Reading: Leader Member Exchange Theory

Abstract

Leader–member exchange (LMX) theory is a relationship-based, dyadic theory of leadership. According to this theory, leadership resides in the quality of the exchange relationship developed between leaders and their followers. High quality exchanges are characterized by trust, liking, and mutual respect, and the nature of the relationship quality has implications for job-related well-being and effectiveness of employees. This article summarizes what LMX is, how it is measured, and the best practices in the study of LMX. Studies conducted on its antecedents, and past research related to its consequences in organizational settings are reviewed. In addition, we review issues related to multiple relationships such as differentiation and relative LMX. Finally, we conclude the article by identifying and summarizing key themes and questions for future research.

Motivating Creativity

It is worth noting, as Atwater and Cameli themselves did, that the focus was on creative involvement rather than creative output.

Tierney, Farmer, and Graen (1999), however, found a strong interaction between LMX and whether a person had an innovative or creative cognitive style (Kirton, 1976; LMX was strongly related to adaptive styles but not with innovative styles); that is, innovative people seem to be creative regardless of a leader’s LMX strategies (Elkins & Keller, 2003).

Creativity literature from the transformational leadership tradition is mixed. Elkins and Keller (2003) concluded from their review that the effects of transformational leadership in R&D organizations are moderated by intervening variables such as project type. Citing Keller (1992), they argued that transformational leadership had a stronger impact on research projects than on research development. They also noted the findings of Waldman and Atwater (1992), who found that the transformational leadership of higher level administrators was more effective for R&D success than it was when project managers expressed this type of leadership. This is generally consistent with the findings of Osborn and Marion (2009), who examined international innovation-seeking alliances. These researchers concluded that the transformational styles of managers in home offices that sponsored alliances were effective for producing a return to the home office, but that transformational styles were not particularly effective for the alliance itself; they further concluded that a more distributed, knowledge-based style (patterning of attention and network development) by project heads was more useful for alliance innovation.

The findings of some researchers would seem to contradict those of Osborn and Marion (2009). For example, Sosik, Kahai, and Avolio (1998) found that, for unidentified leaders, perceptions of transformational behaviors by the leaders increased the emergence of flow, which affected motivation and subsequent creativity. Mumford et al. (2002) may help explain these diverse findings. They observed that transformational leadership was not consistently related to creativity and proposed that the strong vision of a transformational leader may distract the attention of workers and limit their autonomy. This in turn could inhibit creativity. They subsequently proposed that transformational leadership may be (instead of is) related to creativity.

4.2 Cultural Collision and Leadership

Multicultural work settings have been around for centuries. Immigrant workers and cross-national trade have existed around the globe for centuries. However, the intensity of cross-cultural exchanges has a more noticeable economic impact in today’s global society than in the past. In previous centuries, those with resources and military might imposed their ways on others. Now, the competitive global economic market does not allow for a company or a nation to mistreat others for long, as there are competitors and potentially better options. Also, the presence of global communication systems has increased the accountability for social responsibility and human rights in business exchanges.

In today’s global economy, expatriates can be emigrant workers or expatriate managers of large, multinational corporations. Leader–member exchange theory can assist understanding the impact of culture on leadership in a multicultural setting.

The leader–member exchange defines leadership as a relationship. It has demonstrated that leaders have varying levels of intensity in their relationships with different employees. When the leader and the subordinates are in a close, trusting relationship, they enjoy an enriched relationship in which the leader is more apt to engage in transformational behavior styles. When the subordinates are at the periphery of the team, are not closely involved with the leader, and do not have a strong trusting relationship, the leader tends to behave more transactionally with them. Leader–member exchange theory, therefore, demonstrates that like any relationship, the leadership process also goes through a maturation process, from being strangers to being confidants. What makes this model appealing to multicultural setting is that at the core it assumes that leaders and subordinates develop unique dyad relationships.

In a multicultural setting, it is conceivable that leaders will develop closer relationships with subordinates who are similar to them; the leader may stay distant from those who are different. The research on this model is still young, yet the opportunities to explore the role of culture in leader–subordinate dyad relationship building are promising. Exploring the cultural factors that contribute to building strong leadership relationships in multicultural settings is necessary.

