments and action. While ego strength is a Moral Efficacy stable individual-difference construct relative to One can make a sound moral judgment and the more malleable construct of self-efficacy, it feel ownership to act to address a moral inci- provides a similar self-regulatory function in 676 Academy of Management Review October promoting conviction to one’s actions (Rest, (e.g., peer/leader support or whistle-blower re- 1986). porting systems and protection). Because orga- Based on the work of Bandura (1997), we sug- nizations offer varying levels of promotion and gest that moral efficacy is composed of both support for ethical behavior (Victor & Cullen, magnitude (the level of difficulty one expects to 1988), we propose that moral efficacy will be successfully perform in a given moral situation) maximized when both self and means compo- and strength (the extent of certainty one has in nents are high. The effects of moral efficacy on one’s ability to perform to that level of difficulty). moral conation lead to our next proposition. Further, we link moral efficacy to earlier discus- Proposition 7: Higher (lower) levels of sions of self-complexity and the activation of a moral efficacy will be associated with tailored moral working-self. This is because higher (lower) levels of (a) moral mo- self-efficacy is not simply an assessment of tivation and (b) moral action. one’s skills but, rather, of what one can do with those skills in a given situation (Bandura, 1997), thereby influencing moral action in that and Moral Courage only that situation. This suggests individuals The final capacity underlying moral conation hold varying levels of moral efficacy across dif- depicted in Figure 1 is moral courage. Hannah, ferent aspects of their self-identity (e.g., efficacy Avolio, and Walumbwa define moral courage in as a leader to discuss ethical issues with their the workplace as group as opposed to a peer), creating some vari- ability in choices and behavior. 1) a malleable character strength, that 2) provides Linking moral efficacy to self-concept unity, the requisite conation needed to commit to per- sonal moral principles, 3) under conditions where discussed earlier, we propose that through re- the actor is aware of the objective danger in- peated successful experiences regarding moral volved in supporting those principles, 4) that en- action, moral efficacy can be generalized across ables the willing endurance of that danger, 5) in a broader set of moral contexts over time. Thus, order to act ethically or resist pressure to act we believe that moral efficacy, built through a unethically as required to maintain those princi- ples (2011: 560). wide span of rich personal mastery and model- ing/vicarious experiences, would continuously Overcoming threat and perhaps fear for the generalize across a widening expanse of moral sake of morals, virtue, or a higher purpose is tasks and contexts (Bandura, 1997). inherent in definitions of courage (Gould, 2005; Eden made an important distinction between Kidder, 2003). Otherwise, action to overcome a one’s internal (self-efficacy) and external perceived threat can be considered self-serving (means efficacy) sources of efficacy beliefs. Add- or foolhardy. ing to our discussion above, means efficacy rep- Moral courage has been proposed as a critical resents individuals’ beliefs in the quality and factor in promoting ethical behavior in organi- utility of the individuals, tools, methods, and zation members (Verschoor, 2003), and in an ini- procedures available for task performance in a tial field test Hannah et al. (2011) linked moral given context. Research has shown that the ef- courage to externally rated ethical and proso- fects of means efficacy can be isolated from cial behaviors. Describing moral courage, those of self-efficacy (Eden & Sulimani, 2002). Sekerka, Bagozzi, and Charnigo state that These external resources can include imple- “strength of will is needed to face and resolve ments (e.g., equipment and computers), other ethical challenges and to confront barriers that persons (e.g., coworkers, followers, and supervi- may inhibit the ability to proceed toward right sors), or bureaucratic means for accomplishing action,” and, therefore, moral courage is “a qual- work (e.g., policies, procedures, and processes). ity or attribute necessary for ethical behavior in We suggest that both sources of efficacy are organizational settings” (2009: 566). Research necessary to fully explain moral conation. Spe- has differentiated moral courage from other cifically, individuals must believe they not only forms of courage, such as physical or social have the personal capability to address a spe- courage (Rate, Clarke, Lindsay, & Sternberg, cific moral issue (e.g., to disclose unethical ac- 2007; Woodard & Pury, 2007). Yet, as noted by counting practices) but that supporting means Hannah and Avolio (2010), empirical work on are available to allow them to act successfully moral courage is nascent and has not been ad- 2011 Hannah, Avolio, and May 677 equately integrated with other aspects of moral higher (lower) (a) moral motivation psychology, such as we have proposed in Fig- and (b) moral action. ure 1, nor integrated with the processes in Rest et al.’s four-component model. DISCUSSION Providing this integration is important, since we propose that the three components of moral The comprehensive framework shown in Fig- conation are mutually supporting. Individuals ure 1 has implications for theory, methods, and may feel responsibility to act (i.e., have moral practice. The model offers new approaches for ownership) and believe that they have the ca- defining and measuring moral capacities and pacity to do so (i.e., have moral efficacy), yet still can generate new lines of research to predict have insufficient courage to overcome the threat ethical thoughts and behaviors in organizations. they face and to act. Moral ownership, efficacy, and courage, thus, are each necessary yet not Implications for Theory sufficient. Supporting this position, Osswald et As stated earlier, Rest et al. (1999) reported al. (2009) have noted that