LEADER-MEMBER EXCHANGE AS A PREDICTOR OF
INTENTIONS TO LEAVE: DO EMPLOYEE ENGAGEMENT
AND AUTONOMY MATTER?
MASTER THESIS
MSc Human Resource Management (HRM)
University of Groningen, Faculty of Economics and Business
August 2018
Stefan Braam
Student number: s1912046
Supervisor: T.A. de Vries University of Groningen
ABSTRACT
The concept of employee engagement has attracted more and more attention from both scholars and organizations since the last decade. The antecedents as well as the role of middle managers in the process of enhancing employee engagement, however, are rather overlooked and rarely empirically investigated. In addressing this gap and by adopting an engagement perspective, the indirect effect of leader-member exchange (LMX) on intentions to turnover through employee engagement was examined. In addition, the moderating role of autonomy was included, thereby suggesting a moderated mediation effect. The conceptual model was tested by collecting data from 95 employees working at an internationally operating bank in the Netherlands. Results showed that LMX had a significant, positive influence on engagement and a negative impact on intentions to turnover. Furthermore, employee engagement acted as a partial mediator in the relationship between LMX and intentions to turnover. Other than expected, LMX was positively related to employee engagement under low levels of autonomy. Under high levels of autonomy, this connection was unrelated. Lastly, the overall moderated mediation was not found to be significant. The findings suggest that immediate leaders play a critical role in engaging employees through their direct relationship with these individuals. As such, organizations should invest in enhancing the quality of LMX. This will not just positively influence engagement levels, but also limit staff’s intention to turnover and the high costs associated with it. Finally, some directions for future research are discussed.
INTRODUCTION
Although scholarly endeavors on the topic of employee engagement have only recently attracted attention, employee engagement is a flourishing concept among contemporary organizations and consulting firms alike for quite some years now (Saks, 2006; Wollard & Shuck, 2011). Indeed, multiple studies found empirical evidence that higher engagement levels positively impact on firm performance measures like job performance, customer loyalty, profitability (Christian, Garza, & Slaughter, 2011; Harter, Schmidt, & Hayes, 2002) and nonmonetary rewards (Belgio, 2018). Similarly, Agarwal, Datta, Blake-Beard, and Bhargava (2012) found that work engagement positively correlates with innovative work behavior. Results of a study by Saks (2006) further showed that employee engagement is negatively related to workers’ intentions to quit.
Despite the fact that in general quite some research thus shows that employee engagement can significantly influence business outcomes, empirically little is known about its antecedents (Wollard & Shuck, 2011). To date many factors are brought up in literature as being potential predecessors of engagement. In fact, by means of a structured literature review, Wollard and Shuck (2011) listed a total of 42 possible antecedents of employee engagement. Half of which pertained to antecedents directed and governed by the organization and the other half relating to individually antecedents. The suggested antecedents are numerous, yet the authors state only few have been extensively empirically tested and validated. Moreover, “there is little evidence for their use and … antecedents that have been identified are scattered throughout a large literature base” (Wollard & Shuck, 2011: 431). In their studies on antecedents and consequences of employee engagement, Saks (2006) and Bailey, Madden, Alfes, and Fletcher (2017) also conclude that employee engagement is a meaningful concept that requires further research into finding more potential predictors of engagement. An understanding and appreciation of these antecedents is, after all, fundamental in reaping the benefits of an engaged workforce (Rich, Lepine, & Crawford, 2010).
almost no research focused on middle managers in general and the quality of LMX in specific as predicting employee engagement (Agarwal et al., 2012). Some research did show that supervisory feedback (Menguc, Auh, Fisher, & Haddad, 2013), perceived organizational support, and supervisory support (Bailey et al., 2017) all precede engagement, presuming a unilateral and top-down relation from the organization or the supervisor/manager towards the subordinate. LMX, however, represents a bilateral bond between the worker and his/her direct manager and thereby addresses a void in scholarly work on engagement. Another motivation lies in the trivial role middle management fulfills within organizations. Especially in engaging staff they are ascribed a crucial role through their (in)ability to inspire and support employees and facilitate employee development (Frank, Finnegan, & Taylor, 2004; Thornhill, & Saunders, 1998).
