Politics and Justice: A Mediated Moderation Model

By Miller, Brian K Nicols, Kay McGlashan

Organizational politics are complex phenomena, particularly because their existence is interpreted through the perceptions of individual organizational members. Each individual’s perception of organizational politics (POP) is a function of his or her own particular characteristics, the social relationships that they have developed within the organization, and certain outcomes and consequences such as reward allocations and job attitudes. Ferris, Russ, and Fandt (1989) developed a conceptual framework and working definition of organizational politics that identified antecedents, outcomes, and moderators of POP. They define politics as a “social influence process in which behavior is strategically designed to maximize short-term or long-term self-interest, which is either consistent with or at the expense of others’ interests” (1989: 145). Other researchers have subsequently refined and extended this model (e.g., Ferris et al., 2002; Kacmar et al., 1999; Valle and Perrewe, 2000). These researchers suggest that certain situations in organizations (e.g., those that are ambiguous, uncertain, or subjectively determined) are conducive to political perceptions. Some researchers (e.g., Ferris et al, 1996; Valle and Perrewe, 2000; Vigoda, 2000) have examined the pivotal role of POP as a mediator of the relationship between particular antecedents and outcomes, while others have examined political perceptions as a moderator of workplace phenomena (e.g., Harris et al., 2005; Hochwarter et al., 2000). Antecedents found to be significantly related to POP include hierarchical level (Ferris et al., 1999), feedback (Kacmar et al., 1999), interaction with supervisors (Valle and Perrewe, 2000), and career development opportunities (Kacmar et al., 1999; Parker et al., 1995). Outcomes significantly related to POP include job stress (Kacmar et al., 1999), job involvement (Ferris and Kacmar, 1992), job satisfaction (Ferris et al., 1996), and pay satisfaction (Zhou and Ferris, 1995). Previous research has treated antecedents as independent of one another in empirical testing of the POP model. We believe that it is important to begin to explore the highly interactive nature of antecedent variables commonly included in models of POP, and the transmitted effects of such interactions on outcomes associated with organizational politics. Thus, the purpose of this study is to examine the interactive roles of one type of individual difference, locus of control, and an aspect of one’s relationship with their supervisor, leader-member exchange quality, in the prediction of perceptions of distributive justice, as mediated by POP.

In the following sections, we introduce the theoretical framework associated with our mediated moderation model. First, we introduce our two antecedents, work locus of control and leader-member exchange. Next, we discuss the concept of POP. We then discuss our outcome variable, distributive justice. Lastly, we develop hypotheses in order to test the relationships in our model (see Figure I).

A COGNITIVE SOCIAL LEARNING MODEL OF POLITICAL PERCEPTIONS

According to cognitive social learning (CSL) theory, perception is cognitively constructed and the nature of the relationship between job perceptions and job attitudes is purely one of reciprocity (James and Jones, 1980; James and Tetrick, 1986). An individual makes inferences about events in a social setting based on stimuli that are captured and interpreted individually. Such inferences are then impacted by an individual’s unique characteristics (e.g., personality traits), pre-existing expectations and schemata, attributions, and relationships with others, as well as other influences. Thus, employee perceptions about organizational settings are affected by individual interpretations of a situation, rather than being a function of an objective description of the relevant environment (Fiske and Taylor, 1991). The variables of interest in this study interact within individuals’ socially constructed realities. Work locus of control is one such characteristic that impacts an individual’s interpretation of events. Individuals’ perceptions of organizational politics, judgments about the quality of their relationship with their supervisor, and perceptions regarding distributive justice also result directly from social interaction with others in an organizational setting. This study contributes to our understanding of the social construction of reality by examining the interplay among these variables.

Work Locus of Control

Locus of control grew out of social learning theory (Rotter, 1954), which suggests that reinforcement will strengthen the expectancy that a particular behavior will be followed by that reinforcement in the future. Rotter states that “people . . . have developed generalized expectancies in learning situations in regard to whether or not reinforcement, reward, or success in these situations is dependent upon their own behavior or is controlled by external forces, particularly luck (or) chance” (1966: 25). Thus, individual differences account for much of how we interpret events and the effect of those events upon ourselves. Spector (1982) suggests that locus of control is related to multiple other individual differences such as motivation, effort, satisfaction, perception of the job, and supervisory style. He states that internals tend to believe that they control the work environment, while externals maintain that the environment controls them. Cognitive social learning theory suggests that it is the social milieu that provides context for the interpretation of events and from those events individuals learn and adjust their behavior. Because individual differences like locus of control play such an important role in the interpretation of organizational events, it is likely that persons with an internal locus of control will be more cognizant of potentially disruptive external issues imbedded in politics and justice.

Leader-Member Exchange

Unlike theories of leadership that focus on leader traits and behaviors (Bass, 1990; Mintzberg, 1973), leadermember exchange (LMX) theory suggests that the effectiveness of a supervisor is largely determined by the perception of die quality of die dyadic relationship between leaders/supervisors and members/subordinates (Dansereau et al, 1973; Dansereau et al, 1975; Graen, 1976; Graen and Cashman, 1975). Leader-member exchange suggests that leaders treat members differently and, accordingly, an in-group and an out- group of members are formed. In-group membership is preferred by members, as such affiliation is often associated with preferential treatment and characterized by favorable attitudinal correlates for in-group members. Leader-member exchange is strongly associated with subjective performance ratings from leaders and with member affective states (Gerstner and Day, 1997). Additionally, training interventions where leaders are trained to foster highquality relationships with members have increased member performance ratings and member satisfaction (Graen et al., 1982).