feelings of responsibil- that moral judgment as measured by the DIT ity and competence are required to support typically only explains about 20 percent of the moral courage. Similarly, in their model of gen- variance in actual ethical behavior. These au- eral courage (not necessarily moral courage), thors note the need to develop models that bet- Hannah, Sweeney, and Lester (2010) propose ter explain the entire ethical process and that that both self- and means efficacy are required although “there is much diversity of constructs, to reinforce one to act with courage. This is in processes, phenomena, and starting points for part because courage is often relative to expe- the psychology of morality, the greater chal- rienced fear, and negative emotions such as fear lenge is to formulate how all these different occur when individuals assess that a given parts fit together” (1999: 6). We have made one threat exceeds their perceived ability to face attempt to circumscribe such an integrated that threat (Beck, Emery, & Greenberg, 1985). framework, identifying developed capacities re- Moral efficacy would thus bolster perceived ca- lated to each of Rest’s (1986) four stages. Impor- pability, and hence moral courage, all working tant to promoting future research, we demon- together to produce higher levels of moral strate how each component in our model can be conation. operationalized and measured. Finally, consistent with the concept of differ- Benefits of a complexity approach. A primary entiation and integration/unity presented ear- contribution of our model is the incorporation of lier, individuals will possess greater or lesser theories of cognitive complexity and self- amounts of courage across identity subdimen- complexity throughout the model. While Kohl- sions. For example, a person may have more berg’s interview process and Rest’s ethical di- moral courage in his or her role as a leader than lemma exercises are meant to denote general as a follower. He or she will then demonstrate levels of development in logic or moral sche- greater moral courage when the leader role is mas, these techniques are abstract, focus only active (e.g., confronting a follower) than when on moral judgment, and do not recognize the the follower role is active (e.g., confronting his or important role of identity and accompanying her own senior leader). Thus, moral courage, self-regulatory functions in explaining an indi- like moral efficacy, is contextualized and do- vidual’s moral capacity to take action. main specific. Yet Shepela et al. (1999) have sug- Below we show how complexity can be mea- gested that moral courage related to certain val- sured more directly than by Kohlberg’s or Rest’s ues can be core to a person (what we have techniques through schema mapping, Q-sort defined as high unity) and therefore especially tasks, self-complexity matrices, and other tech- powerful in prompting moral courage across a niques. Employing such techniques will refine broader set of domains. This leads to our final our measurements and understanding of what constitutes moral complexity. Further, a self- proposition. complexity approach will allow researchers to Proposition 8: Higher (lower) levels of extend beyond current “global” measures of moral courage will be associated with moral identity, moral efficacy, and moral cour- 678 Academy of Management Review October age and, instead, to compartmentalize mea- deeper understanding of individuals’ moral sures to explore how higher or lower levels of maturation, we must measure moral identity these constructs are represented across differ- and complexity along with associated meta- ent aspects of self-identity. Such an approach cognitive abilities whenever possible within should increase the predictive validity of mea- domains, and then test their effects when in- sures, as well as explore within-person variance dividuals are confronted with moral dilemmas across situations as individuals’ working selves within and between those domains. Rest et are activated or suppressed (Markus & Wurf, al.’s scenarios, such as the “Heinz” dilemma 1987). Further, these techniques will allow re- (regarding whether the target individual searchers to mathematically express the level of should steal a drug his wife needs from a inclusion of various moral attributes across self- pharmacist), address the broad, abstract rea- identities and, therefore, the level of unity each soning of cases most would never face, versus participant has related to attributes of interest the more common concrete and complex situ- to researchers (e.g., a certain set of moral val- ations individuals face in organizational life. ues). These measures of unity can thus be used Further, the DIT purports to measure cognitive to predict self-concordant ethical behaviors. moral development (CMD) levels by having par- Contribution of moral conation. We believe ticipants read scenarios and then rate and rank our framing of the moral conation constructs offers various standardized decision criteria based on new insights into the transference of moral judg- the importance of those criteria in making a ments, through intentions, into action. Discus- decision. Each of those criteria is designated as sions of moral courage date to antiquity, and it personal interest, maintaining order,orpostcon- is the most theoretically developed of the three ventional reasoning levels. Thus, the more im- moral conation capacities. Yet the study of portance an individual ascribes to items associ- moral courage in organizational contexts has ated with postconventional reasoning, the been limited (Hannah & Avolio, 2010; Sekerka et higher that individual will score on postconven- al., 2009). Further, we are not aware of any prior tional CMD. One weakness of this system is that work on the construct of moral ownership in the the DIT ostensibly is a recognition task, where literature and have noted only limited theoreti- participants can read and select items represent- cal development regarding moral efficacy (e.g., ing decision criteria that they may never have Hannah et al., 2005; May et al., 2003). even thought