Intentions to turnover is conceptualized as the outcome variable in this study. The focus on this variable is explained by the fact that this concept is understudied in engagement literature (Byrne, Hayes, & Holcombe, 2017) and as such provides an interesting research opportunity. Moreover, one of the most crucial managerial challenges today is retaining valued and talented personnel (Reina, Rogers, Peterson, Byron, & Hom, 2018). Especially considering the current tight labor market conditions and the significantly high costs involved in employee turnover (Kim & Koo, 2017) this seems a more than justifiable undertaking.
In addition, this study indicates that autonomy may moderate the relationship between the quality of LMX and employee engagement. Autonomy is an interesting job resource, since an essential characteristic of being a professional is by having autonomy in one’s job (Farr-Wharton, Brunetto, & Shacklock, 2011). Also, the degree of autonomy is regarded as a job design feature that is greatly valued by employees (Bakker & Demerouti, 2007). Thus, the degree to which managers are willing to loosen their control on employees’ work is expected to affect the link between LMX and engagement.
FIGURE 1 Conceptual Model
THEORY AND RESEARCH HYPOTHESES
Employee engagement and intentions to turnover
There have been debates over the meaning of employee engagement since the term came into vogue in the 1990s and from then is increasingly used by both scholars as well as practitioners (Macey & Schneider, 2008; Hallberg & Schaufeli, 2006). Founded on the idea that individuals deploy varying levels of physical, cognitive and emotional effort in performing their jobs, the term engagement was first introduced by Kahn (1990). He distinguished two opposite forms, personal engagement and personal disengagement. The former refers to the attachment of organization members’ selves to their work roles, whereas the latter is defined as the detachment of organization members’ selves from their work roles. Based hereupon, Harter et al. (2002) defined employee engagement as an employee’s involvement and satisfaction with as well as enthusiasm for work. Others label engagement as going the extra mile expressed by employees bringing discretionary effort to the workplace by means of extra time, brainpower and energy put into the job (Frank et al., 2004). An alternative line of thought in defining employee engagement can be found in the work of burnout scholars as they regard engagement as the positive antithesis of a burnout (Maslach, Schaufeli, & Leiter, 2001). Burnout is characterized by exhaustion, cynicism, and ineffectiveness, engagement by contrast entails high levels of energy, involvement, and efficacy.
The present paper, however, follows Schaufeli and colleagues’ (2001) call for future research to concentrate on factors that boost and positively impact employee engagement. Thus, in this paper employee engagement is conceptualized as “a positive, fulfilling, work-related
Employee
Engagement Intentions to turnover
Leader Member Exchange
state of mind that is characterized by vigor, dedication and absorption” (Schaufeli, Salanova, González-Romá, & Bakker, 2001: 74). Vigor refers to the readiness to invest effort in one’s job, high levels of energy and mental resilience while working, and persistence when faced with difficulties. Dedication comprises a sense of enthusiasm, pride, challenge and inspiration. Lastly, absorption is characterized as being absorbed in work where time passes quickly and one cannot easily detach oneself from work. Employee engagement in essence thus focuses on the work itself rather than on the organization (Maslach, Schaufeli, & Leiter, 2001). Saks (2006), therefore, explicitly proposed to distinguish between job and organization engagement. Though related, he found evidence for the two being distinct constructs. Job engagement focuses on the work role of organizational members, whereas organization engagement’s emphases is on employees’ role as a member of an organization. This conceptual difference is also empirically endorsed by Hallberg and Schaufeli (2006). Since for this study the definition of Schaufeli et al. (2002) is adopted, which centers a work-related state of mind, in the remainder of this paper employee engagement will embody the job engagement aspect of engagement.
tendency to quit (Reina et al., 2018; Shuck et al., 2011). In a similar vein, Agarwal et al.’s (2012) study into organizations operating in the service sector in India highlights that employees that experience emotions as interest, joy and enthusiasm in their work, have less intentions to leave. From this, one can argue that the more employees are engaged, the less likely it is that they are inclined to leave the organization. This gives rise to the following hypothesis.