It is the reciprocal nature of leadership in general, and LMX more specifically because of its dyadic premise, that lends itself well to interpretation in a CSL framework. Cognitive social learning theory suggests that individuals learn from their supervisor as part of the organizational milieu in which they find themselves. The high quality relationship of ingroup members with their supervisor reinforces their behavior as the scarce organizational resources which are garnered for these favored subordinates leads to reinterpretation of what it means to be so favored-i.e. lower levels of dysfunctional politics and higher levels of perceived justice.

Perceptions of Organizational Politics

Kacmar and Baron (1999) attribute the beginning of empirical research on micro-level organizational politics to Gandz and Murray (1980). For the remainder of that decade, research waned as struggles to define and operationalize organizational politics ensued. Over time, the conceptualization of organizational politics has been refined as researchers consistently identify organizational politics as something that is usually perceived of negatively, even though some have attempted to reconceptualize the construct as something positive (Ferris et al., 1995).

In the conceptualization of perceptions of organizational politics, it is important to distinguish it more specifically from other politics constructs. Political perceptions are different from political tactics (an increasingly common construct of interest to researchers), which refer to specific behaviors by specific individuals (Zanzi and O’Neill, 2001), rather than the perception of political policies, rules, procedures, and outcomes like those conceptualized as POP. Ferris et al. (2002) differentiate among the constructs of organizational politics, political behavior, and politics perceptions. They state that “political behavior deals with influence attempts that occur at the individual and group level, while organizational politics examines the extent to which such behaviors are pervasive in the work, decision-making, and resource allocation processes within the organization” (2002: 183). Rather than being an objective reality, organizational politics is subjectively perceived by individuals in a workplace (Ferris et al, 1989; Ferris et al, 2002). Much of the research (e.g., Ferris and Kacmar, 1992; Ferris et al., 1996; Kacmar et al, 1999; Valle and Perrewe, 2000) has therefore focused on perceptions of organizational politics, in response to Gandz and Murray (1980) who suggested that the most accurate way to study organizational politics is by measuring individuals’ perceptions about it.

Distributive Justice

We utilized distributive justice perceptions as the outcome in our model. Previous research has demonstrated a significant relationship between distributive justice and organizational politics perceptions (Andrews and Kacmar, 2001; Parker et al, 1995). Organizational justice describes the perception of the fairness of distribution, processes, and personal interactions in the workplace (Colquitt, 2001; Greenberg, 1987). Distributive justice is concerned with the fairness of allocation of rewards and is often seen as the antithesis of organizational politics (Andrews and Kacmar, 2001). The two are inversely related because as individuals perceive that resources are allocated unfairly or without regard for need or merit, they are more likely to perceive that organizational politics is the culprit. Andrews and Kacmar (2001) found evidence of discriminant validity in their confirmatory factor analysis of justice, politics, and perceived organizational support. They suggest that politics and justice are likely to be different constructs, but that they are also moderately inversely related.

Dulebohn (1997) suggests that individuals are motivated to use influence tactics in order to achieve particular goals or objectives. Individuals see political behaviors as instrumental in obtaining outcomes of interest to them. Individuals tend to evaluate processes and outcomes separately (Greenberg, 1990; Tyler, 1994). They attempt to realize more favorable outcomes by influencing processes. A favorable outcome is one that is perceived as beneficial to the recipient’s self-interest (Dulebohn, 1997). Political behavior is specifically defined as a social influence process designed to maximize self-interest (Ferris et al., 1989). As such, individuals’ perceptions of the level of organizational politics at play within their organization pertain to the extent to which others are attempting to “work the system” to their own benefit (which may or may not negatively impact the perceiver’s own outcomes). Some may argue that procedural justice is a more appropriate outcome variable to examine in a model of perceptions of organizational politics, because political behaviors are designed to affect organizational processes in order to influence outcomes. However, we argue that perceptions of distributive justice are a more concrete representation of the success of political behaviors. Political behaviors are specifically used to further individuals’ self-interested outcomes. Individuals’ perceptions of distributive justice answers the “so what” question regarding perceptions of organizational politics-if an individual perceives that organizational politics is operating within the work environment, he/ she will most likely be concerned with how this impacts his/her own self-interested, desired outcomes. Thus, we propose that individuals perceiving a great deal of political behavior within their organizations would perceive that such behaviors affect the fairness of reward outcomes distributed to them.

The Mediating Role of Perceptions of Organizational Politics

We suggest that the effects of leader-member exchange and work locus of control (and their interaction) are transmitted to distributive justice via POP. That is, political perceptions serve to explain haw the antecedents affect the outcome. Thus, to test our model, we examine the mediating role of POP in the relationship between the interaction of work locus of control and leader-member exchange in the prediction of distributive justice. The Baron and Kenny (1986) framework for examining mediation requires that certain preliminary relationships be found (namely that the antecedents are statistically related to both the mediator and the outcome and that the mediator is statistically related to the outcome). Each of these relationships is examined below, and their existence is required before an examination of the mediating influence of POP in these relationships can be undertaken.

Influence of Work Locus of Control on Distributive Justice. Sweeney et al. (1991) found a strong direct linkage between an internal locus of control and perceptions of justice in the workplace. Andrews and Kacmar (2001) found that locus of control was significantly associated with perceptions of distributive justice such that individuals with an external locus of control reflected lower distributive justice ratings than individuals with a more internal locus of control. Persons with an internal locus of control tend to look inward for sources of both their successes and failures. As such, we suggest that they are more likely to be believers in the just world hypothesis (Lerner, 1980) that people get what they deserve in life. They are likely to believe that life in general, and their organization specifically, are both fair. Locus of control is operationalized in this study such that higher values represent more internal locus of control, while lower values represent more external locus of control. Thus,

Hypothesis 1: Work locus of control and distributive justice are positively related.