of on their own if not prompted by Prior theorizing of moral efficacy has also failed the responses given in the measure. to incorporate how contextual factors influence We believe levels of moral cognitive complex- an individual’s level of moral efficacy to act. To ity can be determined through more direct meth- address this gap, we have included means effi- ods by using schema mapping (also known as cacy in our construct definition. Finally, we have cognitive or causal mapping) techniques to as- integrated the moral conation constructs and sess individuals’ concepts of morality (i.e., argued that moral ownership, efficacy, and knowledge content), as well as the structure courage each provide necessary yet not suffi- with which they store that content. Schema map- cient contributions to moral conation and, thus, ping has been used successfully to assess other should be studied and developed together. areas of individuals’ knowledge structures in organizational research (e.g., Eden, Ackermann, & Cropper, 1992; Markoczy & Goldberg, 1995). Operationalizing Moral Maturation This technique has also been used to map ethical One criticism of Kohlberg’s stages of devel- schemas of journalists (Lind, Rarick, & Swenson- opment has been the requirement for inter- Lepper, 1997). views and the subjectivity associated with in- Based on the work of Schroder et al. (1967), terpretations and coding of participant complexity integration and differentiation can responses. Furthermore, both Kohlberg’s inter- also be measured using thematic apperception view and Rest et al.’s (1999) written scenario tests, such as picture story exercises (Baker- response technique in the DIT utilize stan- Brown et al., 1992; Tetlock, Peterson, & Berry, dardized ethical dilemmas that neglect the 1993), where open-ended responses are exter- contextualized nature of moral knowledge we nally rated for levels of complexity by trained presented earlier. If we are to develop a coders. These techniques, however, would need 2011 Hannah, Avolio, and May 679 to be modified to focus on moral content. Fur- conation lend themselves more readily to survey ther, there are recent advances in neuroscience methodology and self- and other reports. Han- that may enable researchers to assess complex- nah and Avolio (2010) recently developed a new ity through brain mapping techniques. For ex- multidimensional measure representing the ample, a measure called “coherence” deter- three components of moral conation, based on mines how differentiated the human brain is prior work on psychological ownership (Van either at rest or while processing tasks and may Dyne & Pierce, 2004), self-efficacy (Bandura, be one metric that can be used to assess how 1997), and moral courage (Gould, 2005; Kidder, individuals process moral tasks (see Thatcher, 2003). They tested this three-factor measure and North, & Biver, 2008). found across two studies that moral ownership, Adopting a self-complexity approach to exam- moral efficacy, and moral courage are discrim- ining moral identity will also require a funda- inant constructs yet share variance in that they mentally different approach to measurement form a higher-order construct. Their initial than the current single-survey techniques com- model testing supports our theorizing that the monly used. Moral self-complexity can be as- components of moral conation are each neces- sessed by having participants produce the var- sary yet not sufficient. Sekerka et al. (2009) also ious key roles of their identity (e.g., “self as team has conducted an initial validation of an orga- leader” or “self as coworker”) in free-response nizational moral courage measure. formats, or by using roles as designated by the Based on the tenets of self-complexity theory researcher, and then by sorting ethical attri- presented earlier, moral ownership, courage, butes of interest to the researcher (e.g., values or and efficacy will be represented in varied forms traits) as to whether, and the extent to which, and levels across one’s self-identity. Therefore, each attribute applies to each role using the we suggest that researchers assess these con- Q-sort technique of Linville (1987) or the survey structs of moral conation across different as- format used by Woolfolk et al. (2004). Regardless pects of participants’ self-identity. This can be of the technique used, data can then be format- done by presenting respondents with separate ted into a matrix (self-aspects or roles ϫ attri- scales (e.g., moral courage) for each of their var- butes). Once compiled, the h-statistic (Scott, ious relevant social roles, preferably separating 1969) can then be used to analyze the complexity each survey administration across time to re- of the matrix. duce carryover effects. For example, survey item The h-statistic represents an index of the num- stems for a moral efficacy measure might ask, ber of independent dimensions underlying any “When leading my top management team I set of attribute ratings. Such matrices can also can...”and, on a second version, “When work- illuminate what moral attributes are repre- ing with my peers I can....” Responses from sented across a breadth of social roles, denoting various roles can then be collapsed into a ma- what we discussed above as constituting self- trix to assess dispersion of moral efficacy across concept unity. Finally, measurement methods roles, allowing researchers to investigate the for metacognitive ability also exist in the edu- effects of moral efficacy across contexts based cation literature that could be modified to as- on individuals’ efficacy magnitude and strength sess participants’ knowledge and regulation of associated with each context. The differentia- moral cognition (Baker & Cerro, 2000; Dunlosky tion of moral efficacy across the self-identity can & Metcalfe, 2009). also be assessed using the h-statistic based on In sum, having the ability to measure these this matrix, as described previously. Similar components allows researchers not only to test methods can be used for assessing each moral our proposed theory but also to measure moral conation construct. maturation over time. In combination, we hope to advance both the science and practice asso- Implications for Practice ciated with moral capacity development. We believe that we have offered a set of con- structs associated with moral maturation and Operationalizing Moral Conation moral conation that addresses the capacities Unlike the more complicated constructs of needed across all stages of ethical processing, moral maturation, the constructs of moral from the stage of sensitivity through to action. 680 Academy of Management Review October

