Assignment

Journal of Leadership & Organizational Studies 19(2) 142 –151 © Baker College 2012 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/1548051811433358 http://jlos.sagepub.com

Power in organizations can be interpreted as the potential influence that one individual exhibits over another (Emer- son, 1962; Pfeffer, 1992; Weber, 1947), and its study has intrigued scholars for decades. Sociological research focused on structural and authority/position-related expla- nations represented early attempts to explain hierarchical power (e.g., Emerson, 1962; Pfeffer & Salancik, 1978). However, much has remained to be understood about how individuals acquire power in organizations, when such power is not prescribed by hierarchical level or position that is formally designated (Brass & Burkhardt, 1993).With regard to leader–follower relationships, it is typical to think that leaders hold more power over their followers (Mintz- berg, 1983; Weber, 1947) because of the traditional hierar- chy of authority in bureaucratic organizational structures. However, there are certainly cases where the opposite is true (Mechanic, 1962; Pfeffer, 1992, 2010).

Differences in power between leaders and followers repre- sent significant issues for leader–follower attitudes, behavior, and work relationships, though little direct empirical research has been conducted to investigate the phenomenon (e.g., Ferris et al., 2009; Ragins & Dutton, 2007). This is rather surprising in light of the recent research attention focused on both the nature of work relationships and on shared leadership, power sharing between leaders and followers, and empowerment (e.g., Graen, 2009; Pearce & Conger, 2003). Power is an

important consideration because most researchers implicitly assume that leader–follower relationships are entered into, and maintained, by both parties volitionally (Rousseau & Schalk, 2000), which may not always be an accurate assumption. Furthermore, despite its recognized importance in the organi- zational sciences, power has remained under investigated in leader–follower relationships.

Therefore, the major objective of the present study is to examine the effects of power levels perceived by leaders and followers on central aspects of their work relationships (i.e., work relationship quality and job tension). As such, this investigation attempts to make contributions to the leader- ship, social power, and work relationships literatures.

Theoretical Foundations and Hypothesis Development A model of power in dyadic relationships is presented in Figure 1, which specifies that leader power affects work

433358 JLOs19210.1177/1548051811433358Martin ez et al.Journal of Leadership & Organizational Studies © Baker College 2012

Reprints and permission: http://www. sagepub.com/journalsPermissions.nav

1Illinois State University, Normal, IL, USA 2Florida State University, Tallahassee, FL, USA 3Florida Department of Children and Families, Tallahassee, FL, USA

Corresponding Author: Arthur D. Martinez, Illinois State University, 250 College of Business Building, Campus Box 5580, Normal, IL 61790-5580, USA Email: [email protected]

Power in Leader–Follower Work Relationships

Arthur D. Martinez1, Rachel E. Kane2, Gerald R. Ferris2, and C. Darren Brooks3

Abstract

There is perhaps no more important dyadic relationship than that between a leader and a follower. Nonetheless, few studies examine the implications of both leader and follower power on important work outcomes. Therefore, using resource dependence and role theories, the authors examined the process by which leader power affects important work outcomes, namely, work relationship quality and job tension, through met relationship expectations. Additionally, the authors suggest that the leader power–met expectations relationship is conditional on follower power. A state agency was sampled to obtain and analyze 100 leader–follower work relationship dyads, whereby both dyadic partners were surveyed. Results indicated that leader power affected both leader–follower relationship quality and job tension through followers’ met relationship expectations. However, contrary to our hypothesis, the leader power–met expectations relationship was not conditional on follower power. Contributions of this study, strengths and limitations, and directions for future research are discussed.

Keywords

leader–follower dyads, power, work relationships

Martinez et al. 143

relationship quality and employee well-being (i.e., job ten- sion) through followers’ met expectations of the leader– follower relationship. In addition, the model suggests that the effects of leader power on followers’ met expectations are moderated by the follower power, emphasizing the con- ditional nature of the relationship.

Leader Power There is perhaps no more important dyadic relationship than that between a leader and a follower (Ferris et al., 2009), and within such relationships, power and power dynamics are routinely at play and fundamentally inter- twined. Power stems from the notion of resource depen- dence, which maintains that the power of Person A over Person B is determined by the extent to which Person B is dependent on Person A for resources that are necessary for Person B to meet his/her needs, desires, and goals (Blau, 1964; Emerson, 1962, 1964).

The resource dependency perspective of power views social power as an attribute of social relations and structures not as an attribute of a person making up the relationship (Thibaut & Kelley, 1959). Within organizations, leaders (supervisors) are traditionally considered to hold more power over their followers (subordinates; Weber, 1947), and followers are assumed to be at least partly dependent on their leaders for both tangible resources (e.g., supplies) and intangible resources (e.g., self-verification, instrumental support; Farmer & Aguinis, 2005).