Influence of Leader-Member Exchange on Distributive Justice. Scandura (1999) suggests that justice perceptions are a result of economic exchanges (i.e., pay increases, promotions) being considered fair or unfair. However, these economic exchanges take place within a larger social context. The leader-member exchange process is a social exchange that affects individual perceptions of the fairness of a particular economic exchange. Vecchio et al (1986) and Manogran et al (1994) found significant positive relationships between distributive justice and leader-member exchange, such that individuals with higher quality relationships with their immediate supervisors perceived higher distributive justice than those individuals with lower quality relationships with their leaders. Andrews and Kacmar (2001) found a non-significant relationship between leader-member exchange and distributive justice. However, Wayne et al. (2002) found a significant positive relationship between LMX and distributive justice.

Out-group members in particular may focus more on distributive justice than in-group members, because the nature of the relationship between the leader and members of the outgroup tend to be more formal and contractual in nature (Liden et al, 1997) and “both parties are more concerned with the flow of tangible resources than with the quality of the relationship” (Davis and Gardner, 2004: 446). Therefore, members with low quality relationships with their leader are likely to have heightened sensitivity to an inequitable distribution of rewards and are likely to perceive low levels of distributive justice. Leader-member exchange quality is measured in this study such that higher values represent higher quality exchange relationships, Thus,

Hypothesis 2: leader-member exchange and distributive justice are positively related.

Interaction of Work Locus of Control and Leader-Member Exchange on Distributive Justice. Kinicki and Vecchio (1994) found a positive relationship between individuals’ internal locus of control and quality of relations developed with their immediate supervisor. Martin et al. (2005) also found a significant positive relationship such that individuals with an internal locus of control perceived higher quality relationships with their supervisors. Thus, locus of control and LMX relationships appear to be significantly interrelated. Our previous discussion of these variables reflects their individual positive relationships with distributive justice. Due to the strong positive association between locus of control and leader-member exchange as well as their individual positive effects on distributive justice demonstrated in previous research, we propose that the two variables will interact in our model, and that the interaction of the two is likely to further positively influence individuals’ interpretation of the fairness of organizational outcomes.

Hypothesis 3: There is an interaction effect between leader- member exchange and work locus of control on distributive justice, such that the highest levels of distributive justice will arise from high levels of leader-member exchange and work locus of control.

Influence of Work Locus of Control on Perceptions of Organizational Politics. Valle and Perrewe (2000) suggest that because individuals with an external locus of control are more likely to perceive that they have little or no control on the events that impact them, they are more likely to perceive high levels of organizational politics. Because politics are frequently construed of as phenomena external to one’s self, CSL theory suggests that the social construction of reality depends at least partly upon one’s attributions of causality and one’s locus of control. That is, as an individual attempts to make sense of his/her organizational environment he/she seeks to attribute causality to these encounters. Attribution theory (Weiner, 1980) and CSL theory suggests that while most observers of these phenomena agree on perceived causality, “in contrast, . . . locus of control posits that stable individual differences among perceivers influence causal inference” (Fiske and Taylor, 1991: 42). With this in mind, we suggest that persons with an internal locus of control would view the organizational political process as something for them to more effectively manage, and that some insufficiency of their own efforts (as opposed to organizational constraints) might have allowed for a greater existence of organizational politics. Hypothesis 4: Work locus of control and perceptions of organizational politics are negatively related.

Influence of Leader-Member Exchange on Perceptions of Organizational Politics. A low quality relationship between member and leader is also likely to result in the member’s perception of higher levels of organizational politics. Because the in-group status defined by LMX suggests that favoritism runs concurrently with such status, members of the outgroup are likely to perceive that politics, as evidenced by the inequitable distribution of rewards without regard for merit, is responsible for their undesirable status. Davis and Gardner (2004) suggest that the quality of the leader-member exchange influences the tone of subordinate reactions to, and interpretations of, perceived political actions. Ferris and Kacmar (1992) and Valle and Perrewe (2000) found direct evidence of a significant inverse relationship between quality of LMX relationship and POP. Individuals with low- quality relationships with their superior may tend to perceive leader favoritism toward in-group members that is based on political factors rather than on objective performance factors (Davis and Gardner, 2004; Kacmar and Ferns, 1991). On the other hand, Andrews and Kacmar (2001) found a positive relationship between leadermember exchange and organizational politics. They posited that ingroup members with higher quality relationships with supervisors may actually be more “in-the-know” about political machinations in their organizations. However, at least three other studies (Ferris and Kacmar, 1992; Kacmar et al, 1999; Valle and Perrewe, 2000) have all found negative correlations between interactionwith-supervisors and POP. Thus, it appears likely that a high quality relationship with one’s supervisor will decrease one’s perceptions of organizational politics.

Hypothesis 5: Leader-member exchange and perceptions of organizational politics are negatively related.

Interaction of Work Locus of Control and Leader-Member Exchange on Perceptions of Organizational Politics. As previously discussed, prior research has demonstrated a positive relationship between locus of control and LMX, such that individuals with more internal locus of control perceive higher quality relationships with their supervisors. Individuals perceive higher levels of organizational politics when they have an external locus of control, and when they perceive low LMX quality. Therefore, we posit that individuals with an internal locus of control combined with perceptions of high quality relationships with their supervisors will likely perceive the lowest levels of organizational politics within their organizations.