These capacities are both malleable and mea- “Which of your core beliefs are at stake?” “How surable and can thus be used for selection pur- central are those values to you?” and/or “How poses and to implement and assess moral de- are your emotions influencing your thoughts velopment interventions over time in those and behaviors right now?” Further, teaching organizations seeking to enhance the moral ca- skills to methodically review moral issues pacity of their members.4 We briefly introduce through multiple perspectives—for example, de- some methods for developing each capacity in ontological (rules, duties, or norms), teleological Figure 1 simply to highlight the model’s utility (utilitarian, consequence, or goal-based), or val- and to encourage future research on moral ca- ues-based reasoning—would force individuals pacity development. to access and “exercise” various moral sche- Developing moral maturation capacity. Moral mata, enhancing their level of moral complexity complexity can foremost be developed through and metacognitive moral processing ability. social learning, including rich personal mastery Finally, moral identity can be particularly de- or vicarious experiences in specific domains. veloped through exposure to moral role models Other authors have similarly noted the power of that provide an ideal to strive for, and it serves social learning in ethical decision making based to motivate and guide others’ development (Lord on moral approbation, where individuals learn & Brown, 2004; Mayer et al., 2009). Research has the response consequences of ethical actions also shown that dialogue with others at higher (Jones & Ryan, 1998). The development of moral levels of CMD promotes moral development by maturation occurs as exposure to moral experi- offering the individual new perspectives with ences and conflicts triggers the development of which to think about ethical issues (Dukerich, new associations between concepts held in an Nichols, Elm, & Volrath, 1990). individual’s mental representations, thereby en- Developing moral conation capacity. Based on hancing the individual’s level of complexity the work of Bandura (1991, 1999), we believe that (Street et al., 2001; Walker, 1983; Young & Wasser- organizations can heighten moral ownership by man, 2005). In a training setting moral complexity (1) placing issues in humanistic terms, (2) dis- can be developed through moral discourse, in- couraging euphemisms and sanitizing lan- cluding (1) teaching cognitive moral reasoning guage (e.g., “collateral damage”), (3) encourag- skills (e.g., logic, role-taking, or justice concepts), ing responsibility, (4) making salient the (2) instructing and facilitating group reasoning injurious effects of potential actions, and (5) lim- through exercises and case analysis, and (3) dis- iting attribution of blame to or dehumanization cussing ethics in practical applied areas an indi- of victims. This is consistent with previous eth- vidual may face (e.g., dentistry, journalism, medi- ics research specifying the important role of cine; Bebeau, 2002; Hartwell, 1995). For example, a highlighting the consequences of one’s behav- review of twenty-three ethics programs reported iors to increase moral engagement (e.g., Ferrell by Rest and Thoma (1986) showed that those using & Gresham, 1985; Hunt & Vitell, 1986). This is group moral dilemma discussions had an average also consistent with Jones’s (1991) theory of .41 effect size in raising levels of cognitive moral moral intensity, which has shown that people development versus .09 for those without moral become more morally engaged by raising the discussion. perceived “intensity” associated with an ethical We believe that moral metacognitive ability dilemma (cf. Butterfield, Treviño, & Weaver, can be developed through teaching techniques 2000; Watley & May, 2004). Similarly, Jones and to process moral dilemmas through deeper self- Ryan (1998) have proposed that organizations reflection (Petty & Cacioppo, 1986; Setterlund & can reinforce ethical behavior by raising the Niedenthal, 1993) and practicing executive con- perceived severity of consequences, moral cer- trol over moral processing (Metcalfe & Shima- tainty, degree of complicity, and pressure to mura, 1994). Example executive control “drills” comply. could include guiding questions, such as Organizational reward and control systems may also enhance moral ownership (Treviño, Brown, & Pincus-Hartman, 2003). These systems 4 We recognize that some organizations may be ambiva- lent about or even condone their members’ unethical acts or can signal what is valued in organizations, and that they do not otherwise want to invest the time and research has shown that although individuals resources in developing members’ moral capacity. may initially comply with norms for strategic 2011 Hannah, Avolio, and May 681 self-presentation, over time, such norms can to the effects of a leader’s efficacy in bolstering cause identity changes that can impact the in- the collective leadership efficacy of his or her dividual’s sense of responsibility to take moral group (Hannah et al., 2008). Since knowledge can action (e.g., Tice, 1992). If not properly aligned be distributed across members of a group and with moral action, however, such reward sys- combined through social interaction as needed tems may create a negative pressure to comply to enhance group processing (Kozlowski & Ilgen, with unethical actions (Jones & Ryan, 1998). 