This common conceptualization of where power and dependence lie within the leader–follower relationship is likely spawned from individuals’ expectations concerning the roles characteristic of both leaders and followers. Early research on roles suggests that a portion of individual behavior can be explained by the roles one is perceived to hold and by one’s accompanying beliefs about such roles (Merton, 1957). Within the organizational context, role the- ory posits that leaders and followers engage early on in a role-making process; within such a role-making process,

both dyad members develop beliefs concerning the capabili- ties of the other dyad member, as well as about the outcomes expected from the relationship (Graen & Scandura, 1987; Tsui, 1984; Young & Perrewé, 2000). Furthermore, followers are considered members of their leader’s “role set,” which suggests that they interact, share interests with, and hold expectations of their leader (Katz & Kahn, 1978).

Although there are a wide range of qualities, attitudes, and behaviors likely to define the typical and ideal role of a leader, we suggest that followers should expect their leaders to be in possession of, among many things, power. We hypothesize that the notion of leaders being in possession of power coincides with many followers’ beliefs about the requirements and characteristics of a successful leader. This is to say that not only is power typically associated with those individuals in a hierarchical position to lead, but it is also likely considered a component necessary for the leader to be capable of fulfilling the dependencies of followers, be they tangible or intangible. Furthermore, power imbalances and power differential are traditional of, and to be expected of, relationships that span hierarchical levels. As such, based on role theory, we suggest that leader power is posi- tively related to follower met expectations. Therefore, the following hypothesis is posited:

Hypothesis 1: Leader power is positively related to follower met relationship expectations.

Leader Power Effects Moderated by Follower Power Although early research on power focused intently on the bases of power (French & Raven, 1959), positional and structural determinants of power (e.g., Pfeffer, 1981), and/ or personal characteristics that are influential in acquiring power (e.g., political will; Mintzberg, 1983), very little research has examined the interplay of both leader and fol- lower power simultaneously. Nonetheless, we believe that considerations of leader power and their effect on follow- ers’ met expectations are incomplete without considering the power standing of the follower.

In this support of this notion, research on mentoring sug- gests that a protégés’ met expectations of their relationship with their mentors have little to do with the amount of sup- port received from the mentor and more to do with the extent to which sufficient support is provided based on what the protégés expect and require (Young & Perrewé, 2000). Thus, we suggest that follower power acts to decouple the relationship between leader power and met relationship expectations. More specifically, we suggest that powerful followers expect and need of less support from their leaders than do nonpowerful followers, as they are capable of pro- curing some of their own resources, be they tangible or not.

Furthermore, although the resource dependence perspec- tive of power maintains that followers (i.e., subordinates)

Follower Leader Power

Follower Met

Expecta�ons

Follower Rela�onship

Quality

Leader Follower Power

Follower Job Tension

Figure 1. Model of leader–follower work relationships

144 Journal of Leadership & Organizational Studies 19(2)

are dependent on their leaders (i.e., supervisors) for tangible and intangible resources (Farmer & Aguinis, 2005), followers who are themselves powerful are less likely to be dependent on their leaders for the fulfillment of both physical and psy- chological resources. In other words, followers who possess their own power are able to, in a sense, break their depen- dence on their leader. Finally, a substantial body of research suggests that outcomes of work relationships (e.g., relation- ship quality, satisfaction) depend on the characteristics of both dyad members, not just one or the other (e.g., Kane, Martinez, Treadway, & Ferris, in press; Tsui, Xin, & Egan, 1995).

In mutually powerless dyads, relationship benefits and costs will be low because there will be few valued goods to exchange and little reason to make offers (Blau, 1964). When both persons are powerful, relationship benefits will be high as plenty of valued social rewards will be available for exchange (Blau, 1964). Also in mutually powerful dyads, costs will tend to be moderate because the other’s high demands may be tempered with one’s own power to resist them (Emerson, 1962).

Taken together, we suggest that the positive relationship between leader power and follower met relationship expec- tations is conditional on follower power. More formally, we hypothesize the following:

Hypothesis 2: Follower power will moderate the positive relationship between leader power and follower met expectations, such that for higher (lower) power followers, the positive relationship between leader power and follower met expecta- tions will be stronger (weaker).

Met Relationship Expectations and Work Relationship Quality Work relationship research in the organizational sciences is heavily influenced by leader–member exchange (LMX) theory (Ferris et al., 2009; Graen & Uhl-Bien, 1995). LMX theory postulates that dyadic relationship quality is based on mutual respect for each other’s capabilities, mutual trust, and reciprocal obligations. Capabilities are potential exchange goods that become actual exchange goods when they are respected and valued by others. Likewise, trust and obligations are social exchange goods themselves, much like credit. Hence, work relationship quality is determined by the exchange of social goods, such as valued capabili- ties, trust, and obligations.

Borrowing from the mentoring literature, research has found that mentors who engage in prototypical mentoring behavior inspire greater perceptions of relationship effec- tiveness as well as increased trust from their protégés through met relationship expectations (Young & Perrewé, 2000). Therefore, the following hypothesis is put forward:

Hypothesis 3: Follower met relationship expectations are positively related to follower work relationship quality perceptions.