Hypothesis 6: There is an interaction effect between leader- member exchange and work locus of control on perceptions of organizational politics, such that the highest levels of perceptions of organizational politics will arise from low levels of leader- member exchange and locus of control.

Mediating Role of Perceptions of Organizational Politics. Perceptions of organizational politics is the central construct in our model. To test our model, we use mediated moderation regression analysis to examine these relationships. We borrow the language of Baron and Kenny (1986) when we suggest that politics is a generative mechanism through which one’s relationship with their supervisor and their perception of their ability to influence events around them is transmitted to, and predicts, their perception of the fairness of organizational rewards.

Hypothesis 7: Perceptions of organizational politics mediates the relationship between work locus of control and distributive justice.

Hypothesis 8: Perceptions of organizational politics mediates the relationship between leader-member exchange and distributive justice.

Hypothesis 9: Perceptions of organizational politics mediates the relationship between the interaction of leader-member exchange and work locus of control with distributive justice.

METHOD

Participants

The participants in this study were full-time employees enrolled in an MBA program at a large state university in the southern United States. Demographic information was gathered via a categorical response sheet. Seventy-eight percent of participants were managers, 67% were male, and the racial breakdown was as follows: 78.0% Caucasian, 10.2% African American, 3.9% Hispanic, 2.4% Asian, and 5.5% Other. Fifty-eight percent were over 35 years of age. Regarding current job tenure, 35% had less than one year, 32% reported between one and five years, 6% had six to 10 years of job tenure, 12% reported 11 to 15 years, and 15% had more than 15 years of current job tenure.

Procedure

Participants were solicited for their voluntary participation in a survey of their organizational experiences, attitudes, and perceptions. One hundred eighty-three surveys were distributed to the participants during class time. They completed the survey on their own time and returned them at the next class meeting. One hundred twenty-five usable surveys were returned. They did not record their name or any identifying information on the survey, and the data were therefore completely anonymous. Extra credit was awarded for participation.

Measures

Perceptions of Organizational Politics. A fifteen-item scale measuring perceptions of organizational politics by Kacmar and Carlson (1997) was used. The scale uses a five-point Likert response anchored by 1 = strongly disagree and 5 = strongly agree. High scores indicate that respondents experience high levels of ambiguity and uncertainty in the workplace and are more likely to encounter favoritism and highly self-interested behavior. Six items were reverse scored. Examples of items include: “People in this organization tend to build themselves up by tearing others down,””Favoritism rather than merit determines who gets ahead around here,” and “Promotions in this department generally go to top performers (reverse scored).” The Cronbach’s alpha for scores on this scale was .89.

Distributive Justice. The six-item Distributive Justice Index by Price and Mueller (1986) was used to assess perceptions of distributive justice. High scores indicate that respondents perceive that they are fairly rewarded by their employer. The scale items use a five-point Likert response anchored by 1 = very unfairly and 5 = very fairly. Examples of items include: “To what extent are you fairly rewarded considering the responsibilities that you have?” and “To what extent are you fairly rewarded for work that you have done well?” The Cronbach’s alpha for scores on this scale was .94.

Work Locus of Control. Spector’s (1988) 16-item Work Locus of Control Scale measured locus of control. Half of the items indicate an internal locus of control, while half indicate an external locus of control and were reverse scored. High scores on the scale therefore indicate an internal work locus of control. A high internal work locus of control suggests that respondents believe that they influence their environment more than their environment influences them and that they have the ability to manipulate events that occur around them. The scale uses a five-point Likert response anchored by 1 = strongly disagree and 5 = strongly agree. Examples of items include “A job is what you make of it,””Making money is primarily a matter of good fortune (reverse scored),” and “Most people are capable of doing their jobs well if they make the effort.” Cronbach’s alpha for scores on this scale was .78.

Leader-Member Exchange. A sevenitem scale by Graen et al. (1982) was administered from the perspective of employees regarding the quality of their relationship with their supervisors. Thus, the subordinate (i.e., member) perspective on the dyadic relationship was the focus in this study. High scores indicate that employees (i.e., members) favorably perceive of their relationship with their supervisor (leader). The scale uses a five-point Likert response anchored by 1 = strongly disagree and 5 = strongly agree. Examples of items include: “I can count on my supervisor to ‘bail me out’ at his or her expense when I really need it,””My immediate supervisor understands my problems and needs,” and “I usually know where I stand with my immediate supervisor.” Cronbach’s alpha for scores on this scale was .93.

Analysis

Multiple regression analysis was used to test our hypotheses. We used the Baron and Kenny (1986) framework for testing for mediation and the suggestions of Muller et al. (2005) to examine mediated moderation. Referring again to our model in Figure I, step one in testing for mediation is to establish the existence of a relationship between the independent variables and their interaction in the prediction of the dependent variable. Step two involves testing for a relationship between the independent variables and their interaction in the prediction of the mediator. If, and only if, steps one and two are satisfied can one proceed to steps three and four. If steps one and two result in non-significant relationships, then the relationships to be mediated do not exist and the analysis should stop there.

