Leader member exchange theory essay

Journal of Leadership & Organizational Studies 2016, Vol. 23(4) 456 –466 © The Authors 2016 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1548051816630227 jlo.sagepub.com

Article

Leader–member exchange (LMX) theory posits that fol- lowers’ work-related attitudes and behaviors depend on how their leaders treat them. According to social exchange theory (Blau, 1964), followers feel obligated to reciprocate by working hard to benefit their leader when treated favor- ably (Graen & Uhl-Bien, 1995). Several meta-analyses have provided strong support to LMX theory and found positive relationships between LMX quality and work per- formance (Gerstner & Day, 1997), citizenship behavior (Ilies, Nahrgang, & Morgeson, 2007), and attitudes such as affective and normative commitment and job satisfaction (Dulebohn, Bommer, Liden, Brouer, & Ferris, 2012). Finally, a recent meta-analysis of LMX across 23 countries concluded that the relationships of LMX with work perfor- mance and organizational commitment were not affected by national culture (Rockstuhl, Dulebohn, Ang, & Shore, 2012). Accordingly, a high-quality LMX relationship is a strong and robust predictor of followers’ work performance and important work-related attitudes. Furthermore, it is typ- ically assumed that LMX relationships fall on a single con- tinuum from high to low quality, where high-quality exchange relationships represent social exchange relation- ships, and low-quality exchange relationships represent economic exchange, or transactional exchange, relation- ships (e.g., Bernerth, Armenakis, Feild, Giles, & Walker, 2007; Graen & Uhl-Bien, 1995; Wayne et al., 2009). The other side of the coin is, therefore, that a transactional LMX

relationship will be associated with lower work perfor- mance and organizational commitment. Unlike a high-qual- ity, or social LMX relationship, a transactional LMX relationship is characterized by economic exchange behav- ior, formal role-defined relations, and unidirectional down- ward influence (Graen & Uhl-Bien, 1995). Indeed, Graen and Uhl-Bien (1995) argue that low-quality, or transactional LMX is analogous to transactional leadership in that “the leader makes requests based upon his/her hierarchical status within the organization, and the follower complies because of his/her formal obligation to the leader and because of the economic rewards the leader controls” (p. 232). In this case, followers’ motivations are assumed to be based on the satis- faction of self-interests rather than the prosocial motivation emanating from social exchange of favors (Graen & Uhl- Bien, 1995).

Unfortunately, some dyads may not advance much beyond transactional LMX, as social LMX relationships can be hard to develop owing to personality differences,

630227 JLOXXX10.1177/1548051816630227Journal of Leadership & Organizational StudiesBuch et al. research-article2016

1Oslo and Akershus University College of Applied Sciences, Oslo, Norway 2BI Norwegian Business School, Oslo, Norway

Corresponding Author: Robert Buch, Oslo and Akershus University College of Applied Sciences, Pilestredet 35, 0166 Oslo, Norway. Email: robert.buch@hioa.no

Transactional Leader–Member Exchange Relationships and Followers’ Work Performance: The Moderating Role of Leaders’ Political Skill

Robert Buch1, Geir Thompson2, and Bård Kuvaas2

Abstract In this study, we test whether leader political skill moderates the relationship between more transactional leader–member exchange (LMX) relationships and follower work performance. A field study was conducted among 753 followers and 153 leaders from several Norwegian business organizations. The results showed that the negative relationship between more transactional LMX relationships and follower work performance was weaker for employees with a highly politically skilled leader. Thus, leader political skill seemed in part to mitigate the negative effects of transactional LMX. This is an important observation, since finding alternative routes to enhanced work performance of followers in less favorable transactional LMX relationships is essential. Implications for future research and practice are discussed.

Keywords leader political skill, LMX, follower work performance

Buch et al. 457

style differences, differences in background, incongruent values, and so forth (Breland, Treadway, Duke, & Adams, 2007; Uhl-Bien, 2003). In addition, the time and resources of both followers and leaders are limited (Bauer & Green, 1996), and social LMX relationships require a fair amount of time and energy to be invested in relationship develop- ment. Thus, it follows that leaders favor particular employ- ees over others and only develop social LMX relationships with some, and more transactional LMX relationships with others (e.g., Graen & Uhl-Bien, 1995; Wayne et al., 2009). Accordingly, and to increase overall productivity, the iden- tification of leadership skills that may mitigate the negative impact of more transactional LMX is a theoretically and practically important research endeavor (Kuvaas, Buch, Dysvik, & Haerem, 2012). Furthermore, there is a dearth of knowledge regarding the impact of leaders’ skills on leader- ship effectiveness, and Yukl (2012) recently called for more research on how leader skills may combine with other vari- ables to influence performance.