Independent of this theory of leadership, practitioners and cultural anthropologists have developed cultural guides to help expatriate leaders adjust to working in cross-cultural settings and with people from various cultures. However, in this type of cultural awareness training, two issues are overlooked. The first is that the cultural prototypes are used. That is, in this training, the leader is told what typically may occur in a country or what to expect from people of a certain culture. This can create stereotypes, as most people do not fit the cultural prototypes or stereotypes, and so can lead to misunderstanding. Second, this method of cultural awareness assumes that culture is static. Trompenaars and Hampden-Truner stated that international managers are faced with layers of cultural demands that are fluid and need balancing. In addition, in this type of training there is a tendency for the leader to try to act similar to the natives. This can be a no-win situation, as it has been demonstrated that natives do not trust expatriate leaders who act like them. Consequently, the value of knowing about the norms of the other culture is the awareness of the range of accepted behaviors that can equip leaders with alternative explanations when they observe behaviors that may be strange to their own cultural norms.

In summary, our knowledge about leadership in multicultural settings is still evolving. For years, cross-cultural trainers such as Foa and Triandis have recommended that those interacting in cross-cultural settings defer judgment and view situations through the eyes of the member of the other culture. They also developed the cultural assimilator training programs, which teach the individual faced with a heterogeneous team to be aware of his or her cultural assumptions and potential biases, as well as the beliefs and assumptions of the others. This technique trains the sojourner or expatriate to be more attentive and to think before acting.

Therefore, it is most helpful for the leaders in multicultural settings to be aware of the cultural norms of their own culture and those of the others. This awareness gives them a perspective of the range of behaviors to be expected. In addition, keeping an open mind and asking questions before judging is the best way to adjust to a team composed of various cultures and styles.

3.3 Power and Influence Approaches

Power approaches seek to explain leadership effectiveness in terms of the amount and type of power possessed and exercised by leaders. A classic distinction of forms of power was presented by French and Raven (1960): reward and coercive power, that is, the capability to offer incentives and to make use of organizational sanctions; legitimate power, when followers believe that the organizational or positional power over them is rightful; expert power, when high experience, knowledge, or ability are attributed; and referent power, when a person is referred to, or group norms are identified with, due to appealing personal qualities or values systems. Yukl (2001) describes three qualitatively different outcomes for the followers that result from employing these sources of power: commitment is most likely to be associated with referent and expert power, compliance with legitimate and reward power, and resistance with coercive power. The link between the use of power and behavioral approaches of leadership influence has been established by research on so-called influence tactics, for example, rational persuasion, consultation, ingratiation, exchange, coalition building, or pressure (cf. Yukl 1998).

Leader-member Exchange Theory (LMX) by Graen and his associates focuses on the development and the quality of the mutual relationship between leader and follower. Leaders are assumed to differentiate their followers according to their competence, trustworthiness, and motivation to assume ever more responsibility. Followers whom the leader perceives to display these attributes are categorized as so-called in-group members and in exchange are given more attention, support, and sensitivity by their leaders. The other, so-called out-group members, attract the more routine tasks and maintain a formal relationship with their leaders who in return exert influence with formal authority. Three ‘currencies of exchange’ between follower and leader, namely personal contribution, loyalty, and affect, have been identified. A meta-analytic review of LMX has been presented by Gerstner and Day (1997).

Transformational Leadership (Bass 1998) describes leaders to transform followers' attitudes and values, to activate their higher order motives and to stimulate them to transcend the organizations' higher order goals by their self-interests. The four components of transformational leadership (idealized influence, inspirational motivation, intellectual stimulation, and individualized consideration) are measured in the Multifactor Leadership Questionnaire (MLQ) and were found to be empirically distinguishable and to relate to work unit effectiveness. Transformational leadership is often contrasted with transactional leadership which is based primarily on compliance and norms recognized through leader–follower exchanges, that is, reward and praise are given by the leader for task completion and loyalty given by the follower.

Charismatic leadership is defined more narrowly than transformational leadership. It specifies an idolized leader's characteristics as perceived and attributed by followers. Behaviors typical of charismatic leaders are for example, the articulation of appealing visions, communication of high expectations, and expression of high confidence in followers. Some evidence of the principally positive relationship between personal charisma and effectiveness, especially in situations of crisis, was presented by House et al. (1991) in an investigation of US American presidents and various measures of national effectiveness during each president's term.

There is also a ‘dark side’ to charismatic leadership. Problems that can occur with ‘negative’ charismatics are, for example, that they seek to induce commitment to narrow-minded ideological goals or to themselves, start projects with often unrealistic premises, omit properly investing in the implementation of their visions, and fail to develop competent successors.