Met Relationship Expectations and Job Tension Of both practical and theoretical import, researchers have frequently examined both antecedents and outcomes of job tension (Meurs & Perrewé, 2011). Job tension is defined as stress arising from work-related experiences (Kahn, Wolfe, Quinn, & Snoek, 1964) and is commonly measured with the job tension scale developed by House and Rizzo (1972). Not surprisingly, research suggests that as compared with low levels of tension, high levels of tension result in more dysfunctional organizational outcomes (e.g., job dissatis- faction, intent to turnover; Cropanzano, Howes, Grandey, & Toth, 1997; House & Rizzo, 1972). Research has also found that sources of job tension include a variety of role stressors, including work overload, role conflict, and role ambiguity (e.g., Frone, 1990; O’Driscoll & Beehr, 1994). Taken together, research provides general support for a “less is more” view of job tension.

Meta-analytic research on met expectations, although somewhat concentrated within the mentoring and realis- tic job preview literatures, suggests that met expectations is positively associated with job performance, job sur- vival, organizational commitment, and job satisfaction (Wanous, Poland, Premack, & Davis, 1992). Similarly, more recent research has found that any form of unmet expectations results in dysfunctional organizational out- comes (e.g., job dissatisfaction; Irving & Montes, 2009). As such, a lack of discrepancy between what one expects and what one experiences has consistently resulted in more positive (and less negative) attitudinal and behav- ioral workplace outcomes.

As an extension of the workplace outcomes examined in relation to met expectations, we hypothesize that followers’ met relationship expectations concerning their leader’s power should be negatively associated with job tension. We argue that those followers who perceive their leader to be in possession of power are in a sense reassured that their leader is capable of fulfilling the traditional role of leader. In addi- tion, such power in the hands of a leader was argued above to allow the leader to fulfill the follower’s resource depen- dencies. Consequently, the leader’s ability to fulfill both tangible and intangible needs of a follower should reduce the amount of stress a follower experiences arising from work experiences in general, and from their leader–follower dyadic relationship, in particular. Hence, the following hypothesis is proposed:

Hypothesis 4: Follower met relationship expectations are negatively related to follower job tension.

Martinez et al. 145

Method Participants and Procedure A total of 360 leader–follower dyads from a large south- eastern state government agency were invited to participate in the online survey study. Furthermore, 180 supervisors were selected to answer questions about two of their subor- dinates. The agency provided a list of all leaders/supervi- sors from selected departments (i.e., departments within a particular city) who supervised at least two followers/sub- ordinates, and all of those leaders/supervisors were invited to participate. A review of the job titles of participants indicated that most of the management staff consisted of frontline supervisors, and a much smaller set were the middle managers who supervised the frontline managers.

The employees being supervised were administrative, computer programming, or social service staff. Leaders had to supervise at least two employees to be considered in the study, and the span of control typically ranged from 2 to 12 direct reports. For each leader/supervisor, only two of his or her followers/subordinates were chosen at random in order to minimize nonindependence concerns. Both leaders and followers were asked to provide their names on the surveys in order to match their responses for subsequent analyses. Confidentiality was maintained by deleting respondent names after matching.

Leaders completed surveys for 218 of their followers, resulting in a 60.6% leader response rate, and 150 followers completed surveys, resulting in a 41.7% follower response rate. After combining leader and follower responses, there were a total of 100 useable dyads, resulting in a net response rate of 27.8%. Within dyads, 68% of the followers and 52% of the leaders were female. The average follower’s age was 49.5 years, and the average leader’s age was approximately the same (i.e., 50.2 years). Also, 71.0% of the followers and 62.7% of the leaders self-identified as White, 0.0% of the followers and 20.3% of the leaders self-identified as Hispanic, 23.0% of the followers and 4.2% of the leaders self-identified as Black, 4.0% of the followers and 0.9% of the leaders self-identified as Asian, and 2.0% of the follow- ers and 11.9% of the leaders self-identified as Other. The average job tenure was 5.9 years for followers and 5.4 years for leaders, and the average dyad tenure was 2.8 years.

Measures Follower power. Leaders answered four questions, loosely

adapted from Nesler and colleagues (Nesler, Aguinis, Quig- ley, Lee, & Tedeschi, 1999), to measure follower power. Two items were adapted from the “Resistance and Control Power” scale: “My subordinate can get what he/she wants from me” and “My subordinate can get me to do things I don’t want to do.” The other two items were adapted from Nesler et al.’s (1999) “Global Power” scale: “My subordinate can

influence me to evaluate his/her work performance favor- ably” and “My subordinate can influence me with regard to the types of projects I assign him/her.” A factor analysis using a principal component analysis extraction method was employed. A resulting one-factor model was selected based on retaining factors with eigenvalues greater than one. The one-factor model explained 62.5% of the variable variance. Cronbach’s α of .80 was obtained for the resulting four-item measure.

Leader power. Followers answered four questions, loosely adapted from Nesler et al. (1999), to measure leader power. Two items were adapted from the “Resistance and Control Power” scale: “My supervisor can get what he/she wants from me” and “My supervisor can get me to do things I don’t want to do.” The other two items were adapted from Nesler et al.’s (1999) “Global Power” scale: “My supervisor can influence me to work harder at my job” and “My super- visor can influence the type of projects I become involved in.” A factor analysis using a principal component analysis extraction method was employed. A resulting one-factor model was selected based on retaining factors with eigen- values greater than one. The one-factor model explained 52.4% of the variable variance. Cronbach’s α of .68 was obtained for the resulting four-item measure.