In the present study, we therefore investigate the interac- tive roles of LMX and leader political skill in predicting follower performance. Specifically, we hypothesize and test whether leader political skill can moderate the negative relationship between more transactional LMX and follower work performance. Political skill refers to the “ability to effectively understand others at work and to use such knowledge to influence others to act in ways that enhance one’s personal and/or organizational objectives” (Ferris et al., 2005, p. 127). With an emphasis on skill, political skill is typically viewed as a something that can be devel- oped through training and socialization (Ferris, Anthony, Kolodinsky, Gilmor, & Harvey, 2002; Munyon, Summers, Thompson, & Ferris, 2015), although it probably also has dispositional antecedents (Liu et al., 2007). Followers in social LMX relationships deliver high performance mainly as a result of the prosocial motivation emanating from the relationship itself, probably relatively independently of their leader’s political skill. However, for followers in a more transactional LMX relationship, their leaders’ ability to understand others at work and to use this type of knowl- edge to influence followers to work harder or smarter may represent an instrumental pathway to reducing the negative impact of such LMX relationships.

Theory and Hypothesis

Because organizations can be considered political areas in which social and political skills are vital to managerial suc- cess (Mintzberg, 1983), researchers have begun to investi- gate the nature and the effects of political skill. Research has shown that leader political skill relates to ratings of leader effectiveness (Douglas & Ammeter, 2004), task and contextual job performance (Jawahar, Meurs, Ferris, & Hochwarter, 2008), and followers’ job satisfaction and

organizational commitment (Treadway et al., 2004). Thus, leader political skill shares many of the most important con- sequences of a favorable LMX relationship. Yet motivating followers through political skill is typically described as more instrumental than influencing them by developing trust-based, diffuse, and long-term relationships. Early the- orizing defined political skill as the ability to influence oth- ers through the use of manipulation, persuasion, and negotiation (Mintzberg, 1983). More recently, Ferris et al. (2005) argued that political skill is composed of four aspects: interpersonal influence, networking ability, social astuteness, and apparent sincerity. Because of their high levels of networking ability and social astuteness, leaders high in political skill have a good understanding of the workplace and are able to read situations and people well (Brouer, Harris, & Kacmar, 2011; Kimura, 2013). Because of this ability to diagnose situations, politically skilled lead- ers are able to select the most situationally appropriate behaviors to gain interpersonal effectiveness and to elicit the desired responses from those around them (Ewen et al., 2013). They do not only know exactly what to do in various social situations at work, but also how to do it in a way that disguises any self-serving motives, making them appear sincere (Ferris et al., 2005). Finally, politically skilled lead- ers should be able to influence employees by using influ- ence tactics (Ferris et al., 2007), such as coalition building, upward appeal (Kipnis, Schmidt, & Wilkinson, 1980; Yukl & Falbe, 1990), and rational persuasion, defined as “using logic, factual evidence, or thorough explanations in order to persuade or influence others” (Kolodinsky, Treadway, & Ferris, 2007, p. 1751). The use of political skill is arguably particularly advantageous in situations that require interper- sonal skill and social influence tactics (Bing, Davison, Minor, Novicevic, & Frink, 2011).

Our main premise is that a leader’s political skill is prob- ably not equally important to all followers in terms of facili- tating their work performance. Although it is theoretically possible for leaders to develop social exchange relation- ships with all their followers, the reality is that leaders tend to differentiate between followers and develop more social LMX relationships with a chosen few (Liden & Graen, 1980). Below, we argue that a leader’s political skill will be increasingly more important for follower performance the more transactional the LMX relationship is, and increas- ingly less important the more social the LMX relationship is. Stated differently, we argue that the relationship between LMX and follower work performance is moderated by leader political skill.

First, as pointed out by Graen and Uhl-Bien (1995), a fol- lower in a more transactional LMX relationships is probably motivated by “the satisfaction of his/her own self-interests, without consideration of the good of the group” (p. 232). Hence, the explanation for negative consequences of more transactional LMX relationships is probably that followers

458 Journal of Leadership & Organizational Studies 23(4)

in such relationships adopt a short-term and instrumental approach to their work and their leader (Kuvaas et al., 2012). As an example, followers may not be willing to take on any extra work unless they know precisely what they will get in return for doing so. Leaders high in political skill can miti- gate or compensate for a transactional LMX relationship because they have been shown to set challenging goals, cre- ate accountability, lead by example, and incentivize employ- ees by rewarding goal attainment (Ewen et al., 2013; Smith, 2009). For instance, the politically skilled leader may be bet- ter at bargaining the terms of the more transactional agree- ment as well as framing (Munyon et al., 2015) the terms in such a way that the follower is incentivized to take on the extra work. Furthermore, the more transactional the LMX relationship, the more necessary and influential such behav- iors would probably be in increasing follower work perfor- mance. Followers in more transactional LMX relationships probably need an extra stimulus to perform well, as they worry about their self-interest and future returns and there- fore withhold effort (Buch, Martinsen, & Kuvaas, 2015; Kuvaas et al., 2012).

It is also likely that followers in more transactional LMX relationships engage in feedback behaviors and political influence tactics that require a highly politically skilled leader. Research has, for instance, shown that resource-con- strained conditions (low relative LMX/low perceived orga- nizational support) increase the frequency of upward influence attempts and that employees are likely to employ levels of both hard and soft tactics with a transactional leader (Epitropaki & Martin, 2013). High political skill may allow leaders to respond properly to these influence tactics and better explain to these followers the means for improv- ing their status so they are able to benefit from desired per- formance goal improvements.