Humor in Leadership

A good sense of humor is often considered to be an important characteristic for a leader, along with other attributes such as intelligence, creativity, persuasiveness, and general communication skills. Research on leadership behavior indicates that effective leadership requires skills in the general areas of (1) giving and seeking information, (2) making decisions, (3) influencing people, and (4) building relationships (Yukl & LepSinger, 1990). These broad skill areas have been further divided into a variety of component behaviors, many of which are related to interpersonal relations and communication, such as the ability to communicate and get along well with subordinates, peers, and superiors, to manage conflict, motivate others, and enhance group cohesion and cooperation. Humor can be viewed as an important communication skill that is potentially useful to leaders and managers in many of these areas. For example, the use of humor could be beneficial for teaching and clarifying work tasks, helping to motivate and change behavior, promoting creativity, coping with stress, and generally making the interactions between the manager and subordinates more positive and less tense (Decker & Rotondo, 2001).

The psychology of leadership and ways in which leaders can be most effective has historically been a topic of considerable interest, but has attracted even greater attention over the past few decades. The manner in which leaders incorporate humor into their organizations is an important area of this research because it often plays a key role in how leaders communicate with their employees. Leader–member exchange (LMX) theory suggests that leaders and followers develop unique relationships based on their social exchanges, and the quality of these exchanges within an organization can influence employee outcomes (Graen & Uhl-Bien, 1995; Liden et al., 1997). Humor can be a contributing factor during such exchanges, and the frequency (Hughes & Avey, 2009) and type of humor (Wisse & Rietzschel, 2014) that leaders exhibit can shape followers’ perceptions of their leader.

Survey studies have examined the correlation between sense of humor and perceived leadership qualities by asking workers to rate their supervisors on various dimensions. In a survey of 290 workers, Wayne Decker (1987) found that those who rated their supervisors as being high in sense of humor also reported greater job satisfaction and rated these supervisors as having generally more positive leadership characteristics as compared to participants who rated their supervisors as low in sense of humor (see Fig. 11.4).

Sign in to download full-size image

Figure 11.4. Workers rated supervisors high in sense of humor as having generally more positive leadership characteristics as compared to those low in sense of humor.

Source: Adapted from “Managerial humor and subordinate satisfaction,” by W. H. Decker, 1987, Social Behavior and Personality: An International Journal, 15, p. 225. Copyright 1987 by Scientific Journal Publishers. Adapted with permission.

Similarly, Robert Priest and Jordan Swain (2002) found that military cadets perceived leaders they felt were good or effective as having a warmer, more competent, and benign humor style; they perceived leaders they felt were bad or ineffective as having a colder, ineffective, and mean-spirited humor style.

In his 2009 paper, Larry Hughes proposed that a leader’s style of humor delivery has an impact on workers’ positive emotions (e.g., joy, contentment, interest) and creative performance. He identified three main styles of humor delivery: self-deprecating (leaders make fun of themselves), other-deprecating (leaders make fun of the focal person in the social situation), and nonsense style (the target of the humor is not the leader nor another person in the situation, but a neutral target). Hughes further suggested that followers’ positive emotions mediate the relationship between leader humor and followers’ creative performance.

In a later study, Deog-Ro Lee (2015) examined how leaders’ humor styles influence employee creativity, as well as the moderating effects of employees’ trust in their leader. Using 316 leader–follower pairs from five telecommunication organizations in South Korea, Lee showed that leaders’ use of self-enhancing humor correlated positively with employees’ creativity, while leaders’ use of aggressive humor was negatively associated with employees’ individual creativity. Importantly, employees’ level of trust in their leader moderated the relationship between a leader’s use of self-enhancing humor and employee creativity; when employees trusted their leader, the leader’s use of self-enhancing humor was associated with greater employee creativity. In contrast, insofar as employees did not trust their leader, the leader’s use of self-enhancing humor was associated with less employee creativity.

Researchers have also shown interest in whether humor uses and their outcomes may differ for leaders based on gender or race. Wayne Decker and Denise Rotondo (2001), for instance, conducted a study to determine whether a leader’s use of positive and negative humor has different outcomes depending on the leader’s gender. Decker and Rotondo asked employees in a variety of organizations to indicate the extent to which their manager used positive humor (e.g., enjoyment of jokes, use of humor in daily conversation) and negative humor (e.g., sexual humor, ridicule). They also rated their manager’s overall leadership effectiveness. They found that employees rated their managers as more effective insofar as they used positive humor, and less effective insofar as they used negative humor. Furthermore, this pattern of results was stronger for female than male managers. That is, the use of positive humor by female versus male managers was more strongly associated with workers’ positive perceptions of their leadership skills. By the same token the use of sexual and other negative forms of humor was more strongly associated with workers’ negative perceptions of their leadership skills in female compared to male managers.