Hypothesis 1: Employee engagement has a negative effect on intentions to turnover.
LMX and employee engagement
In scientific literature, the Job Demands-Resources (JD-R) model is the most widely accepted conceptualization of engagement (Bailey et al., 2017). Stemming from this model, Bakker and Demerouti (2007) state that four categories of job resources can be accounted to having a motivational potential and lead to high employee engagement. These are: (1) organizational level (e.g. pay, job security); (2) interpersonal level (e.g. teamwork, supervisor support); (3) work level (e.g. role clarity, participation), and; (4) task level (e.g. autonomy; skill variety). As was pointed out in the introduction section, the lion share of research on antecedents of employee engagement concentrates on either organization or individual work(er) antecedents (Wollard & Shuck, 2011), which by definition reflect three of the job resources classifications: organizational level, work level, and task level. Gruman and Saks (2011), for example, suggest that employee engagement can be enhanced by paying attention to job design, social support, and training. Next, Saks (2006) found empirical evidence that perceived organizational support, job characteristics (e.g. task variety) and procedural justice are all predictors of the level of employee engagement. In addition, he hypothesizes that training programs, incentive compensation, and flexible work arrangements might also be measures to augment engagement. Pitt-Catsouphes and Matz-Costa (2008) examined the influence of the latter and concluded that flexibility indeed is a powerful positive predictor of engagement. Other work examined the influence of task variety, task significance (Christian et al., 2011), job fit, psychological climate and emotional identification (Shuck et al., 2011) and discovered that these factors were all positively related to employee engagement.
et al., 2017). That is, the fourth category of the JD-R model: interpersonal level, is rather understudied. Frank et al. (2004), for example, consider first-line leaders to be notably important in the process of improving employee engagement. The role of middle managers in enhancing employee engagement is further exemplified by Aon Hewitt (2011), a large UK-based consulting firm. In their European Manager Survey 2011 they explored the role of middle managers in the follow-up process of engagement surveys. By collecting data from 731 middle managers from 10 European countries, they found that middle managers are a fundamental linkage in the creation and improvement of employee engagement. Likewise, managerial behaviors and their relations with subordinates are found to play a major role in the extent to which employees invest in the job (Reina et al., 2018).
Early research by Dansereau, Graen, and Haga (1975) on organizational leadership, provides a means to incorporate the role of middle managers in engaging employees. Specifically, they argue the quality of the relationship between subordinates and their direct supervisors is a significant predictor of employees’ commitment to and in their jobs. This theory is known as the leader-member exchange (LMX) theory (Graen & Uhl-Bien, 1995). The foundation of LMX can be explained by the Social Exchange Theory (SET). Hence, employees feel a sense obligation to repay their supervisors for the provision of work opportunities, trust and meaningful work. An example of repayment is to invest high levels of energy and engagement in performing one’s job (Agarwal et al., 2012). Although the effect of LMX on employee engagement is only investigated recently and by a mere few studies thus far, the results nevertheless showed promising results in the sense that all concluded that LMX is a strong and significant antecedent of employee engagement (Agarwal, et al., 2012; Cheng, Lu, Chang, & Johnstone, 2013). Research on 43 five-star hotels in Korea involving 290 employees, in addition, showed that high quality relationships between managers and employees helps the latter group to engage in their job (Kim & Koo, 2017). Moreover, the researchers postulate that LMX is essential for nurturing and enhancing a workplace climate characterized by full engagement. Taking these results into account and underwriting the research models that emphasize the relationship between supervisors and subordinates and so position LMX as predicting engagement; it is proposed that LMX positively affects employee engagement.