Leader political skill can also act as a substitute for a more social LMX, by providing followers with some of the social benefits associated with a favorable leader–follower relationship. Leaders with higher levels of political skill are able to build large bases of social and organizational capital (Ferris et al., 2005). Such capital implies that politically skilled leaders have more resources available to compen- sate followers in more transactional LMX relationships. In support of this, followers of politically skilled leaders have been shown to exhibit higher levels of perceived organiza- tional support, job satisfaction, and commitment (Treadway et al., 2004). Such resource availability may in turn signal to followers in more transactional LMX relationships that their LMX status can be altered.

Accordingly, followers in transactional LMX relation- ships might benefit more from higher levels of leader politi- cal skill by being more instrumentally motivated to increase work effort. Conversely, followers in more social LMX relationships should exhibit higher levels of work perfor- mance because they have a more strongly felt obligation to

reciprocate (see meta-analytic reviews by Dulebohn et al., 2012; Gerstner & Day, 1997) with lower needs for leaders with strong political skills. As noted by Blau (1964, p. 16), “If we feel grateful and obligated to an associate for favors received, we shall seek to reciprocate his kindness by doing things for him.” This is consistent with findings by Erdogan, Kraimer, and Liden (2004), showing that followers in more social LMX relationships are less influenced by work value congruence when it comes to career satisfaction than fol- lowers in more transactional LMX relationships. Accordingly, we hypothesize the following:

Hypothesis 1: The negative relationship between more transactional LMX relationships and work performance is moderated by leader political skill. The higher the political skill, the less negative the relationship.

Method

Participants and Procedure

The respondents were recruited through students enrolled in executive education programs at the business school where two of the authors are employed. Data were collected from supervisors and their employees at different organizational levels (top, middle, and operational) in 35 medium to large business organizations located in eastern part of Norway. The respondents were full-time employed. Each organiza- tion provided us with information that displayed the organi- zational structure, showing leaders and their direct reports. Additionally, we were given access to leaders’ and follow- ers’ e-mail addresses. We distributed questionnaires to the respondents while they were at work. One online survey was used to collect data from the followers, and a separate online survey was used to collect data from the leaders. The respon- dents were informed that participation was voluntary, that the data were being collected for academic research, and confidentiality assurances were given. Every respondent was given a unique link to the questionnaires where they could respond to each item. A total of 987 followers and 148 leaders responded (the overall response rate was nearly 81%). Merging the followers’ and leaders’ responses pro- vided complete data for 753 followers and 153 leaders. Of the leaders, 69.3% were men, and 30.7% were women. Of the followers, 65.3% were men and 34.7% were women. The average age of the leaders was 41.8 years, reporting an average education of 15.1 years. Follower average age and education were 42.9 and 13.9 years, respectively.

Measures

Transactional LMX. Our measurement of transactional LMX (α = .88) by having followers fill out Graen and Uhl-Bien’s (1995) widely cited seven-item scale (LMX-7) is based on

Buch et al. 459

the assumption in the extant literature that (low-quality) transactional LMX and (high-quality) social LMX relation- ships represent opposite poles on a single continuum (e.g., Bernerth et al., 2007; Graen & Uhl-Bien, 1995; Wayne et al., 2009). To critically assess the unidimensionality of this bipolar construct, we performed a supplemental explor- atory factor analysis. The results of this exploratory factor analysis clearly showed that the seven items loaded on a single factor with an average factor loading of .72. Hence, to facilitate the interpretation of the results we reverse- coded the respondents’ LMX-7 scores such that higher scores would represent a more transactional LMX relation- ship. Consequently, the reversed scale represents subordi- nate perceptions of having a low-quality (transactional) exchange, rather than a high-quality (social) exchange rela- tionship with their leader. Sample items include “How well does your supervisor understand your job problems and needs?” and “Regardless of how much power my supervi- sor has built into his/her position, my supervisor would be personally inclined to use his/her power to help me solve problems in my work.” Respondents recorded their responses on a 7-point scale (1 = strongly disagree to 7 = strongly agree).

Leader Political Skill. To measure leader political skill (α = .88), we had leaders fill out the political skill inventory by Ferris et al. (2005). Sample items include “I always seem to instinctively know the right things to say or do to influence others” and “I am particularly good at sensing the motiva- tions and hidden agendas of others.” Respondents recorded their responses on a 7-point scale (1 = strongly disagree to 7 = strongly agree).

Follower Work Performance. To measure follower work per- formance (α = .92), we obtained leader ratings of work per- formance by having supervisors fill out a five-item scale (Liden & Graen, 1980). The aspects of performance evalu- ated included dependability, planning, know-how and judg- ment, overall present performance, and expected future performance. The items were scored using a 7-point scale (1 = unsatisfactory, 7 = outstanding).