Steps three and four are examined in a hierarchical regression where the independent variables and their interaction are entered and then followed by the mediator. If the presence of the mediator in the equation reduces the independent variables or their interaction to non-significance, then there is evidence of full mediation. If the independent variables and their interaction remain significant, but less in magnitude when the mediator is in the equation, then partial mediation exists. Lastly, we centered the independent variables before the interaction term was computed in order to avoid the effects of multicollinearity (Aiken and West, 1991). Sobel Tests

Whereas Baron and Kenny’s (1986) multi-step test for mediation is commonly used by researchers, it only provides a broad characterization of the indirect (i.e., the path from the predictor to the mediator to the criterion) effect by classifying mediation as either full, partial, or none at all. Increasingly, researchers are turning their mediation tests to a single-step procedure that provides a precise measure of the level of statistical significance for the indirect effect of a mediator. For this, MacKinnon et al. (2002) recommend the Sobel (1982) test. The Sobel test uses the standard errors of regression weights and the t-tests of these coefficients to produce a measure interpreted as a Z-score. It should be noted that MacKinnon et al. (1995) consider this test to be very conservative. We conducted Sobel tests using the Aroian (1944) adjustment advocated by Preacher and Leonardelli (2003). The Aroian (1944) adjustment to the test does not assume that the standard errors of the independent variables are exceedingly small, as the original Sobel test does. The Aroian version is an appropriate tool for researchers examining mediation using the Baron and Kenny (1986) framework. The intention of the Sobel test is to provide additional evidence regarding whether the mediator is indeed a generative mechanism by which the influence of the predictor is transmitted to the criterion. With this in mind, we use the Sobel test to provide a precise measure of the likelihood that our mediated effects are truly non-zero in the population to which we hope to generalize.

RESULTS

Correlation Results

We ran two-tailed tests of significance for the relationship between the demographic variables and the focal constructs. See Table 1 for these results. Results revealed that of the demographic variables, tenure (r = -.34, p

Factor Analysis Results

Because each of our variables was collected via self-report in a cross-sectional survey, it is possible that our data may suffer from commonsource, common-method bias (Podsakoff and Organ, 1986). To examine if a four-construct model devoid of nuisance or method factors fit the data, we ran confirmatory factor analysis (CFA) using Lisrel 8.72 software (Joreskog and Sorbom, 2004). This also allowed us to examine whether our respondents conceptualized of the focal constructs in our study as distinct and separate. We examined two models: a one-factor model comprised of all items in all four scales and a four-factor model forcing each item to load on its intended factor. In each model, error variances for the items were not allowed to correlate. If the one-factor model provides a fit of the data that is equivalent to the four-factor model, it would indicate a single underlying latent construct. If the four-factor model should provide the better fit than the one-factor model, then we have evidence that our respondents conceptualize of the constructs in our model as independent and distinct. The Chi-square difference test is used to compare the two models.

The fit statistics for the one-factor model were root mean square error of approximation (RMSEA) of .14, comparative fit index (CFI) of .85, standardized root mean square residual (SRMR) of 0.11, and Chi-square (chi^sup 2^) of 2501.48 (df = 902, p

Regression Results

Because Ferris and colleagues (1989, 2002) suggested that certain demographic characteristics and aspects of work experience would be related to POP, we controlled for age, race, sex, tenure, and hierarchical level in each of our regression models. In step one of the test for mediation, we regressed the dependent variable of distributive justice on the control variables, independent variables, and the interaction term. see Table 2 for these results. The regression resulted in a significant F-score with an R-squared value of .42. The beta weight for work locus of control was significant at .38 (p

In step two of the test for mediation, we regressed POP (our mediator) on the control variables, independent variables, and the interaction term. See Table 3 for these results. The regression resulted in a significant F-score with an R-squared value of .58. The beta weight for work locus of control was significant at-.35 (p

In steps three and four of the test for mediation, we ran a hierarchical regression with distributive justice as the dependent variable, with the control variables and independent variables of work locus of control and leader-member exchange and their interaction entered first, followed hierarchically by the mediator of POP. See Table 4 for these results. The regression resulted in a significant F score with an R-squared value of .51. In the first step of the regression the beta weight for work locus of control was significant at .38 (p

Sobel Test Results

Consistent with the results above, the results of our Sobel tests also suggest that perceptions of politics is a generative mechanism through which the effects of leader-member exchange and work locus of control and their interaction are transmitted to distributive justice perceptions. Our Sobel tests revealed significant indirect effects for each of these three relationships. The Sobel test statistic (interpreted as a Z-test) for work locus of control was 3.38 (p

We found that leader-member exchange, work locus of control, and their interaction are valid predictors of both perceptions of organizational politics (POP) and of distributive justice. Our graph (see Figure H) of the interaction showed support for the relationship between the interactive effects of leader-member exchange and work locus of control in the prediction of distributive justice, and revealed that high levels of the antecedents predict high levels of the outcome. However, our other graph (see Figure III) shows that the highest levels of organizational politics are also perceived when one has a high quality relationship with their supervisor combined with a strong internal work locus of control. This is somewhat unexpected, as most previous research on the main effects of these variables indicates that one would expect low POP when one has a high quality relationship with their supervisor or a strong internal work locus of control.

Our results reflect a mutually reinforcing relationship, in that the interaction between a strong internal work locus of control and high leadermember exchange counter-intuitively predicts high levels of POP. Langfred found this result in his mediated moderation study and referred to it as “too much of a good thing” (2004: 385). Perhaps having a strong internal work locus of control and a great relationship with one’s supervisor is too much of a good thing as well. We suggest that such persons in such situations may have a heightened awareness of organizational politics as a result of the combination of a strong internal locus of control and a good relationship with their supervisor.

Another potential explanation is that these high internals may possibly feel a stronger “frustration effect” of sorts, in that they prefer to think they are in charge of their own fate and have developed positive relations with their supervisors to further their ability to be in control. Yet they may perceive that others’ political behaviors thwart their efforts to be in control, which could cause frustration leading to more negative perceptions of distributive justice. Conversely, strong externals may feel less frustration simply because they do not feel they are in control anyway. Thus, externals may accept negative effects of organizational political behaviors on distributive justice more readily.