Power interaction term. The interaction term was calcu- lated via multiplying the follower power and leader power variables within a dyad. To address multicollinearity issues, the power variables were centered before multiplying, and the centered variables where used in subsequent analyses.

Follower met relationship expectations. Follower met rela- tionship expectations was operationalized with a three-item measure. Two items were modified from Young and Per- rewé’s (2000) Met Expectations Scale: “So far, I have received what I expected to receive from the relationship” and “In retrospect, I didn’t get what I expected from the relationship (reverse-scored).” A third item, not originally included in Young and Perrewé’s (2000) operationalization, was included in efforts to supplement the existing measure. This item read: “Overall, my expectations about my rela- tionship with my supervisor are being: (1) unmet, (2) par- tially met, (3) met, (4) somewhat exceeded, or (5) exceeded.”

We felt it was necessary to include this item as a supple- ment to the original two-item scale in case respondents felt expectations were exceeded, not just met. Furthermore, including an additional item in attitudinal measures that reflects an overall or composite assessment of the construct is not uncommon. A factor analysis using a principal com- ponent analysis extraction method was employed. A result- ing one-factor model was selected based on retaining factors with eigenvalues greater than one. The one-factor model explained 79.3% of the variable variance. Cronbach’s α of .85 was obtained for the three-item measure.

Follower-assessed work relationship quality. Follower- assessed work relationship quality was measured using the

146 Journal of Leadership & Organizational Studies 19(2)

popular seven-item LMX instrument (Graen & Uhl-Bien, 1995). Sample items included “Regardless of how much for- mal authority he/she has built into his/her position, what are the chances that your supervisor would use his/her power to help you solve problems in your work?” and “Again, regard- less of the amount of formal authority your supervisor has, what are the chances that he/she would ‘bail you out,’ at his/ her expense? A factor analysis using a principal component analysis extraction method was employed. A resulting one- factor model was selected based on retaining factors with eigenvalues greater than one. The one-factor model explained 63.4% of the variable variance. Cronbach’s α of .90 was obtained for the seven-item measure.

Follower job tension. The widely used seven-item Likert- type instrument developed by House and Rizzo (1972) was used to measure views of follower job tension. Sample items included “I have felt fidgety or nervous as a result of my job,” “I work under a great deal of tension,” and “My job tends to directly affect my health.” A factor analysis using a principal component analysis extraction method was employed. A resulting one-factor model was selected based on retaining factors with eigenvalues greater than one. The one-factor model explained 64.8% of the variable variance. Cronbach’s α of .91 was obtained for the seven- item measure.

Control variables. Followers were asked three questions regarding gender, job tenure, and dyad tenure, which were used as control variables in subsequent analyses. Gender is related to LMX and job tension. For example, Duchon, Green, and Taber (1986) found that gender predicted in- group/out-group status. Also, Pretty, McCarthy, and Catano (1992) concluded that men and women differ with regard to predictors and processes of burnout. Controlling for gender addresses concerns regarding it as an alternative explana- tion. Gender was coded with Male as 1 and Female as 2.

Tenure is an important consideration because, over time, employees may tend to self-select into and out of work rela- tionships that are compatible or incompatible with their val- ues (e.g., Schneider, Goldstein, & Smith, 1995). As tenure increases, it seems plausible that employees will tend to settle into relationships with relatively more favorable lev- els of relationship quality, met expectations, and job ten- sion. Hence, there may be more variation in relationship quality, met expectations, and job tension with lower ten- ure. Both follower job tenure and dyad tenure (i.e., tenure with leader) were controlled and measured with five rela- tively meaningful divisions for comparison (1 = less than 1 year; 2 = 1-3 years; 3 = 3-5 years; 4 = 5-10 years; and 5 = more than 10 years).

Data Analysis Model testing. The model presented in Figure 1 was tested

via path analyses using LISREL 8.71 software. Model

parameter estimates were derived from maximum likeli- hood estimation procedures. Both leader power and fol- lower power variables were centered to minimize multicollinearity issues caused by their interaction term. Hence, the hypothesized moderation effect was tested via an interaction term. The sample size (N = 100) was rather small, so model fit was evaluated using the standardized root mean square residual (SRMR) and comparative fit index (CFI) two-index strategy (Hu & Bentler, 1999).

Recommended thresholds for the SRMR and CFI two- index strategy are about .08 and .95, respectively (Hu & Bentler, 1999). Hence, SRMR should be less than .08 and CFI should be greater than .95. Model fit was also evaluated using root mean square error of approximation (RMSEA) because this is another fit index robust to small sample sizes (Fan, Thompson, & Wang, 1999). RMSEA values less than .06 typically indicate good model fit (Hu & Bentler, 1999). Finally, Tucker–Lewis index (TLI) also was used as it has been found to be relatively independent of sample size, and the TLI should be greater than .95 to indicate a good fit (Hu & Bentler, 1999; Marsh, Balla, & McDonald, 1988).