Control Variables. To rule out alternative explanations of the observed relationships between more transactional LMX and follower work performance, we controlled for the age and gender of both the leader and the follower. Further- more, given that relational demography could account for variance in LMX, we controlled for age and gender differ- ences or similarities between the leaders and the followers (Kacmar, Witt, Zivnuska, & Gully, 2003; Turban & Jones, 1988), where similar genders between leaders and followers was coded as “0” and different genders as “1.” We calcu- lated age difference as the absolute difference in age between the leader and follower. Finally, we controlled for

span of supervision, as the number of followers per leader might influence interaction frequency and thus the nature of the LMX relationship (Antonakis & Atwater, 2002).

Data Analysis

To examine the adequacy of our measurement model we performed a confirmatory factor analysis (CFA). Since “ordinal variables are not continuous and should not be treated as if they are” (Jöreskog, 2005, p. 10), we used the WLSMV estimator (Muthén, du Toit, & Spisic, 1997) to accommodate the ordered categorical data (e.g., Flora & Curran, 2004). In addition, because the observations in the data set are not independent (i.e., followers clustered within different leaders), the CFA was performed using cluster robust standard errors at the leader level.

Because our data are hierarchical (followers nested within leaders), there may be concerns about “leader effects” for some of the variables (e.g., Brouer, Douglas, Treadway, & Ferris, 2013). Therefore, we used hierarchical linear modeling to test our hypothesis. The appropriateness of using this method was underlined by the estimation of a fully unconditional model (null model) for follower work performance, which demonstrated significant between- group variability in leader ratings of follower work perfor- mance (τ

00 = .17, p < .05). To explore the nature of the

statistically significant interaction, we plotted low versus high scores on transactional LMX and leader political skill (one standard deviation below and above the means using nonstandardized scores; Cohen, Cohen, West, & Aiken, 2003).

Results

The results of a CFA performed on a three-factor model rep- resenting LMX, leader political skill, and follower work performance fit the data well (χ2[402] = 642.05, p < .01; root mean square error of approximation = .024; compara- tive fit index = .97; Tucker–Lewis index = .97) according to commonly used rules of thumb (Bollen, 1989; Fan, Thompson, & Wang, 1999; Steiger, 1989). Furthermore, the scales demonstrated high internal consistency, with coeffi- cient alphas ranging from α = .88 to α = .92. We report descriptive statistics, reliability estimates, and bivariate cor- relations among the study variables in Table 1, and we pres- ent the results of the hierarchical linear modeling analyses in Table 2.

In Step 1, we entered the control variables. Results indicated that leaders’ span of supervision relates nega- tively to followers’ work performance (γ = −.02, p < .05). In Step 2, we entered transactional LMX. The results indi- cated when controlling for leaders’ age, gender, span of supervision, and tenure, there is a negative relationship between transactional LMX and follower work

460 Journal of Leadership & Organizational Studies 23(4)

performance (γ = −.38, p < .001). Next, in Step 3, results indicated that leader political skill (γ = .21, p < .01) relates positively to follower work performance. Finally, in Step 4, we entered the interaction term (transactional LMX × Leader political skill). The introduction of the interaction term (γ = −.17, p < .05) resulted in increase in the pseudo R2 (ΔR2 = .01) and a significant decrease in model deviance (Δχ2 = 4.13, p < .05), thus showing that the relationship

between transactional LMX and follower work perfor- mance is moderated by leader political skill. Specifically, the results displayed in Figure 1 demonstrate a weaker negative relationship between transactional LMX and fol- lower work performance for followers whose leaders reported higher levels of political skill (b

low = −.47, p <

.001, vs. b high

= −.25, p < .001). A supplemental t test revealed significantly different slopes for high versus low

Table 1. Descriptive Statistics, Scale Reliabilities, and Correlations.

Mean SD 1 2 3 4 5 6 7 8 9

1. Leader’s age 41.81 7.70 2. Leader’s gendera .31 .46 −.06 3. Leader’s span of

supervision 11.79 9.03 .14** −.08*

4. Follower’s agea 41.49 11.01 .16** −.12** −.13* 5. Follower’s gender .35 .48 −.08* .26** −.02 −.03 6. Age differenceb 10.15 8.85 −.00 −.03 .11** .10** −.01 7. Gender differencec .33 .47 −.01 .17** −.03 −.03 .36** −.02 8. Transactional LMX 2.13 .62 −.02 −.04 .14** .03 −.00 .10** .00 (.88) 9. Leader political skill 5.43 .61 −.11** .00 .11** −.16** .08* −.02 .11** −.09** (.88) 10. Follower work

performance 5.58 .98 .05 .07* −.23** −.01 .10** −.07 .04 −.28** .13** (.92)

Note. N = 753. LMX = leader–member exchange. aMale = 0; female = 1. bWe calculated age difference as the absolute difference in age between the leader and follower. cWe coded similar genders between leaders and followers as “0” and different genders as “1.” *p < .05. **p < .01.

Table 2. Results of Hierarchical Linear Modeling Analyses.