2006), even moral complexity might be concep- Through social learning where successful tualized and measured as a multilevel con- moral performance is achieved, individuals struct. will not only build greater moral complexity but Repeatedly throughout this article we have also the confidence to enact similar approaches noted that aspects of moral maturation and to address future ethical challenges (Bandura, moral conation should be tailored to specific 1997). Ethical role models can also reinforce ob- contexts. This is because different contexts, servers’ efficacy, as well as the collective effi- such as national or organizational cultures, cacy of the group, to act morally over time (Ban- have different social mores (Margolis & Phillips, dura, 2002). This may be one mechanism 1999; Shweder, 1991), and culture influences eth- explaining how ethical leadership can diffuse to ical processing (see Kish-Gephart, Harrison, & others throughout an organization (Mayer et al., Treviño, 2010; Tenbrunsel & Smith-Crowe, 2008). 2009). Hannah et al. (2011) demonstrate that lead- We believe our model can be tailored to specific ers who are seen as being authentic can bolster cultures (e.g., by measuring complexity based followers’ moral courage and subsequent ethi- on particular cultural mores). Yet our proposed cal and prosocial behaviors. They theorize that model is not context neutral. Researchers should authentic leaders serve as moral exemplars and assess how cultural contingencies and con- also establish transparent climates that encour- straints influence the predictions associated age followers to openly espouse their values with our model (see Treviño et al., 2006). and to act in line with those values coura- geously. Further, training programs have re- cently shown some success in developing moral CONCLUSION courage through teaching behavior routines We propose that rates of unethical behavior (i.e., scripts) individuals can use when facing can be decreased and virtuous behavior in or- threats (e.g., Jonas, Boos, & Brandstätter, 2007; ganizations increased through the development Osswald et al., 2009). Finally, Walker and Hen- of moral capacity—what we have conceptual- ning (2004) suggest that moral exemplars can ized as moral maturation and moral conation. In have a contagion effect on others such that ob- proposing the current model, we hope we have servers come to believe they, too, have the cour- provided a clearer line of sight to the capacities age to successfully meet similar threats. underlying moral development and have pro- vided propositions and measurement ap- Boundaries and Future Research proaches to promote future research. For the sake of parsimony, we have focused on the individual level in developing our model in REFERENCES order to detail the component processes that underlie individual moral maturation and moral Ajzen, I. 1991. The theory of planned behavior. Organizational conation. We suggest, however, that the model Behavior and Human Decision Processes, 50: 179–211. may be applied as a multilevel framework to Aquino, K., Freeman, D., Reed, A., Lim, V. K. G., Felps, W. explain how groups and organizations develop 2009. Testing a social-cognitive model of moral behav- ior: The interactive influence of situations and moral a higher “moral perspective.” For example, self- identity centrality. Journal of Personality and Social Psy- efficacy and collective efficacy are reciprocal chology, 97: 123–141. and entrained (Bandura, 1997). The constructs Aquino, K., & Reed, A. 2002. The self-importance of moral associated with this model lend themselves to identity. Journal of Personality and Social Psychology, investigation of how individual development 83: 1423–1440. can extend to collective development through Baker, L., & Cerro, L. 2000. Assessing metacognition in chil- social learning and influence processes, similar dren and adults. In G. Schraw & J. Impara (Eds.), Issues 682 Academy of Management Review October

in the measurement of metacognition: 99–145. Lincoln: Dukerich, J. M., Nichols, M. L., Elm, D. R., & Volrath, D. A. 1990. Buros Institute, University of Nebraska. Moral reasoning in groups: Leaders make a difference. Human Relations, 43: 473–493. Baker-Brown, G., Ballard, E. J., Bluck, S., deVries, B., Suedfeld, P., & Tetlock, P. 1992. The integrative complexity coding Dunlosky, J., & Metcalfe, J. 2009. Metacognition. London: manual. In C. Smith (Ed.), Handbook of thematic analy- Sage. sis: 605–611. Cambridge: Cambridge University Press. Dutton, J. E., & Jackson, S. E. 1987. Categorizing strategic Bandura, A. 1991. Social cognitive theory of moral thought issues: Links to organizational action. Academy of Man- and action. In W. M. Kurtines & J. L. Gewitz (Eds.), Hand- agement Review, 12: 76–90. book of moral behavior and development, vol. 1: 45–103. Eden, C., Ackermann, R., & Cropper, S. 1992. The analysis of Hillsdale, NJ: Lawrence Erlbaum Associates. cause maps. Journal of Management Studies, 29: 309– Bandura, A. 1997. Self-efficacy: The exercise of control. New 324. York: Freeman. Eden, D. 2001. Means efficacy: External sources of general Bandura, A. 1999. Moral disengagement in the perpetuation and specific subjective efficacy. In M. Erez, U. Kleinbeck, of inhumanities. Personality and Social Psychology Re- & H. Thierry (Eds.), Work motivation in the context of a view, 33: 193–209. globalizing economy: 65–77. Hillsdale, NJ: Lawrence Erl- baum Associates. Bandura, A. 2002. Social cognitive theory in cultural context. Journal of Applied Psychology, 51: 269–290. Eden, D., & Sulimani, R. 2002. Pygmalion training made ef- fective: Greater mastery through augmentation of self- Bebeau, M. J. 2002. The Defining Issues Test and the four efficacy and means efficacy. In B. J. Avolio & F. J. Yam- component model: Contributions to professional educa- marino (Eds.), Transformational/charismatic leadership: tion. Journal of Moral Education, 31: 271–294. The road ahead: 287–308. New York: Elsevier. Beck, A. T., Emery, G., & Greenberg, R. L. 1985. Anxiety Ferrell, O. C., & Gresham, L. G. 1985. A contingency frame- disorders and phobias: A cognitive perspective. New work for understanding ethical decision-making in mar- York: Basic Books. keting. Journal of Marketing, 