Results Descriptive Statistics and Correlations

Descriptive statistics and intercorrelations are provided in Table 1. Leader power correlated significantly with met relationship expectations (r = .40, p < .01). Follower power was not correlated with met expectations. The met relation- ship expectations variable correlated significantly with both work relationship quality and job tension, (r = .74, p < .01) and (r = −.39, p < .01), respectively. Finally, no control variable correlated significantly with work relationship quality or job tension.

Tests of Hypotheses A path analysis was conducted to test the model presented in Figure 1. Again, to minimize multicollinearity issues, the leader and follower power variables were mean-centered and then multiplied to derive the interaction term. The model fit indices suggested that the model reproduced the variance–covariance matrix fairly well. The χ2 was 11.84, with 10 degrees of freedom, and the p value was .30, imply- ing that there was no significant difference between the actual and reproduced matrices. The SRMR was .05, and the CFI was .98. Also, RMSEA was .05, and the TLI was .93. In total, the fit indices suggested adequate model fit.

This study provided strong support for Hypothesis 1 as the path between leader power and met relationship expec- tations (β = .39, t = 4.18) was moderately large, statistically significant, and in the hypothesized direction. The interac- tion hypothesis (i.e., Hypothesis 2) was not supported, as

Martinez et al. 147

the path between the interaction term and met relationship expectations (β = .08, t = 0.79) was small and statistically nonsignificant. The path between met relationship expecta- tions and work relationship quality (β = .75, t = 11.27) was very large, providing strong support for Hypothesis 3. Finally, the path between met expectations and job tension (β = −.38, t = 4.01) was moderately large and in the hypoth- esized direction. Hence, Hypothesis 4 also was supported.

Discussion Contributions of the Study

Recognizing the lack of research examining the power standings of both leader and follower and the implications of such on work outcomes, the purpose of this study was to examine the process by which leader power affects impor- tant attitudinal outcomes (i.e., work relationship quality and job tension) through met relationship expectations, as well to examine what impact follower power has on the relation- ship between leader power and met relationship expecta- tions. Overall, the results supported the general thesis that leader power affects work relationship quality and job ten- sion (i.e., employee well-being) through follower met rela- tionship expectations. Thus, the research suggested that the effect of leader power on met expectations has important implications for follower well-being.

It was argued that followers expect their leaders to be in possession of power, as this is not only characteristic of the role of a leader, but it is also characteristic of the traditional power differentials of hierarchical leader–follower dyads. Furthermore, past research on mentoring linked mentoring behavior to greater perceptions of relationship effectiveness through met expectations (Young & Perrewé, 2000) pro- vided the basis for the hypothesized relationship between met expectations and greater perceptions of work relation- ship quality. Finally, past research linking unmet expecta- tions to a number of negative attitudinal workplace outcomes (e.g., job dissatisfaction; Irving & Montes, 2009)

underpinned our assertion that met expectations would result in decreased job tension.

Results, however, did not lend support for our assertion that the leader power–met expectations relationship would be conditional on follower power, although it was argued that follower power acts to decouple the relationship between leader power and met expectations. More specifi- cally, we suggested that powerful followers are likely to expect and need less support from their leaders than do non- powerful followers, as powerful followers are capable of procuring necessary resources, be they tangible or not, by themselves. Results did not confirm this notion.

Strengths and Limitations A major strength of the study was the incorporation of other-reports for the focal independent variables. Specifically, followers assessed their leaders’ power, and leaders assessed their followers’ power. Other-reports helped minimize issues associated with common method biases (Podsakoff, Mackenzie, Lee, & Podsakoff, 2003). The predominant limitation of the study was the lack of statistical power. The rather small sample size of 100 dyads was insufficient to find paths with smaller effect sizes as statistically significant. Also, the small sample size limited the ability to use stronger data analyses methods, such as confirmatory factor analyses and structural equation mod- eling. Another limitation was that the study relied solely on reflective responses to questionnaires administered once.

A potential problem of common method variance is pres- ent in these data, which needs to be acknowledged. All vari- ables in the model tested (i.e., leader power, met expectations, work relationship quality, and job tension) except follower power (i.e., which was assessed by leaders) were measured from the follower’s perspective. That being said, inspection of the pattern of correlations in Table 1 does not appear to reflect a spuriously inflating mechanism that could be resulting from method bias, so it does not appear to be a serious problem in the present study.

Table 1. Means, Standard Deviations, and Intercorrelations

M SD 1 2 3 4 5 6 7

1. Follower Power 2.43 0.79 .80 2. Leader Power 4.03 0.56 .01 .68 3. Met Expectations 3.71 0.81 .12 .40* .85 4. LMX 3.86 0.73 .11 .38* .74* .90 5. Job Tension 2.63 0.92 .06 −.10 −.39* −.33* .91 6. Gender 1.68 0.47 .02 −.15 .17 .03 −.09 (n.a.) 7. Job Tenure 3.25 1.57 .03 −.07 .07 .14 −.13 .08 (n.a.)