Follower work performance

Variables Step 1 Step 2 Step 3 Step 4

Intercept 5.58*** 5.54*** 5.52*** 5.53*** Leader’s age .01 .01 .01 .01 Leader’s gendera .02 .01 .02 .01 Leader’s span of supervision −.02* −.02* −.02** −.02** Follower’s age .00 .00 .00 .00 Follower’s gendera .17* .17* .16* .16* Age differenceb −.01 −.00 −.00 −.00 Gender differencec −.05 −.04 −.06 −.05 Transactional LMX −.38*** −.37*** −.36*** Leader political skill .21** .22** Transactional LMX × Leader political skill

−.17*

Pseudo R2 .09 .15 .17 .18 ΔR2 .06 .02 .01 Model deviance χ2 1921.53 1872.35 1863.40 1859.27 Decrease in deviance: Δχ2d 49.18*** 8.95** 4.13*

Note. N = 753. Nonstandardized coefficients are displayed. LMX = leader–member exchange. aMale = 0; female = 1. bWe calculated age difference as the absolute difference in age between the leader and follower. cWe coded similar genders between leaders and followers as “0” and different genders as “1.” dThe full ML estimator was used to calculate this decrease in deviance (Δχ2) which can be considered a way of expressing effect size in multilevel modeling. *p < .05. **p < .01. ***p < .001.

Buch et al. 461

levels of leader political skill (t = 2.04, p < .05). Accordingly, we received support for Hypothesis 1.

Finally, researchers have argued that there is a need to better comprehend the four subdimensions of political skill and how they may operate individually (e.g., Ferris, Treadway, Brouer, & Munyon, 2012). Accordingly, to explore which of the dimensions of political skill that can explain the significant interaction, we performed post hoc moderation analyses using the four subdimensions of politi- cal skill (i.e., social astuteness, interpersonal influence, net- working ability, and apparent sincerity).1 In these analyses, we entered the control variables in Step 1, transactional LMX in Step 2, the four subdimensions of political skill in Step 3, and all the interaction terms (i.e., transactional LMX × Social astuteness, transactional LMX × Interpersonal influence, Transactional LMX × Networking ability, and Transactional LMX × Apparent sincerity) in Step 4. The results revealed that the only product term reaching statisti- cal significance was that of transactional LMX and net- working ability (γ = .13, p < .05). In line with the initial results using the single composite measure of leader politi- cal skill, the results displayed in Figure 2 suggest a weaker negative relationship between transactional LMX and fol- lower work performance with higher levels of networking ability (b

low = −.49, p < .001, vs. b

high = −.24, p < .01).

Discussion

Leaders do not treat all followers alike, differentiate between followers, and develop social exchange relation- ships with some and economic exchange, or transactional

relationships with others (Graen & Uhl-Bien, 1995). Consistent with prior research (e.g., Gerstner & Day, 1997; Rockstuhl et al., 2012) our results suggest that more trans- actional LMX relationships are related to lower work per- formance, and more social LMX relationships are related to higher work performance. Since many dyads do not advance much beyond transactional LMX finding alternative routes to the increased work performance of followers in these relationships is essential. In this respect, we have extended research on LMX relationships and follower work perfor- mance by investigating the interactive role of LMX and leader political skill. In support of our hypothesis, the form of moderation reveals a weaker negative relationship between transactional LMX and follower work perfor- mance with higher leader political skill. Accordingly, we found that leader political skill did, to some extent, mitigate the negative relationship between more transactional LMX relationships and follower work performance. The politi- cally skilled leader has the ability to read people and such situations well (i.e., social astuteness) and can therefore act to influence followers by applying the influence tactics that the situations demand. This probably makes politically skilled leaders better able to incentivize followers to per- form well in situations where followers worry about their self-interest and future returns and therefore withhold effort (Buch et al., 2015; Kuvaas et al., 2012). This can be done, for instance, by tailoring the terms of more transactional exchanges (e.g., provision of rewards in exchange for fol- lower performance) to fit the followers’ needs and wishes. In addition, as noted by Munyon et al. (2015, p. 172) politi- cal skill should “enhance the accuracy of perspective taking

Figure 1. The moderating role of leader political skill on the relationship between transactional LMX and follower work performance.

462 Journal of Leadership & Organizational Studies 23(4)

and empathic processes, which facilitate improved negotia- tion and conflict resolution.” The politically skilled leader should thus be able to take the perspective of the follower with whom she/he has a more transactional exchange rela- tionship with and negotiate the exchanges while appearing genuine and sincere, with no ulterior motives (Munyon et al., 2015). Our findings may thus indicate that leader political skill acts as a proxy for more social LMX relation- ships with respect to inducing follower performance. As such, the current study contributes to clarifying the interac- tive impact of leader characteristics (i.e., leader political skill) and situational criteria (leader–member relationship quality) in predicting follower work performance, and rep- resents a step toward a answering Yukl’s (2012) call for more research on how leader skills may combine with other variables to influence performance.

Furthermore, we observed a moderately strong positive relationship between leader political skill and follower work performance, in line with prior research (e.g., Ahearn, Ferris, Hochwarter, Douglas, & Ammeter, 2004; Treadway et al., 2004). Together, the findings of the present study reaffirms previous research indicating that political skill is one of the most essential competencies a leader can possess (Treadway et al., 2004). Still, even though there was a markedly weaker negative relationship between transac- tional LMX and work performance for leaders with strong political skills, political skill may seem to represent a some- what poor consolation for more transactional LMX rela- tionships as the observed relationship was still negative. However, taking into account that LMX is a strong and

robust predictor of work performance and other important employee outcomes, the mitigating role of leader political skill is not negligible.