Overall, these studies provide evidence that employees perceive supervisors who use positive humor as more effective leaders; leaders who use humor inappropriately tend to receive more negative evaluations of their leadership skills. These findings could suggest that greater sense of humor might cause a leader to be more effective, but perhaps not. The findings could just reflect a “halo effect,” whereby greater overall liking of a supervisor may cause employees to perceive him or her as having a better sense of humor as well as better leadership skills. Further research is necessary to more fully delineate the direction of causality in the relationship between leaders’ use of humor and employee outcomes.

3.1 Leader–member exchange

Leader–member exchange theory evolves as an exchange process whereby subordinates may have high-quality exchanges or low-quality exchanges (Graen & Uhl-Bien, 1995). High-quality exchanges are typified by increased effort and personal loyalty to the leader and in exchange, leaders allow these followers more control and influence (Graen & Uhl-Bien, 1995; Schriesheim, Castro, Zhou, & Yammarino, 2001). Moreover, leaders may view followers with high-quality exchanges as the “in-group” or “advisors” and treat the group in a similar fashion, attributing importance to these followers as a subgroup.

Low-quality exchanges are typified by followers who produce efforts that do not exceed expectations and rely on organizational exchange patterns to guide interactions. As such, leaders in low-quality exchanges are less likely to initiate or provide additional benefits (Graen & Uhl-Bien, 1995; Schriesheim et al., 2001). Leaders also may view this subgroup as an “out-group” with less importance than the “in-group” and as such, attribute less value to member input received from the out-group.

Because of the unequal emphasis given to subgroups, LMX leadership can be viewed at the dyad-within-group level of analysis (Dionne & Dionne, 2008; Schriesheim et al., 2001). Concerning team convergence of mental models, as the process moves from individual to group, a leadership style emphasizing select subgroups over other subgroups may be detrimental to the convergence process. Specifically, within the composition form of emergence, LMX may inhibit the progress from individual level of analysis to team level of analysis in that the dyad-within-group focus of the leader stalls the movement to group level of analysis. Moreover, the unequal emphasis on resource and benefits acquisition between the in-group and the out-group is likely to affect all phases of the convergence process.

1.1 Leader–member exchange

In describing supervisor–subordinate dyadic relationships, leader–member exchange (LMX) theory proposed by Graen and his colleagues explains that the relationship between a superior and a subordinate develops because of their workplace interactions (Dansereau, Graen, & Haga, 1975; Graen & Scandura, 1987; Graen & Uhl-Bien, 1995; Graen & Wakabayashi, 1994; Liden & Graen, 1980). Based on role theory and social exchange theory, the LMX framework focuses on the differentiated exchange relationships that leaders develop and maintain with subordinates within workgroups, suggesting that leaders develop differential dyadic relationships with the various group members (Graen & Uhl-Bien, 1995). Evidence from prior work demonstrates that LMX substantially influences employees’ task commitment, task performance, job satisfaction, helping behaviors, and turnover intention (e.g., Gerstner & Day, 1997).

LMX was first posited as a multi-dimensional construct by both Dienesch and Liden (1986) and Liden and Maslyn (1998) who identified the potential for “currencies of exchange” for supervisors and subordinates aside from their work behaviors. This multidimensional approach to LMX has received favorable support (Dienesch & Liden, 1986; Liden & Maslyn, 1993; Liden & Maslyn, 1998; Maslyn & Uhl-Bien, 2001). Currencies of exchange in LMX theory include affect, loyalty, and/or professional respect between members. These currencies can occur in varying “amounts” and combinations.

Studies indicate that supervisors and subordinates focus on different currencies of exchange from their partners. For example, supervisors may seek more work-related currencies, while subordinates may prefer more socially related currencies (Day & Crain, 1992; Dockery & Steiner, 1990; Liden, Sparrowe, & Wayne, 1997). Importantly, currencies are perceived, so a work-related currency is a perceived contribution, and is thus defined as the “perception of the amount, direction, and quality of work-oriented activity each member puts forth toward the mutual goals (explicit or implicit) of the dyad” (Dienesch & Liden, 1986, p. 45). On the other hand, social currencies consist of perceived affect (i.e., “the mutual affection members of the dyad have with each other, based primarily on interpersonal attraction”), perceived loyalty (i.e., “the extent to which the leader and member are loyal to one another”), and perceived professional respect (i.e., “the perception of the degree to which each member of the dyad has built a reputation within and outside of the organization”) (Dienesch & Liden, 1986; Liden & Maslyn, 1998). Consistent with the reasoning behind this research, Liden et al. (1997) argue that treating LMX as a multidimensional construct is most appropriate and characteristic of the complex leader–member relationship, and thus allows for a more complete description of its multifaceted relationships (e.g., including important individual and organizational outcomes).