At the heart of our study, our mediation tests show that POP partially mediates each of the main effects of work locus of control and leader-member exchange in their relationship with distributive justice, but fully mediates the relationship between the interaction of work locus of control and leader-member exchange in its relationship with distributive justice. Furthermore, our Sobel tests show that the main effects of internal work locus of control and leader-member exchange, and their interaction, are mediated by POP in the prediction of distributive justice. Therefore, we have strong evidence that POP is a generative mechanism through which the relationship between a subordinate and supervisor and the subordinate’s perception that they control the events in which they find themselves are transmitted to, and a predictor of, perceptions of distributive justice. In essence, when one has a good relationship with their supervisor combined with a strong internal work locus of control, they are not necessarily likely to perceive that they are fairly rewarded because they are also likely to perceive higher levels of organizational politics, and we found that perceptions of high levels of politics yields low levels of perceptions of fairness. Thus, political perceptions are so strongly present in the perceptions of those high in leader-member exchange and work locus of control, that because of these political perceptions they may perceive that they are unfairly rewarded.

Limitations

Like many survey-based studies, ours potentially suffers from the fact that our sample size is small, but still above the cutoffs provided by Tabachnik and Fidell (2001) for the interpretation of multiple regression results. A larger sample would have allowed us to include more variables in our analysis which might have alleviated model misspecification. It may also have allowed us to utilize structural equation modeling (SEM) rather than regression analysis. However, in our sample the number of subjects per estimated parameter in our model is below the commonly accepted threshold for SEM of five to ten (Kerlinger, 1986); so SEM was inadvisable for our study. Therefore, we suggest that our use of hierarchical regression analysis and the Baron and Kenny (1986) framework was appropriate.

Managerial Implications

As in previous studies (e.g., Andrews and Kacmar, 2001), our initial correlational analyses (see Table 1) reflect the interrelatedness of the variables in our model. Each of our focal measures was shown to be significantly correlated with one another. In our full model, we found that those with the high internal locus of control combined with the high quality relationship with one’s supervisor actually increases, rather than reduces, individuals’ perceptions that organizational politics are at play in their organizations. Perhaps such high internals with high quality relationships with their supervisors may feel more cognizant of their political system, and may feel greater frustration that they are unable to utilize their strong sense of control and relationship with the boss to impact their organizational rewards as effectively as they would wish to.

Our study’s use of the cognitive social learning (CSL) framework provides some context for understanding the complex dynamics of the workplace. Specifically, as managers seek to foster high quality relationships with their subordinates (as they usually should), they should recognize that subordinates with strong feelings regarding their own ability to manipulate their work experiences may become so in tune with organizational dynamics that by nature of their relationship with their supervisor they are also more aware of organizational politics. This may in turn affect individuals’ perceptions regarding the fairness of the organization’s reward system.

This study examines the mediating influence of perceptions of organizational politics in the relationship between the interactive effects of work locus of control and leader-member exchange on distributive justice. We use the tests for mediation described by Baron and Kenny (1986) and amended for mediated moderation by Muller, Judd, and Yzerbyt (2005). We found that the perception of organizational politics partially mediates the relationships between work locus of control and leader-member exchange with distributive justice. We also found that the perception of organizational politics fully mediates the interaction of work locus of control and leader-member exchange in the prediction of distributive justice. Our results suggest that organizational politics acts as a generative mechanism through which the interaction of one’s relationship with their supervisor and their perception of their own ability to influence events around them is transmitted to their perception of the fairness of organizational rewards.

References

Aiken, L. S. and S. G. West. 1991. Multiple Regression: Testing and Interpreting Interactions. Thousand Oaks, CA: Sage Publications.

Andrews, M. C. and K. M. Kacmar. 2001. “Discriminating Among Organizational Politics, Justice, and Support.” Journal of Organizational Behavior 22: 347-366.

Aroian, L. A. 1944. “The Probability Function of the Product of Two Normally Distributed Variables.” Annals of Mathematical Statistics 18: 265-271.

Baron, R. M. and D. A. Kenny. 1986. “The Moderator-Mediator Variable Distinction in Social Psychological Research: Conceptual, Strategic, and Statistical Considerations.” Journal of Personality and Social Psychology 51: 1173-1182.

Bass, B. M. 1990. Bass and Stodgill’s Handbook of Leadership (3rd ed.). New York, NY: Free Press.

Colquitt, J. A. 2001. “On the Dimensionality of Organizational Justice: A Construct Validation of a Measure.” Journal of Applied Psychology 86: 386-400.

Dansereau, F., J. Cashman and G. Graen. 1973. “Instrumentality Theory and Equity Theory as Complementary Approaches in Predicting the Relationship of Leadership and Turnover among Managers.” Organizational Behavior and Human Performance 10: 184-200.

_____, G. Graen and W. J. Haga. 1975. “A Vertical Dyad Linkage Approach to Leadership Within Formal Organizations: A Longitudinal Investigation of the Role Making Process.” Organizational Behavior and Human Performance 13: 46-78.

Davis, W. D. and W. L. Gardner. 2004. “Perceptions of Politics and Organizational Cynicism: An Attributional and Leader-Member Exchange Perspective.” Leadership Quarterly 15: 439-465.

Dulebohn, J. H. 1997. “Social Influence injustice Evaluations of Human Resource Systems.” In Research in Personnel and Human Resources Management (Vol. 15). Ed. G. R. Ferris. Greenwich, CT: JAI Press, pp. 241-291.