49: 87–96. Blasi, A. 1980. Bridging moral cognition and moral action: A George, B. 2007. True north: Discover your authentic leader- critical review of the literature. Psychological Bulletin, ship. San Francisco: Jossey-Bass. 88: 1–45. Gilligan, C. 1982. In a different voice. Cambridge, MA: Har- Blasi, A. 1993. The development of identity: Some implica- vard University Press. tions for moral functioning. In G. Noam & T. Wren (Eds.), The moral self: 99–122. Cambridge, MA: MIT Press. Gollwitzer, P. M., & Schaal, B. 1998. Metacognition in action: The importance of implementation intentions. Personal- Bower, G. H., & Hilgard, E. R. 1981. Theories of learning. ity and Psychology Review, 2: 124–136. Saddle River, NJ: Prentice Hall. Gould, N. H. 2005. Courage: Its nature and development. Brown, M. E., Treviño, L. K., & Harrison, D. A. 2005. Ethical Journal of Humanistic Counseling, Education and Devel- leadership: A social learning perspective for construct opment, 44(1): 102–116. development and testing. Organizational Behavior and Human Decision Processes, 97: 117–134. Hannah, S. T., & Avolio, B. J. 2010. Moral potency: Building the capacity for character-based leadership. Consulting Butterfield, K. D., Treviño, L. K., & Weaver, G. R. 2000. Moral Psychology Journal, 62: 291–310. awareness in business organizations: Influences of is- sue-related and social context factors. Human Relations, Hannah, S. T., Avolio, B. J., Luthans, F., & Harms, P. D. 2008. 53: 981–1018. Leadership efficacy: Review and future directions. Lead- ership Quarterly, 19: 669–692. Campbell, J. D., Trapnell, P. D., Heine, S. J., Katz, I. M., La- vallee, L. F., & Lehman, D. R. 1996. Self-concept clarity: Hannah, S. T., Avolio, B. J., & Walumbwa, F. O. 2011. The Measurement, personality correlates, and cultural relationships between authentic leadership, moral cour- boundaries. Journal of Personality and Social Psychol- age, and ethical and pro-social behaviors. Business Eth- ogy, 70: 141–156. ics Quarterly, 21: 555–578. Carver, C. S., & Scheier, M. F. 1998. On the self-regulation of Hannah, S. T., Lester, P. B., & Vogelgesang, G. R. 2005. Moral behavior. New York: Cambridge University Press. leadership: Explicating the moral component of authen- tic leadership. In W. B. Gardner, B. J. Avolio, & F. O. Chen, G., Gully, S. M., & Eden, D. 2001. Validation of a new Walumbwa (Eds.), Authentic leadership theory and general self-efficacy scale. Organizational Research practice: Origins, effects, and development: 3–42. Ox- Methods, 4(1): 62–83. ford: Elsevier. Dollinger, M. J. 1984. Environmental boundary spanning and Hannah, S. T., Sweeney, P. J., & Lester, P. B. 2010. The coura- information processing effects on organizational perfor- geous mindset: A dynamic personality system approach mance. Academy of Management Journal, 27: 351–368. to courage. In C. Pury & S. Lopez (Eds.), The psychology Donaldson, T. 2003. Editor’s comments: Taking ethics seri- of courage: Modern research on an ancient virtue: 125– ously—A mission now more possible. Academy of Man- 148. Washington, DC: American Psychological Associa- agement Review, 28: 363–366. tion. 2011 Hannah, Avolio, and May 683

Hannah, S. T., Uhl-Bien, M., Avolio, B. J., & Cavarretta, F. Kozlowski, S. W. J., & Ilgen, D. R. 2006. Enhancing the effec- 2009a. A framework for examining leadership in ex- tiveness of work groups and teams. Psychological Sci- treme contexts. Leadership Quarterly, 20: 897–919. ence in the Public Interest, 7(3): 77–124. Hannah, S. T., Woolfolk, R. L., & Lord, R. L. 2009b. Leader LeBoeuf, R. A., Shafir, E., & Bayuk, J. B. 2010. The conflicting self-structure: A framework for positive leadership. Jour- choices of alternating selves. Organizational Behavior nal of Organizational Behavior, 30: 269–290. and Human Decision Processes, 111: 48–61. Hardy, S. A., & Carlo, G. 2005. Identity as a source of moral Lind, R. A., Rarick, D. L., & Swenson-Lepper, T. 1997. Cogni- motivation. Human Development, 48: 232–256. tive maps assess news viewer ethical sensitivity. Jour- nal of Mass Media Ethics, 12: 133–147. Hartwell, S. 1995. Promoting moral development through ex- periential teaching. Clinical Law Review, 1: 505–539. Linville, P. W. 1987. Self-complexity as a cognitive buffer against stress-related illness and depression. Journal of Hill, P. L., & Roberts, B. W. 2010. Propositions for the study of Personality and Social Psychology, 52: 663–676. moral personality development. Current Directions in Psychological Science, 19: 380–383. Lord, R. G., & Brown, D. J. 2004. Leadership processes and follower self-identity. Mahwah, NJ: Lawrence Erlbaum Hunt, S., & Vitell, S. 1986. A general theory of marketing Associates. ethics. Journal of Macromarketing, 6(1): 5–16. Lord, R. G., Hannah, S. T., & Jennings, P. L. 2011. A framework Janoff-Bulman, R., Sheikh, S., & Hepp, S. 2009. Proscriptive for understanding leadership and individual requisite versus prescriptive morality: Two faces of moral regu- complexity. Organizational Psychology Review, 1: 104– lation. Journal of Personality and Social Psychology, 96: 127. 521–537. Margolis, J. D., & Phillips, R. A. 1999. Toward an ethics of Jausovec, N. 2008. Metacognition: A psychophysiological per- organizations. Business Ethics Quarterly, 9: 619–638. spective. In M. F. Shaughnessy, M. V. J. Veenman, & C. Kleyn-Kennedy (Eds.), Meta-cognition: A recent review Markoczy, L., & Goldberg, J. 1995. A method for eliciting and of research, theory and perspectives: 45–62. New York: comparing causal maps. Journal of Management, 21: Nova Science. 