8. Dyad Tenure 2.42 1.31 −.14 .07 −.01 .14 .01 .11 .41*

NOTE: LMX = leader–member exchange. N = 100. Cronbach α values are in boldface in the diagonal. *Correlation significant at .01 level (2-tailed).

148 Journal of Leadership & Organizational Studies 19(2)

Another limitation concerns the met relationship expec- tations construct. Although it was the only measure of its kind that we could identify, we must acknowledge that a three-item measure, with largely unknown psychometric properties, is far from ideal. Future research should focus on the development of a better measure of met relationship expectations, which reflects the rigorous development and construct validation reported by Hinkin (1995, 1998). Finally, there is potential dependency in the data that needs to be acknowledged as a potential concern. Because180 supervisors were selected to answer questions about two of their subordinates, there is a possibility that leaders did not evaluate followers power independently but rather reflected cross-person rating bias. However, as we noted above in the description of the sample and procedures: “For each leader/ supervisor, only two of his or her followers/subordinates were chosen at random in order to minimize non-indepen- dence concerns.” That being said, we admit this could raise some concern (see Gooty & Yammarino, 2011, for a recent review of such issues with particular reference to dyads).

Directions for Future Research The power variables in the leader–follower dyad are prob- ably related to several other important outcome variables. For example, how does power relate to performance vari- ables, such as in-role performance, organizational citizen- ship behaviors directed at individuals, and organizational citizenship behaviors directed at the organization (e.g., Williams & Anderson, 1991)? This study can be extended to other domains (e.g., behaviors, attitudes, and feelings), and met expectations might be considered an instrumental mediator. Of course, other mediators and moderators can be investigated as well. Future research might also consider testing a potential “negotiation, acquiescence, and capitula- tion hypothesis,” that is, specifically, that the conditions of negotiation and acquiescence by followers in interactions with leaders might be more beneficial than capitulation to leader demands. This notion starts to get at more specific focus on the precise nature of the work relationships that are developed and endure between leaders and followers as a function of their power.

As such, future research might investigate particular work relationship dimensions. Recent research on the nature of dyadic relationships has argued for a focus on the underlying dimensions of work relationships, such as trust, respect, time, distance, and affect (Ferris et al., 2009). Particularly relevant to the present study is the relationship dimension of distance, which Ferris et al. (2009) referred to as

the quality of closeness or separation in space and time, and it can be reflected in the two concepts of physical distance and psychological distance. In work relationships, physical distance might be reflected in how closely two people work in terms of physical

location. Psychological distance refers to the close- ness or separation of the perceptions, attitudes, and feelings of two people, usually measured through the use of a Euclidean distance measure, or a profile similarity or matching process. (pp. 1391-1392)

Napier and Ferris (1993) argued that “functional dis- tance” is most important for work relationship effectiveness and that there is not necessarily a linear relationship between distance and effective work relationships. Instead, they characterized functional distance as the optimal degree of psychological distance that provides leaders with perspec- tive and allows them to be more objective in their evalua- tion and treatment of followers. Antonakis and Atwater (2002) discussed leader distance, referring to it as perceived social distance, physical distance, and perceived interaction frequency of leaders with their followers. Leader distance is important because it reflects the perspective from which leaders perceive, interpret, and process information and eval- uate member outcomes, and also it appears that it could reso- nate with leader and follower power. It might be interesting to extend the present results to investigate the “distance” dimension of work relationships and see to what extent leader and follower power are associated with an optimal level of functional distance in the work relationship.

Power is an important aspect of the context of the leader– follower relationship. In future studies, other contextual factors such as the nature of work, leader–follower relation- ship expectations, and individual power relation prefer- ences could be more explicitly considered within the underlying social context of organizations. For example, future studies could investigate variables such as the organi- zation’s decision-making framework that formally defines the power relationship among supervisors and subordinates, the organizational and subunit cultural norms, and the power relation norms that may serve as important contex- tual factors potentially affecting leader and follower power and their work relationships.

Finally, future research might consider the investigation of follower-assessed work relationship quality and some of the factors that might affect such relationship perceptions. For example, scholars might measure follower attractive- ness to the leader and the degree to which personal qualities influence the power/dependence relationship (e.g., percep- tions about leader ambition, energy, focus, etc., may play a role in the leader–follower work relationship). These repre- sent just a few ideas for future research on the impact of leader and follower power, which is an area of inquiry of which we have only begun to scratch the surface of fully comprehending.

Practical Implications The social context in which power dynamics occurs is an important factor that frames organizational decisions,

Martinez et al. 149

actions, and behaviors of leaders and followers in the course of their daily interactions. The shift in the economy to a more knowledge-based workforce raises the impor- tance for practitioners to develop an informed understand- ing of the dynamic power relationship between followers and leaders and its subsequent impact on employee atti- tudes, behavior, and organizational performance.

When considering the impact of context on the exchange of social power, practitioners also must recognize that orga- nizational structures, routines, and culture influence the context of how such power dynamics occur. Specifically, practitioners must evaluate the influence of their organiza- tion on the nature of how work is done and decisions are made. For example, in the current study, differences in leader and follower power were perceived positively by fol- lowers. This may be the result of regulatory policy or legis- lative mandate that creates a culture that encourages such differential levels. As a result, followers may understand how the system of work and adopt a strategy that allows them to work effectively within the system.