As we move toward a better understanding of which aspects of LMX relationships impact various individual attitudes and behaviors, we will not only be able to further develop and refine LMX theories, but also be in a better position to fill some critical (and gaping) lacunas in the extant research. For example, past research on supervisor–subordinate dyads has suggested that demographic characteristics (e.g., gender, age, organizational tenure) can play an important role in supervisor–subordinate expectations of their LMX quality. Specifically, high tenure differences in the supervisor–subordinate dyad are associated with lower levels of psychological attachment among work group members, while a supervisor's perceptions of similarity to a subordinate is positively related to the supervisor's rating on the subordinate's performance evaluations (Turban & Jones, 1988; Turban et al., 1990). In addition, gender differences in supervisor–subordinate dyadic relationships have been found to be related to selection for international assignments (Tejeda, 2006; Tsui & Gutek, 1999; Tsui et al., 2002; Turban & Jones, 1988; Varma & Storh, 2001). We also concur with researchers who argue that more attention on LMX needs to come from the standpoint of the supervisor's point of view (Liden et al., 1997). This argument is consistent with that of Maslyn and Uhl-Bien (2001) who sees LMX relationships as a series of dyadic exchanges.

Within the framework of LMX and the Malaysian workplace, prior studies have demonstrated links between LMX quality and various work outcomes. For example, LMX quality has been found to have a positive direct impact on perceived organizational citizenship behavior and satisfaction (Bhal, Gulati, & Ansari, 2009; Ishak & Alam, 2009; Lo, Ramayah, & Hui, 2006), and organizational commitment and turnover intention (Ansari, Hung, & Aafaqi, 2007; Khong, 2009; Lo, Ramayah, Min, & Songan, 2010). From the above, and both in and out of Malaysia, we see that previous research has provided valuable evidence regarding supervisor and subordinate demographic influences on the quality of LMX. However, this same research has bypassed the critical domains of supervisor–subordinate dyadic demographic effects on leader–member exchange relationships, as well as related subordinate's work attitudes. Nor has this research examined the effects of relational demography as a proxy of cultural norms. Studies such as this one thus fill a research void in that we are capturing and measuring the LMX relationship from both the supervisor's and the subordinate's perspective (see Greguras & Ford, 2006; Liden et al., 1997; Liden & Maslyn, 1998).

2.3.5 Leader–member exchange

The idea that leadership is based on a transaction or exchange between leaders and followers is emphasized in leader–member exchange (LMX) theory (Graen, Novak, & Sommerkamp, 1982; Graen & Scandura, 1987; Liden, Sparrowe, & Wayne, 1997). The focus in LMX theory is on how leaders and followers engage together to generate high quality work relationships that allow them to produce effective leadership outcomes (Graen & Uhl-Bien, 1995). The earliest work on vertical dyad linkage described these followers as trusted assistants to leaders who helped them meet the task-related challenges of the work unit (Dansereau, Graen, & Haga, 1975). Later work focused on how followers act as active partners to leaders when they operate in the context of high quality relationships (Uhl-Bien, Graen, & Scandura, 2000). While LMX theory does acknowledge followers in the relational process, it is still more leadership – than followership – focused in that it privileges the leader as the driver of the relationship-building process (Uhl-Bien et al., 2000).

Abstract

While leader–member exchange (LMX) has evolved, a richer understanding continues to evade scholars due to the sustained focus on the leader–member dyad. We argue that LMX theory remains incomplete until contextual factors surrounding these pivotal relationships are accounted for including the impact of coworker exchange relationships, peer exchange, and perceptions of justice regarding individual LMX relationships. A fully crossed experimental design manipulating participant LMX and distributive justice and coworker LMX and distributive justice was employed to understand how these constructs affect coworker exchange relationships. Support was found for a causal model which specifies that justice moderates the causal relationship between LMX similarity and CWX, which subsequently leads to increased sharing of resources among coworkers. The data suggest that workgroup members are savvy to differences within individual leader–member relationships, where the sharing of resources among peers is determined by their own LMX, the LMX of their coworker and perceptions of justice of their respective LMX levels. These findings are contextualized in extant leadership research and practice.