305–333. Jonas, K., Boos, M., & Brandstätter, B. (Eds.) 2007. Training Markus, H., & Wurf, E. 1987. The dynamic self-concept: A moral courage: Theory and practice. Göttingen: Hogrefe. psychological perspective. Annual Review of Psychol- ogy, 38: 299–337. Jones, T. M. 1991. Ethical decision-making by individuals in organizations: An issue-contingent model. Academy of May, D. R., Chan, A. Y. L., Hodges, T. D., & Avolio, B. J. 2003. Management Review, 16: 366–395. Developing the moral component of authentic leader- ship. Organizational Dynamics, 32: 247–260. Jones, T. M., & Ryan, L. V. 1997. The link between ethical judgment and action in organizations: A moral appro- May, D. R., Gilson, R. L., & Harter, L. 2004. The psychological bation approach. Organization Science, 8: 663–680. conditions of meaningfulness, safety, and availability and the engagement of the human spirit at work. Journal Jones, T. M., & Ryan, L. V. 1998. The effect of organizational of Occupational and Organizational Psychology, 77(1): forces on individual morality: Judgment, moral appro- 11–37. bation, and behavior. Business Ethics Quarterly, 8: 431– 445. Mayer, D. M., Kuenzi, M., Greenbaum, R., Bardes, M., & Sal- vador, R. 2009. How low does ethical leadership flow? Kidder, R. M. 2003. Moral courage. New York: William Mor- Test of a trickle-down model. Organizational Behavior row. and Human Decision Processes, 108: 1–13. Kihlstrom, J. F., Beer, J. S., & Klein, S. B. 2003. Self and identity Metcalfe, J., & Shimamura, A. P. 1994. Metacognition: Know- as memory. In M. R. Leary & J. P. Tangney (Eds.), Hand- ing about knowing. Cambridge, MA: MIT Press. book of self and identity: 68–90. New York: Guilford Monin, B., Pizarro, D. A., & Beer, J. S. 2007. Deciding versus Press. reacting: Conceptions of moral judgment and the rea- Kish-Gephart, J. J., Harrison, D. A., & Treviño, L. K. 2010. Bad son-affect debate. Review of General Psychology, 11: apples, bad cases, and bad barrels: Meta-analytic evi- 99–111. dence about sources of unethical decisions at work. Narvaez, D. 2010. Moral complexity: The fatal attraction of Journal of Applied Psychology, 95: 1–31. truthiness and the importance of mature moral function- Kohlberg, L. 1981. The philosophy of moral development. San ing. Perspectives on Psychological Science, 5: 163–181. Francisco: Harper & Row. O’Fallon, M. J., & Butterfield, K. D. 2005. A review of the Kohlberg, L. 1986. A current statement on some theoretical empirical ethical decision-making literature: 1996–2003. issues. In S. Modgil & C. Modgil (Eds.), Lawrence Kohl- Journal of Business Ethics, 59: 375–413. berg: Consensus and controversy: 485–546. Philadelphia: Osswald, S., Greitemeyer, T., Fischer, P., & Frey, D. 2009. Falmer. What is moral courage? Definition, and classification of Kohlberg, L., & Candee, D. 1984. The relationship of moral a complex construct. In C. Pury & S. Lopez (Eds.), The judgment to moral action. In W. Kurtines & J. Gerwitz psychology of courage: Modern research on an ancient (Eds.), Morality, moral behavior and moral development: virtue: 94–120. Washington, DC: American Psychological 52–73. New York: Wiley. Association. 684 Academy of Management Review October

Petty, R. E., & Cacioppo, J. T. 1986. The elaboration likelihood Shepela, S. T., Cook, J., Horlitz, E., Leal, R., Luciano, S., & model of persuasion. Advances in Experimental Social Lutfy, E. 1999. Courageous resistance: A special case of Psychology, 123–205. altruism. Theory and Psychology, 9: 787–805. Piaget, J. 1965. (First published in 1932.) The moral develop- Shweder, R. A. 1991. Thinking through cultures. Cambridge, ment of a child. New York: Free Press. MA: Harvard University Press. Pierce, J. L., Kostova, T., & Dirks, K. T. 2003. The state of Sonenshein, S. 2007. The role of construction, intuition, and psychological ownership: Integrating and extending a justification in responding to ethical issues at work: The century of research. Review of General Psychology, 7: sensemaking-intuition model. Academy of Management 84–107. Review, 32: 1022–1040. Rafaeli-Mor, E., & Steinberg, J. 2002. Self-complexity and Stahlberg, D., Peterson, L., & Dauenheimer, D. 1999. Prefer- well-being: A review and research synthesis. Personal- ences for and evaluation of self-relevant information ity and Social Psychology Review, 6: 31–58. depending on the elaboration of the schema involved. European Journal of Social Psychology, 29: 489–502. Rate, C. R., Clarke, J. A., Lindsay, D. R., & Sternberg, R. J. 2007. Implicit theories of courage. Journal of Positive Psychol- Stajkovic, A. D., & Luthans, F. 1998. Self-efficacy and work- ogy, 2: 80–98. related performance: A meta-analysis. Psychological Bulletin, 124: 240–261. Reicher, S., & Levine, M. 1994. Deindividuation, power rela- tions between groups and the expression of social iden- Street, M. D., Douglas, S. C., Geiger, S. W., & Martinko, M. J. tity. British Journal of Social Psychology, 33: 145–163. 2001. The impact of cognitive expenditure on the ethical decision-making process: The cognitive elaboration Rest, J. R. (Ed.) 1986. Moral development: Advances in re- model. Organizational Behavior and Human Decision search and theory. New York: Praeger. Processes, 86: 256–277. Rest, J. R. 1994. Background, theory and research. In J. R. Rest Streufert, S., & Nogami, G. 1989. Cognitive style and complex- & D. Narvaez (Eds.), Moral development: Advances in ity: Implications for I/O psychology. In C. L. Cooper & I. research and theory (2nd ed.): 59–88. New York: Praeger. Robertson (Eds.), International review of industrial and Rest, J. R., Narvaez, D., Bebeau, M. J., & Thoma, S. J. 1999. organizational psychology: 147–181. Chichester, UK: Wiley. Postconventional moral thinking: A neo-Kohlbergian ap- Swann, W. B. 1983. Self-verification: Bringing social reality proach. Mahwah, NJ: Lawrence Erlbaum Associates. into harmony with the self. In J. Suls & A. G. Greenwald Rest, J. R., & Thoma, S. J. 1986. Education programs and (Eds.), Psychological perspectives on the self: 33–66. interventions. In J. R. Rest (Ed.), Moral development: Ad- Hillsdale, NJ: Lawrence Erlbaum Associates. vances in research and theory: 59–88. New York: Prae- Swanson, H. L., & Hill, G. 1993. Metacognitive aspects of ger. moral reasoning and behavior. Adolescence, 28: 711–735. Reynolds, S. J. 2006. A neurocognitive model of the ethical Tenbrunsel, A. E., & Messick, D. M. 2004. Ethical fading: The decision-making process: Implications for study and role of self-deception in unethical behavior. Social Jus- practice. Journal of Applied Psychology, 91: 737–748. tice Research, 17: 223–236. Reynolds, S. J. 2008. Moral attentiveness: Who pays attention Tenbrunsel, A. E., & Smith-Crowe, K. 