Consequently, organizations might employ power dynamics in strategic interventions that support follower performance and optimize organizational productivity, all within the context of their operational environment. In more structured contexts, potential interventions might include leader training in employee engagement activities, resulting in followers being able to more fully use their knowledge, skills, and abilities within a more hierarchically based orga- nization. Conversely, in a less traditionally structured envi- ronment, educating leaders on ways to minimize power differences with followers would be preferable. Such efforts might consider increased efforts at the implementation of shared leadership or participative management processes.

Conclusion Results indicated that leader power affected both leader– follower relationship quality and job tension through fol- lowers’ met relationship expectations. However, contrary to our hypothesis, the leader power–met expectations relation- ship was not conditional on follower power. In other words, leader power appears to be a central factor in leader–fol- lower work relationships because it shapes follower expec- tations, and whether or not these relationship expectations are met affects important follower well-being outcomes. Nonetheless, as organizational begin to flatten, future research should not ignore the possibility that follower power could alter the relationship between leader power and work outcomes.

Declaration of Conflicting Interests

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

Funding The authors received no financial support for the research, author- ship, and/or publication of this article.

References

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

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

Brass, D. J., & Burkhardt, M. E. (1993). Potential power and power use: An investigation of structure and behavior. Acad- emy of Management Journal, 36, 441-470.

Cropanzano, R., Howes, J., Grandey, A., & Toth, P. (1997). The relationship of organizational politics and support to work behaviors, attitudes, and stress. Journal of Organizational Behavior, 18, 159-180.

Duchon, D., Green, S. G., & Taber, T. D. (1986). Vertical dyad linkage: A longitudinal assessment of antecedents, measures, and consequences. Journal of Applied Psychology, 71, 56-60.

Emerson, R. M. (1962). Power-dependence relations. American Sociological Review, 17, 31-41.

Emerson, R. M. (1964). Power-dependence relations: Two experi- ments. Sociometry, 27, 282-298.

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

Farmer, S. M., & Aguinis, H. (2005). Accounting for subordinate perceptions of supervisor power: An identity-dependence model. Journal of Applied Psychology, 90, 1069-1083.

Ferris, G. R., Liden, R. C., Munyon, T. P., Summers, J. K., Basik, K., & Buckley, M. R. (2009). Relationships at work: Toward a multidimensional conceptualization of dyadic work relation- ships. Journal of Management, 35, 1379-1403.

French, J. R. P., & Raven, B. (1959). The bases of social power. In D. Cartwright & A. Zander (Eds.), Group dynamics (pp. 150- 167). New York, NY: Harper & Row.

Frone, M. (1990). Intolerance of ambiguity as a moderator of the occupational role stress-strain relationship: A meta-analysis. Journal of Organizational Behavior, 11, 309-320.

Gooty, J., & Yammarino, F. J. (2011). Dyads in organizational research: Conceptual issues and multilevel analyses. Organi- zational Research Methods, 14, 456-483.

Graen, G. B. (2009). Growing powerful using cherry-picking strategies: Coworker networks as cherry trees. In D. Tjosvold & B. Wisse (Eds.), Power and interdependence in organiza- tions (pp. 103-117). Cambridge, England: Cambridge Univer- sity Press.

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

150 Journal of Leadership & Organizational Studies 19(2)

Graen, G. B., & Uhl-Bien, M. (1995). Relationship-based approach to leadership: Development of LMX theory of leadership over 25 years. Leadership Quarterly, 6, 219-247.

Hinkin, T. R. (1995). A review of scale development practices in the study of organizations. Journal of Management, 21, 967-988.

Hinkin, T. R. (1998). A brief tutorial on the development of measures for use in survey questionnaires. Organizational Research Methods, 1, 104-121.

House, R. J., & Rizzo, J. R. (1972). Role conflict and ambigu- ity as critical variables in a model of organizational behav- ior. Organizational Behavior and Human Performance, 7, 467-505.

Hu, J., & Bentler, P. M. (1999). Cutoff criteria for fit indices in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1-55.

Irving, P. G., & Montes, S. D. (2009). Met expectations: The effects of expected and delivered inducements on employee satisfaction. Journal of Occupational and Organizational Psy- chology, 82, 431-451.

Kahn, R. L., Wolfe, D. M., Quinn, R. P., & Snoek, J. D. (1964). Organizational stress: Studies in role conflict and ambiguity. New York, NY: John Wiley.

Kane, R. E., Martinez, A. D., Treadway, D. C., & Ferris, G. R. (in press). Supervisor power x subordinate political skill interac- tion on work relationship quality: Alternative interaction forms as a function of dyadic influence dynamics. In D. Svyantek & K. Mahoney (Eds.), Received wisdom, kernels of truth, and boundary conditions in organizational studies. Charlotte, NC: Information Age.

Katz, D., & Kahn, R. L. (1978). The social psychology of organi- zations (2nd ed.). New York, NY: Wiley.