Finally, our post hoc analyses suggest that networking ability is the most important political skill a leader can have with respect to mitigating the negative relationship between more transactional LMX and follower work performance. Although more research is needed, this observation aligns well with our theorizing that leaders with higher levels of political skill are able to build large bases of social and organizational capital (Ferris et al., 2005), which provides them with more resources available to compensate follow- ers in more transactional LMX relationships. Consider the example above, where followers in more transactional LMX relationships withhold effort because they worry about their self-interest and future returns. In this case, the benefits of having more resources available to negotiate better follower performance in return for some tangible reward enables the leader to match the available resources with the needs and wishes of the follower. The leader could for instance leverage extra vacation days, participation in decision making, interesting tasks, more responsibility, and so forth, in return for extra effort. Indeed, as noted by Ferris et al. (2012) networking ability is expected to be important for the effectiveness of leaders since “well positioned and networked leaders are better able to maximize resource attainment for their units, which builds greater appreciation and support from their teams” (p. 510). Furthermore, Sparrowe and Liden (2005) theorized that the “structure and composition of an individual’s network of informal

Figure 2. Post hoc analysis of the moderating role of networking ability on the relationship between transactional LMX and follower work performance.

Buch et al. 463

contacts allows him or her to identify strategic opportuni- ties, marshal resources, assemble teams, and win support for innovative projects that benefit the organization” (p. 505). Finally, a leader with networking ability is believed to have the capability of effectively handling negotiations and conflict (Ferris et al., 2012), which should mitigate poten- tial conflicts that may arise in more transactional LMX rela- tionships. Taken together, the present study aligns well with previous research demonstrating the importance of net- working ability for leader effectiveness (Douglas & Ammeter, 2004; Ferris et al., 2012) and represents a step toward a better understanding of the different roles of the four dimensions of political skill.

Limitations, Strengths, and Research Directions

Our study has certain limitations that need to be acknowl- edged. First, we cannot demonstrate causal relations between the variables. For causal inferences to be drawn, experimental studies would be necessary (Shadish, Cook, & Campbell, 2001).

Second, we relied on same-source data with respect to the measurement of the moderator (leader political skill) and the dependent variable (follower work performance). This may cause concerns about common method bias and inflated ratings (Podsakoff, MacKenzie, & Podsakoff, 2012). Still, there was a modest correlation between leader political skill and follower work performance (r = .13), and our results align well with prior research where leader polit- ical skill and follower outcomes had been assessed indepen- dently (e.g., Ahearn et al., 2004; Ewen et al., 2013). Furthermore, while one of the procedural remedies recom- mended by Podsakoff, MacKenzie, Lee, and Podsakoff (2003) is separating the measurement of the predictor and the criterion variables in time, a strength of the current study is that we obtained LMX ratings (the predictor vari- able) and leader performance ratings (criterion variable) from different sources. Doing so represents another proce- dural remedy to alleviate potential common method vari- ance (Podsakoff et al., 2012). Still, future research should ideally measure LMX and follower work performance inde- pendently and include a temporal gap between their measurements.

Relatedly, given that we gathered corresponding data from different sources (leaders and followers), we could not guarantee the respondents’ complete anonymity, as an identi- fying variable was required to match the data. As noted by Podsakoff et al. (2003), a lack of anonymity could “reduce their willingness to participate or change the nature of their responses” (p. 887). However, in attempting to alleviate such potential problems, we informed the respondents that a link- ing variable not related to their identity was used and assured them strict confidentiality. In this respect, Podsakoff et al. (2003) argue that emphasizing respondent confidentiality

should reduce the likelihood that respondents “edit their responses to be more socially desirable, lenient, acquiescent, and consistent with how they think the researcher wants them to respond” (p. 888).

With one-time measurement of LMX, researchers (e.g., Graen & Scandura, 1987) have advocated the measurement of LMX from the perspective of the followers, since leaders are more likely to answer in a socially desirable way indi- cating they treat all followers the same (Howell & Hall- Merenda, 1999). Although our research satisfies this condition, research indicates that leaders’ and followers’ perceptions of the relationship frequently do not converge (e.g., Sin, Nahrgang, & Morgeson, 2009). Future research should thus measure LMX from the perspective of both the follower and the leader to see whether similar results would be obtained.

Finally, the problem of generalizability poses a potential limitation, as our sample is characterized by mostly male lead- ers and followers in a specific culture (Western). Accordingly, future research should investigate whether our findings can be replicated across different contexts and cultures.

In addition to performing similar studies in other con- texts by means of experimental designs, an interesting ave- nue for future research might be to incorporate the notion of relative LMX or a focal individual’s LMX relative to the LMXs of coworkers (Hu & Liden, 2013; Vidyarthi, Liden, Anand, Erdogan, & Ghosh, 2010). For instance, it may be that leader political skill may fully mitigate the negative relationship between transactional LMX and follower work performance when followers perceive to be relationally dis- advantaged in comparison with their coworkers. In such a scenario, a leader’s political skill may be of utmost impor- tance for the compensation of lack of a favorable LMX relationship.