2008. Ethical decision to the moral aspects of life? Journal of Applied Psychol- making: Where we’ve been and where we’re going. ogy, 93: 1027–1041. Academy of Management Annals, 2: 545–607. Reynolds, S. J., & Ceranic, T. L. 2007. The effects of moral Tetlock, P. E., Peterson, R. S., & Berry, J. M. 1993. Flattering judgment and moral identity on moral behavior: An em- and unflattering personality portraits of integratively pirical examination of the moral individual. Journal of simple and complex managers. Journal of Personality Applied Psychology, 92: 1610–1624. and Social Psychology, 64: 500–511. Ryan, L. V., & Riordan, C. M. 2000. The development of a Thatcher, R. W., North, D., & Biver, C. 2008. Development of measure of desired moral approbation. Educational and cortical connectivity as measured by EEG coherence Psychological Measurement, 60: 448–462. and phase. Human Brain Mapping, 12: 1400–1415. Schroder, H. M., Driver, M. J., & Streufert, S. 1967. Human Thoma, S. J., Rest, J. R., & Davidson, M. L. 1991. Testing a information processing: Individuals and groups func- moderator of the moral judgment and action relation- tioning in complex social situations. New York: Rinehart ship. Journal of Personality and Social Psychology, 61: & Winston. 659–669. Scott, W. A. 1969. Structure of natural cognitions. Journal of Tice, D. M. 1992. Self-concept change and self-presentation: Personality and Social Psychology, 12: 261–278. The looking glass self is also a magnifying glass. Jour- Sekerka, L. E., Bagozzi, R. P., & Charnigo, R. 2009. Conceptu- nal of Personality and Social Psychology, 63: 435–451. alizing and measuring professional moral courage. Treviño, L. K. 1986. Ethical decision-making in organizations: Journal of Business Ethics, 89: 565–579. A person-situation interactionist model. Academy of Setterlund, M. B., & Niedenthal, P. M. 1993. “Who am I? Why Management Review, 11: 601–617. am I here?”: Self-esteem, clarity, and prototype match- Treviño, L. K. 1992. Moral reasoning and business ethics: ing. Journal of Personality and Social Psychology, 65: Implications for research, education and management. 769–780. Journal of Business Ethics, 11: 445–459. 2011 Hannah, Avolio, and May 685

Treviño, L. K., Brown, M., & Pincus-Hartman, L. 2003. A qual- Walker, L. J. 1983. Sources of cognitive conflict for stage itative investigation of perceived executive ethical lead- transition in moral development. Developmental Psy- ership: Perceptions from inside and outside the execu- chology, 19: 103–110. tive suite. Human Relations, 56: 5–37. Walker, L. J., & Henning, K. H. 2004. Differing conceptions of Treviño, L. K., Weaver, G. R., & Reynolds, S. J. 2006. Behav- moral exemplarity: Just, brave, and caring. Journal of ioral ethics in organizations: A review. Journal of Man- Personality and Social Psychology, 86: 629–647. agement, 32: 951–990. Watley, L. D., & May, D. R. 2004. Enhancing moral intensity: Treviño, L. K., & Youngblood, S. A. 1990. Bad apples in bad Personal and consequential information in ethical deci- barrels: A causal analysis of ethical decision making sion-making. Journal of Business Ethics, 50: 105–126. behavior. Journal of Applied Psychology, 75: 447–476. Weaver, G. 2006. Virtue in organizations: Moral identity as a Van Dyne, L., & Pierce, J. L. 2004. Psychological ownership and foundation for moral agency. Organization Studies, 27: feelings of possession: Three field studies predicting em- 341–368. ployee attitudes and organizational citizenship behavior. Journal of Organizational Behavior, 25: 439–459. Werhane, P. 1999. Moral imagination and management de- cision making. New York: Oxford University Press. Veenman, M., & Elshout, J. J. 1999. Changes in the relation between cognitive and metacognitive skills during the Woodard, C., & Pury, C. 2007. The construct of courage: Cat- acquisition of expertise. European Journal of Psychology egorization and measurement. Consulting Psychology of Education, 14: 509–523. Journal: Practice and Research, 59: 135–147. Verplanken, B., & Holland, R. W. 2002. Motivated decision Woolfolk, R. L., Gara, M. A., Allen, L. A., & Beaver, J. D. 2004. making: Effects of activation and self-centrality of val- Self-complexity: An assessment of construct validity. ues on choices and behavior. Journal of Personality and Journal of Social and Clinical Psychology, 23: 463–474. Social Psychology, 3: 434–447. Yearley, L. H. 1990. Mencius and Aquinas: Theories of virtue Verschoor, C. C. 2003. Eight ethical traits of a healthy orga- and conceptions of courage. Albany: State University of nization. Strategic Finance, 85(3): 20–30. New York Press. Victor, B., & Cullen, J. B. 1988. The organizational bases of Young, M. E., & Wasserman, E. A. 2005. Theories of learning. ethical work climates. Administrative Science Quar- In K. Lamberts & R. Goldstone (Eds.), Handbook of cog- terly, 33: 101–125. nition: 161–182. Thousand Oaks, CA: Sage.

Sean T. Hannah ([email protected]) is a U.S. Army Colonel and the director of the Center for the Army Profession and Ethic at West Point–United States Military Academy. He received his Ph.D. from the University of Nebraska–Lincoln. His research interests include leadership development, moral development, and self-identity. Bruce J. Avolio ([email protected]) is the Marion B. Ingersoll Professor and the executive director of the Center for Leadership and Strategic Thinking in the Michael Foster School of Business, University of Washington. He received his Ph.D. from the Univer- sity of Akron. His current interests are in the areas of leadership development and distributed strategic leadership. Douglas R. May ([email protected]) is professor and codirector of the International Center for Ethics in Business, University of Kansas. He received his Ph.D. from the University of Illinois at Urbana-Champaign. His research interests include the rela- tions between moral efficacy, moral courage, moral meaningfulness, and constructive moral behaviors in the workplace. Copyright of Academy of Management Review is the property of Academy of Management 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.