Ferris, G. R., G. Adams, R. W. Kolodinsky, W. A. Hochwarter and A. P. Ammeter. 2002. “Perceptions of Organizational Politics: Theory and Research Directions.” In Research in Multi-level Issues: The Many Faces of Multi-level Issues (Vol. 1). Eds. F. J. Yammarino and F. Dansereau. Amsterdam: JAI Press, pp. 179-254.

_____, D. D. Frink, T. A. Beehr and D. C. Gilmore. 1995. “Political Fairness and Fair Politics: The Conceptual Integration of Divergent Constructs.” In Organizational Politics, Justice, and Support: Managing the Social Climate of the Workplace. Eds. R. S. Cropanzano and K. M. Kacmar. Westport, CT: Quorum, pp. 21-36.

_____, _____, M. C. Galang, J. Zhou, K M. Kacmar and J. L. Howard. 1996. “Perceptions of Organizational Politics: Prediction, Stress-related Implications and Outcomes.” Human Relations 49: 233- 266.

_____and K. M. Kacmar. 1992. “Perceptions of Organizational Politics. “Journal of Management 18: 93-116. _____, G. S. Russ and P. M. Fandt. 1989. “Politics in Organizations.” In Impression Management in the Organization. Eds. R. A. Giacalone and P. Rosenfield. Hillsdale, NJ: Lawrence Erlbaum. pp. 143-170.

Fiske, S. T. and S. E. Taylor. 1991. Social Cognition (2nd ed.). Reading, MA: Addison-Wesley Publishing.

Gandz, J. and V. V. Murray. 1980. “The Experience of Workplace Politics.” Academy of Management Journal 23: 237-251.

Gerstner, C. R. and D. V. Day. 1997. “Meta-Analytic Review of Leader-Member Exchange Theory: Correlates and Construct Issues.” Journal of Applied Psychology 82: 827-844.

Graen, G. B. 1976. “Role-making Processes within Complex Organizations.” In Handbook of Industrial and Organizational Psychology. Ed. M. D. Dunnette. Chicago, IL: Rand McNally. pp. 1201- 1245.

_____ and J. Cashman. 1975. “A Role-making Model of Leadership in Formal Organizations: A Developmental Approach.” In Leadership Frontiers. Eds. J. G. Hunt and L. L. Larson. Kent, OH: Kent State University, pp. 143-165.

_____, M. Novak and P. Sommerkamp. 1982. “The Effects of Leader- Member Exchange and Job Design on Productivity and Satisfaction: Testing a Dual Attachment Model.” Organizational Behavior and Human Performance 30: 109-131.

Greenberg, J. 1990. “Organizational Justice: Yesterday, Today, and Tomorrow.” Journal of Management 16: 399-342.

_____. 1987. “A Taxonomy of Organizational Justice Theories.” Academy of Management Review 12: 9-22.

Harris, K. J., M. James and R. Boonthanon. 2005. “Perceptions of Organizational Politics and Cooperation as Moderators of the Relationship Between Job Strains and Intent to Turnover. “Journal of Managerial Issues 17 (1): 26-42.

Hochwater, W. A., L. A. Witt and K. M. Kacmar. 2000. “Perceptions of Organizational Politics as a Moderator of the Relationship Between Conscientiousness and Job Performance.” Journal of Applied Psychology 85: 472-478.

Hu, L. and P. M. Bentler. 1999. “Cutoff Criteria for Fit Indexes in Covariance Structure Analysis: Conventional Criteria versus New Alternatives.” Structural Equation Modeling 6: 1-55.

_____ and _____. 1998. “Fit Indices in Covariance Structure Modeling: Sensitivity to Underparameterized Model Misspecification.” Psychological Methods 3: 424-453.

James, L. R. and A. P.Jones. 1980. “Perceived Job Characteristics and Job Satisfaction: An Examination of Reciprocal Causation.” Personnel Psychology 33: 97-135.

_____ and L. E. Tetrick. 1986. “Confirmatory Analytic Tests of Three Causal Models Relating Job Perceptions to Job Satisfaction.” Journal of Applied Psychology 71: 77-82.

Joreskog, K. and D. Sorborn. 2004. LISREL 8.72 [computer software]. Chicago, IL: Scientific Software International.

Kacmar, K. M. and R. A. Baron. 1999. “Organizational Politics: The State of the Field, Links to Related Processes, and an Agenda for Future Research.” In Research in Personnel and Human Resource Management (Vol. 17). Ed. G. R. Ferris. Amsterdam: JAI Press, pp. 1- 39.

_____, D. P. Bozeman, D. S. Carlson and W. P. Anthony. 1999. “An Examination of the Perceptions of Organizational Politics Model: Replication and Extension.” Human Relations 52: 383-415.

_____ and D. S. Carlson. 1997. “Further Validation of the Perceptions of Politics Scale (POPS): A Multiple Sample Investigation.” Journal of Management 23: 627-658.

_____ and G. R. Ferris. 1991. “Perceptions of Organizational Politics Scale (POPS): Development and Construct Validation.” Educational and Psychological Measurement 51: 193-205.

Kerlinger, F. N. 1986. Foundations of Behavioral Research (3rd ed.). New York, NY: Holt, Rinehart and Winston.

Kinicki, A. J. and R. P. Vecchio. 1994. “Influences on the Quality of Supervisor-Subordinate Relations: The Role of Time- Pressure, Organizational Commitment, and Locus of Control.” Journal of Organizational Behavior 15: 75-82.