Practical Implications

Notwithstanding the aforementioned limitations, our study holds some potentially important practical implications. First, if it is difficult to learn the abilities necessary to develop more social, as opposed to more transactional LMX relationships as perceived by followers (Uhl-Bien, 2003), organizations should select and promote candidates to lead- ership positions who already have these abilities. Second, organizations may draw on our findings to tailor selection and/or promotion practices toward politically skilled indi- viduals. Finally, since political skill is viewed as theoreti- cally malleable, training programs could be designed to develop the talents of those with lower levels of political skill (Ahearn et al., 2004). In this respect, Munyon et al. (2015, p. 167) suggest that “organizations that provide ongoing training and professional development opportuni- ties for their employees could include workshops on politi- cal skill in their course offerings” and note that such

464 Journal of Leadership & Organizational Studies 23(4)

organization-sponsored training should be team-based and include roleplay and simulations. Organizations could also help leaders develop political skills by providing formal mentoring programs. Blass and Ferris (2007), for instance, viewed political skill as representative of a form of tacit personal learning that could be developed through a form of “relational job learning” (p. 8) where they learn the “infor- mal rules of the game” (p. 9). Indeed, research on the indi- vidual dimensions of political skill suggests that networking ability is predicted by mentoring (e.g., Ferris et al., 2008). Hence, political skill, and networking ability in particular, can probably be fostered in organizations by providing indi- viduals with experienced, politically skilled mentors (Ferris et al., 2012). This might be especially useful when the size of a leader’s span of supervision exceeds his or hers ability to develop social exchange relationships with all followers (Schyns, Maslyn, & van Veldhoven, 2012).

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Note

1. We would like to thank one of the anonymous reviewers for this idea.

References

Ahearn, K. K., Ferris, G. R., Hochwarter, W. A., Douglas, C., & Ammeter, A. P. (2004). Leader political skill and team perfor- mance. Journal of Management, 30, 309-327. doi:10.1016/j. jm.2003.01.004

Antonakis, J., & Atwater, L. (2002). Leader distance: A review and a proposed theory. Leadership Quarterly, 13, 673-704.

Bauer, T. N., & Green, S. G. (1996). Development of leader-mem- ber exchange: A longitudinal test. Academy of Management Journal, 39, 1538-1567. doi:10.2307/257068

Bernerth, J. B., Armenakis, A. A., Feild, H. S., Giles, W. F., & Walker, H. J. (2007). Leader-member social exchange (LMSX): Development and validation of a scale. Journal of Organizational Behavior, 28, 979-1003. doi:10.1002/Job.443

Bing, M. N., Davison, H. K., Minor, I., Novicevic, M. M., & Frink, D. D. (2011). The prediction of task and contextual perfor- mance by political skill: A meta-analysis and moderator test. Journal of Vocational Behavior, 79, 563-577. doi:10.1016/j. jvb.2011.02.006

Blass, F. R., & Ferris, G. R. (2007). Leader reputation: The role of mentoring, political skill, contextual learning, and adapta- tion. Human Resource Management, 46, 5-19. doi:10.1002/ hrm.20142

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

Bollen, K. A. (1989). Structural equations with latent variables. New York, NY: Wiley.

Breland, J. W., Treadway, D. C., Duke, A. B., & Adams, G. L. (2007). The interactive effect of leader-member exchange and political skill on subjective career success. Journal of Leadership & Organizational Studies, 13(3), 1-14.

Brouer, R. L., Douglas, C., Treadway, D. C., & Ferris, G. R. (2013). Leader political skill, relationship quality, and leader- ship effectiveness: A two-study model test and constructive replication. Journal of Leadership & Organizational Studies, 20, 185-198. doi:10.1177/1548051812460099

Brouer, R. L., Harris, K. J., & Kacmar, K. M. (2011). The mod- erating effects of political skill on the perceived politics-out- come relationships. Journal of Organizational Behavior, 32, 869-885. doi:10.1002/job.718

Buch, R., Martinsen, Ø. L., & Kuvaas, B. (2015). The destruc- tiveness of Laissez-Faire leadership behavior: The mediat- ing role of economic leader-member exchange relationships. Journal of Leadership & Organizational Studies, 22, 115- 124. doi:10.1177/1548051813515302

Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple regression/correlation analysis for the behavioral sciences (3rd ed.). London, England: Erlbaum.