Marsh, H. W., Balla, J. R., & McDonald, R. P. (1988). Goodness- of-fit indexes in confirmatory factor analysis: The effect of sample size. Psychological Bulletin, 103, 391-410.

Mechanic, D. (1962). Sources of power of lower participants in complex organizations. Administrative Science Quarterly, 7, 349-364.

Merton, R. K. (1957). Social theory and social structure. New York, NY: Free Press.

Meurs, J. A., & Perrewé, P. L. (2011). Cognitive activation theory of stress: An integrative theoretical approach to work stress. Journal of Management, 37, 1043-1068.

Mintzberg, H. (1983). Power in and around organizations. Englewood, Cliffs, NJ: Prentice-Hall.

Napier, B. J., & Ferris, G. R. (1993). Distance in organizations. Human Resource Management Review, 3, 321-357.

Nesler, M. S., Aguinis, H., Quigley, B. M., Lee, S. J., & Tedeschi, J. T. (1999). The development and validation of a scale mea- suring global social power based on French and Raven’s power taxonomy. Journal of Applied Social Psychology, 29, 750-771.

O’Driscoll, M. P., & Beehr, T. A. (1994). Supervisor behaviors, role stressors, and uncertainty as predictors of personal out- comes for subordinates. Journal of Organizational Behavior, 15, 141-155.

Pearce, C. L., & Conger, J. A. (2003). Shared leadership: Reframing the hows and whys of leadership. Thousand Oaks, CA: SAGE.

Pfeffer, J. (1981). Power in organizations. Boston, MA: Pitman. Pfeffer, J. (1992). Managing with power. Cambridge, MA: Harvard

University Press. Pfeffer, J. (2010). Power: Why some people have it and others

don’t. New York, NY: HarperCollins. Pfeffer, J., & Salancik, G. R. (1978). The external control of orga-

nizations: A resource dependence perspective. New York, NY: Harper & Row.

Podsakoff, P. M., Mackenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 879-903.

Pretty, G. M., McCarthy, M. E., & Catano, V. M. (1992). Psy- chological environments and burnout: Gender considerations within the corporation. Journal of Organizational Behavior, 13, 701-711.

Ragins, B. R., & Dutton, J. E. (2007). Positive relationships at work: An introduction and invitation. In J. E. Dutton & B. R. Ragins (Eds.), Exploring positive relationships at work: Build- ing a theoretical and research foundation (pp. 1-25). Mahwah, NJ: Lawrence Erlbaum.

Rousseau, D. M., & Schalk, R. (2000). Psychological contracts in employment: Cross-national perspectives. Thousand Oaks, CA: SAGE.

Schneider, B., Goldstein H. W., Smith D. B. (1995). The ASA framework: An update. Personnel Psychology, 48, 747-773.

Thibaut, J. W., & Kelley, H. H. (1959). The social psychology of groups. New York, NY: John Wiley.

Tsui, A. S. (1984). A role set analysis of managerial reputation. Organizational Behavior and Human Decision Processes, 34, 64-96.

Tsui, A. S., Xin, K. R., & Egan, T. D. (1995). Relational demogra- phy: The missing link in vertical dyad linkage. In S. E. Jackson & M. N. Ruderman (Eds.), Diversity in work teams: Research paradigms for a changing workforce (pp. 97-129). Washington, DC: American Psychological Association.

Wanous, J. P., Poland, T. D., Premack, S. L., & Davis, K. S. (1992). The effect of met expectations on newcomer attitudes and behaviors: A review and meta-analysis. Journal of Applied Psychology, 7, 288-297.

Weber, M. (1947). The theory of social and economic organiza- tion. Oxford, England: Oxford University Press.

Williams, L. J., & Anderson, S. E. (1991). Job satisfaction and organizational commitment as predictors of organizational citizenship and in-role behaviors. Journal of Management, 17, 601-617.

Young, A. M., & Perrewé, P. L. (2000). What did you expect? An examination of career-related support and social support among mentors and protégés. Journal of Management, 26, 611-632.

Bios

Arthur D. Martinez is Assistant Professor of Management and Quantitative Methods at Illinois State University. He received his

Martinez et al. 151

Ph.D. in Management from Florida State University. His research investigates social power in organizational contexts.

Rachel E. Kane is a Ph.D. candidate in Organizational Behavior and Human Resource Management at Florida State University. Her current research interests include antecedents and boundary conditions of political skill, organizational politics perceptions, and leadership. She is also interested in advanced data analysis methods such as structural equations modeling.

C. Darren Brooks is the Human Resource Shared Services Director at the Florida Department of Children and Families; in addition, he is an Adjunct Professor of Management and Visiting

Instructor of Educational Psychology and Learning Systems at Florida State University. He received his Ph.D. in Instructional Systems from Florida State University. His research investigates social influences and learning factors that affect performance in organizational contexts, adaptive learning systems, and human performance technology.

Gerald R. Ferris is the Francis Eppes Professor of Management and Professor of Psychology at Florida State University. He received a Ph.D. in Business Administration from the University of Illinois at Urbana-Champaign. Ferris has research interests in the areas of social influence and effectiveness processes in orga- nizations, and the role of reputation in organizations.