The mediating role of employee engagement
Recent studies on engagement which emerged in the last decade with the growing interest in the subject, consider the concept of employee engagement as mediating the relationship between its antecedents and outcomes (Bakker & Demerouti, 2007; Christian et al., 2011; Konrad, 2006; Saks, 2006). This coincides with the claim that job resources (e.g. LMX) function as a predictor of employee well-being which consequently has an impact on organizational outcomes, as proposed by the JD-R model. LMX, for instance, is considered as a meaningful antecedent to increased levels of energy of employees and the effort they put into the job (Reina et al., 2018); one of the key characteristics of engaged workers. As a result, this high energy and work motivation is expressed by staff’s behavior (Kahn, 1990). In fact, Rich et al. (2010) state that employee engagement serves as a mechanism through which employee characteristics and organizational factors lead to job performance. They found statistical significant evidence for this by surveying 245 firefighters employed by four municipalities. Conclusion was that engagement fully accounted for the relationship between antecedents (value congruence, organizational support, self-evaluations) and outcomes (task performance, organizational citizenship behavior). Other research showed similar results. Hence, a partial indirect relation through engagement was found for, amongst others, job characteristics, supervisor support, rewards and recognition on the one hand and organizational commitment and job satisfaction on the other (Saks, 2006). What’s more, when employees perceive high supervisory support, this high quality relationship will have an effect on the subordinates’ job participation and engagement (Cheng et al., 2013), and as a consequence, employees are less tempted to leaving the organization (Agarwal et al., 2012; Saks, 2006)
The first two hypotheses predicted that (1) leader-member exchange is positively related to the level of engagement and that (2) employee engagement, in turn, is negatively related to intentions. Accordingly, it is proposed here that employee engagement mediates the relationship between LMX and intentions to turnover.
Hypothesis 3: The relationship between LMX and intentions to turnover is mediated by employee engagement.
Autonomy as a moderator
(Morgeson & Humphrey, 2006). Autonomy was already underpinned as a resource with motivational potential within jobs by Hackman and Oldham (1975). Moreover, and in line with the JD-R model, Bakker and Demouteri (2007) claim that autonomy is strongly related to extra-role performance through its relationship with engagement. Hence, its availability increases engagement of employees, which in turn fosters positive employee outcomes (Agarwal et al., 2012). The question then arises in what way and how autonomy is exactly related to engagement. Kahn (1990), in that light, states that individuals feel safer when they have some freedom and control over their job and that situations where managers are unwilling to loosen their control on employees’ work, this will send a message that employees are not to be trusted. This will then lead to employees investing less of their selves at work as well as exerting less energy. The principles of SET apply. That is, it is more likely for employees who are provided with tasks that are high on the job characteristics (e.g. autonomy), to reciprocate with greater job engagement (Bakker & Demerouti, 2007). ‘More likely’ since an employee is given more opportunities and freedom in decision-making and methods to perform one’s job. Seeing that supportive managers and high quality interpersonal relationships directly foster feelings of psychological safety that increase willingness to fully engage in work roles (Saks, 2006); this supposes a direct moderation of job autonomy on the relationship between interpersonal (i.e. supervisor-subordinate) relations and employee engagement. Specifically, LMX theory suggests that middle managers employ a different style of leadership over time depending on the relationship they have with their subordinates and they “convey their expectations of subordinates through their work assignments” (DeConinck, 2009: 1082). The degree of autonomy in those tasks can vary and thereby have an effect on the relationship between LMX and employee engagement. Hence, in the present paper it is posited that autonomy has a complementary impact on the relationship between LMX and employee engagement.
in situations of high work autonomy, individuals are less dependent on their employer.Volmer, Spurk, and Niessen’s (2012) work on the moderating role of job autonomy on the relation between LMX and creative work involvement showed that LMX was positively related to creative work involvement under high levels of job autonomy. Under minimum levels of job autonomy, however, this connection proved to be unrelated. Further research showed that at high levels of perceived autonomy, supervisory support had a positive effect on employee engagement (Menguc et al., 2013). That is to say, at high levels of autonomy employees are better able and more willing to embrace and integrate supervisory support in their jobs and thereby striving to flourish by exerting more energy and devotion. On the contrary, when employees perceived little autonomy, they showed little motivation and empowerment to put the, still sufficient, supervisory support into action.
Translating the previous notions to the objective of this paper implies that a high quality bond between supervisors and their subordinates in itself is insufficient in explaining the connection of LMX and its effects. Incorporating job design features in the research design is, therefore, a well worth undertaking (Bakker & Demerouti, 2007; Volmer et al., 2012). Taken together it is, therefore, posited that the link between LMX and employee engagement hinges on the level of autonomy that employees experience. Taking SET as the basis and taking previous research outcomes into account, it is thus expected that autonomy acts as a moderator in the relationship between leader-member exchange and employee engagement such that high levels of autonomy have a positive effect on the relationship between LMX and engagement.