Langfred, C. W. 2004. “Too Much of a Good Thing? Negative Effects of High Trust and Individual Autonomy in Self-Managing Teams.” Academy of Management Journal 47: 385-399.

Lerner, M. J. 1980. The Belief in a Just World: A Fundamental Delusion. New York, NY: Plenum Press.

Liden, R. C., R. T. Sparrowe and S. J. Wayne. 1997. “Leader- Member Exchange Theory: The Past and Potential for the Future.” Research in Personnel and Human Resource Management 15: 47-119.

MacKinnon, D. P., C. P. Lockwood, J. M. Hoffman, S. G. West and V. Sheets. 2002. “A Comparison of Methods to Test the Significance of the Mediated Effect.” Psychological Methods 7: 83-104.

_____, G. Warsi and J. H. Dwyer. 1995. “A Simulation Study of Mediated Effect Measures.” Multivariate Behavioral Research 30: 41- 62.

Manogran, P., J. Stauffer and E. J. Conlon. 1994. “Leader-Member Exchange as a Key Mediating Variable between Employees’ Perceptions of Fairness and Organizational Citizenship Behavior.” National Academy of Management Meeting Proceedings: 249-267.

Martin, R., G. Thomas, K. Charles, O. Epitropaki and R. McNamara. 2005. “The Role of Leader-Member Exchanges in Mediating the Relationship between Locus of Control and Work Reactions.” Journal of Occupational and Organizational Psychology 78: 141-147.

Mintzberg, H. 1973. The Nature of Managerial Work. New York, NY: Longman.

Muller, D., C. M. Judd and V. Y. Yzerbyt. 2005. “When Moderation is Mediated and Mediation is Moderated.” Journal of Personality and Social Psychology 89: 852-863.

Parker, C. P., R. L. Dipboye and S. L. Jackson. 1995. “Perceptions of Organizational Politics: An Investigation of Antecedents and Consequences.” Journal of Management 21 (5): 891- 912.

Podsakoff, P. M. and D. W. Organ. 1986. “Self-Reports in Organizational Research: Problems and Prospects. “Journal of Management 12: 531-544.

Preacher, K.J. and G. J. Leonardelli. 2003. Calculation of the Sobel Test. Retrieved July 17, 2006 from http://www.unc.edu/ ~preacher/sobel/sobel.htm

Price, J. L. and C. W. Mueller. 1986. Handbook of Organizational Measurement. Marshfield, MA: Pittman.

Rotter, J. B. 1966. “Generalized Expectancies for Internal Versus External Control of Reinforcement.” Psychological Monographs: General and Applied 80: 1-28.

_____. 1954. Social Learning and Clinical Psychology. Englewood Cliffs, NJ: Prentice-Hall.

Scandura, T. A. 1999. “Rethinking Leader-Member Exchange: An Organizational Justice Perspective.” Leadership Quarterly 10: 25- 40.

Sobel, M. E. 1982. “Asymptotic Intervals for Indirect Effects in Structural Equations Models.” In Sociological Methodology. Ed. S. Leinhart. San Francisco, CA: Jossey-Bass. pp. 290-312.

Spector, P. E. 1988. “Development of the Work Locus of Control Scale.” Journal of Occupational Psychology 61: 335-340.

_____. 1982. “Behavior in Organizations as a Function of Employee’s Locus of Control.” Psychological Bulletin 91: 482-497.

Sweeney, P. D., D. B. McFarlin and J. S. Cotton. 1991. “Locus of Control as a Moderator of the Relationship between Perceived Influence and Procedural Justice.” Human Relations 44: 333-342.

Tabachnik, B. G. and L. S. Fidell. 2001. Using Multivariate Statistics (4th ed.). Needham Heights, MA: Allyn and Bacon.

Tyler, T. R. 1994. “Psychological Models of the Justice Motive: Antecedents of Distributive and Procedural Justice. “Journal of Personality and Social Psychology 67: 850-863.

Valle, M. and P. L. Perrewe. 2000. “Do Politics Perceptions Relate to Political Behavior? Tests of an Implicit Assumption and Expanded Model.” Human Relations 53: 359-385.

Vecchio, R. P., R. W. Griffeth and P. W. Horn. 1986. “The Predictive Utility of the Vertical Dyad Linkage Approach.” Journal of Social Psychology 126: 617-625.

Vigoda, E. 2000. “Internal Politics in Public Administration System: An Empirical Examination of its Relationship with Job Congruence, Organizational Citizenship Behavior, and In-RoIe Performance.” Public Personnel Management 29: 185-210.

Wayne, S. J., L. M. Shore, W. H. Bommer and L. E. Tetrick. 2002. “The Role of Fair Treatment and Rewards in Perceptions of Organizational Support and Leader-Member Exchange. “Journal of Applied Psychology 87: 590-598.

Weiner, B. 1980. Human Motivation. New York, NY: Holt, Rinehart and Winston.

Zanzi, A. and R. M. O’Neill. 2001. “Sanctioned Versus Non- sanctioned Political Tactics.” Journal of Managerial Issues 13 (2): 245-262.

Zhou, J. and G. R. Ferris. 1995. “The Dimensions and Consequences of Organizational Politics Perceptions: A Confirmatory Analysis. “Journal of Applied Social Psychology 25 (19): 1747-1764.

Brian K. Miller

Assistant Professor of Management

Texas State University

Kay McGlashan Nicols

Assistant Professor of Management

Texas State University

Copyright Pittsburg State University, Department of Economics Summer 2008

(c) 2008 Journal of Managerial Issues; JMI. Provided by ProQuest Information and Learning. All rights Reserved.