Douglas, C., & Ammeter, A. P. (2004). An examination of leader political skill and its effect on ratings of leader effective- ness. Leadership Quarterly, 15, 537-550. doi:10.1016/j. leaqua.2004.05.006

Dulebohn, J. H., Bommer, W. H., Liden, R. C., Brouer, R. L., & Ferris, G. R. (2012). A meta-analysis of antecedents and con- sequences of leader-member exchange: Integrating the past with an eye toward the future. Journal of Management, 38, 1715-1759. doi:10.1177/0149206311415280

Epitropaki, O., & Martin, R. (2013). Transformational– transactional leadership and upward influence: The role of relative leader–member exchanges (RLMX) and perceived organizational support (POS). Leadership Quarterly, 24, 299- 315. doi:10.1016/j.leaqua.2012.11.007

Erdogan, B., Kraimer, M. L., & Liden, R. C. (2004). Work value congruence and intrinsic career success: The compensatory roles of leader-member exchange and perceived organiza- tional support. Personnel Psychology, 57, 305-332.

Ewen, C., Wihler, A., Blickle, G., Oerder, K., Ellen Iii, B. P., Douglas, C., & Ferris, G. R. (2013). Further specification of the leader political skill–leadership effectiveness relation- ships: Transformational and transactional leader behavior as mediators. Leadership Quarterly, 24, 516-533. doi:10.1016/j. leaqua.2013.03.006

Fan, X. B., Thompson, B., & Wang, L. (1999). Effects of sample size, estimation method, and model specification on structural equation modeling fit indices. Structural Equation Modeling: A Multidisciplinary Journal, 6, 56-83.

Ferris, G. R., Anthony, W. P., Kolodinsky, R. W., Gilmor, D. C., & Harvey, M. G. (2002). Development of political skill. In C. Wankel & R. DeFillippi (Eds.), Research in management education and development. Volume 1: Rethinking manage- ment education for the 21st century (pp. 3-25). Greenwich, CT: Information Age.

Ferris, G. R., Blickle, G., Schneider, P. B., Kramer, J., Zettler, I., Solga, J., . . . Meurs, J. A. (2008). Political skill con-

Buch et al. 465

struct and criterion-related validation: A two-study inves- tigation. Journal of Managerial Psychology, 23, 744-771. doi:10.1108/02683940810896321

Ferris, G. R., Treadway, D. C., Brouer, R. L., & Munyon, T. P. (2012). Political skill in the organizational sciences. In G. R. Ferris & D. C. Treadway (Eds.), Politics in organizations: Theory and research considerations (pp. 487-529). New York, NY: Routledge/Taylor & Francis.

Ferris, G. R., Treadway, D. C., Kolodinsky, R. W., Hochwarter, W. A., Kacmar, C. J., Douglas, C., & Frink, D. D. (2005). Development and validation of the politi- cal skill inventory. Journal of Management, 31, 126-152. doi:10.1177/0149206304271386

Ferris, G. R., Treadway, D. C., Perrewé, P. L., Brouer, R. L., Douglas, C., & Lux, S. (2007). Political skill in organizations. Journal of Management, 33, 290-320.

Flora, D. B., & Curran, P. J. (2004). An empirical evaluation of alternative methods of estimation for confirmatory factor analysis with ordinal data. Psychological Methods, 9, 466- 491.

Gerstner, C. R., & Day, D. V. (1997). Meta-analytic review of leader-member exchange theory: Correlates and con- struct issues. Journal of Applied Psychology, 82, 827-844. doi:10.1037/0021-9010.82.6.827

Graen, G., & Scandura, T. A. (1987). Toward a psychology of dyadic organizing. In L. L. Cummings & B. M. Staw (Eds.), Research in organizational behavior (Vol. 9, pp. 175-208). Greenwich, CT: JAI Press.

Graen, G., & Uhl-Bien, M. (1995). Relationship-based approach to leadership: Development of leader-member exchange (LMX) theory of leadership over 25 years: Applying a multi-level multi-domain perspective. Leadership Quarterly, 6, 219-247. doi:10.1016/1048-9843(95)90036-5

Howell, J. M., & Hall-Merenda, K. E. (1999). The ties that bind: The impact of leader-member exchange, transformational and transactional leadership, and distance on predicting follower performance. Journal of Applied Psychology, 84, 680-694. doi:10.1037/0021-9010.84.5.680

Hu, J., & Liden, R. C. (2013). Relative leader–member exchange within team contexts: How and when social comparison impacts individual effectiveness. Personnel Psychology, 66, 127-172. doi:10.1111/peps.12008

Ilies, R., Nahrgang, J. D., & Morgeson, F. P. (2007). Leader- member exchange and citizenship behaviors: A meta- analysis. Journal of Applied Psychology, 92, 269-277. doi:10.1037/0021-9010.92.1.269

Jawahar, I. M., Meurs, J. A., Ferris, G. R., & Hochwarter, W. A. (2008). Self-efficacy and political skill as comparative predictors of task and contextual performance: A two-study constructive replication. Human Performance, 21, 138-157. doi:10.1080/08959280801917685

Jöreskog, K. G. (2005, February 2005). Structural equation mod- eling with ordinal variables using LISREL. Retrieved from http://www.ssicentral.com/lisrel/techdocs/ordinal.pdf

Kacmar, K. M., Witt, L. A., Zivnuska, S., & Gully, S. M. (2003). The interactive effect of leader-member exchange and com- munication frequency on performance ratings. Journal of Applied Psychology, 88, 764-772. doi:10.1037/0021- 9010.88.4.764