Hypothesis 4: The relationship between LMX and employee engagement is moderated by autonomy, such that this relationship is stronger under high autonomy and weaker under low autonomy.
When combining the foregoing hypotheses, based on the conceptual model (see figure 1) to be tested and presented in the introduction, this paper predicts a moderated mediation effect. This leads to the fifth and last hypothesis.
METHOD
Sample and procedure
To study the conceptual model and thereby testing the hypotheses, a survey based on purposive sampling was conducted amongst employees working at an internationally operating Dutch bank. As a provider of services, a bank places great emphasis on human resources and employee engagement. In addition, banks are traditionally hierarchical organizations in which the vast majority of the workforce is led by middle managers. This organizational layer plays a decisive role in the quality of LMX and the creation and enhancement of the level of engagement amongst staff, both in a positive as well as a negative way.
In total 137 surveys (in English) were distributed by including a weblink in the emails that were sent to employees working directly or indirectly on a global process improvement project. In the body of the email, the content and purpose of the study and the survey were explained and introduced to each respondent. To rule out the chance of missing data, the Qualtrics Survey tool was used. This tool facilitates the feature to make questions in the digital survey mandatory to be filled before submitting the survey. To guarantee confidentiality, the survey was fully anonymous and participation was on voluntary basis. With a response rate of 69%; 95 (N = 95) completed and usable surveys were registered. The sample was almost equally distributed over men (n = 47, 49.5%) and women (n = 48, 50.5%). The age distribution was as follows: under 25 years n = 3 (3.2%); 25 – 34 years n = 19 (20.0%); 35 – 44 years n = 37 (38.9%); 45 – 54 years n = 31 (32.6%); 55 years and over n = 5 (5.3%). Most participants worked more than 5 years in their current job (n = 38, 40.0%). The other tenure categories showed the following distribution: less than 1 year n = 11 (11.6%); 1 year to less than 2 years n = 13 (13.7%); 2 years to less than 3 years n = 7 (7.4%); 3 years to less than 5 years n = 26 (27.4%).
Measures
Employee engagement. Rich et al.’s (2010) emotional job engagement scale was used to assess employee engagement. Some example items of this six-item scale are “I feel energetic at my job”, “I am interested in my job”, and “I feel positive about my job”. Participants indicated their response on a five-point Likert scale with anchors (1) strongly disagree to (5) strongly agree. In the present study, the reliability (α) of the scale was .84.
Leader-member exchange was measured using five items adopted from DeConinck (2009). “My immediate supervisor recognizes my potential” and “my working relationship with my immediate supervisor is extremely effective”, are two examples of items that participants were requested to rate using using a five-point Likert-type scale (1 = strongly disagree, 5 = strongly agree). The Cronbach’s alpha coefficient for the scale equaled .89.
Autonomy was measured using six items adopted from Morgeson and Humphrey’s (2006) study on measures for assessing work and job design. Participants were asked to indicate the extent to which they experience autonomy in performing their job. They responded using a five-point Likert response scale ranging from (1) strongly disagree to (5) strongly agree to items such as, “the job allows me to make a lot of decisions on my own” and “the job allows me to decide on my own how to go about doing my work” (α = 0.93).
All the above-mentioned variables under study demonstrated good levels of reliability and the items to measure each are widely tested and used in scientific research and literature. A total overview of all survey items per measure can be found in appendix A.
Finally, the control variables: gender, age, and tenure were included. Prior research, namely, demonstrated that employees’ age, tenure, and gender can predict employees’ motivation, engagement (Avery, McKay, & Wilson, 2007; Sarker, Crossman, & Chinmeteepituck, 2003), and turnover (Hom, Ellis, & Roberson, 2008). All three control variables were measured by one survey item each. Age and tenure were both measured using a five-point scale. The former had the following categories: “under 25”; “25-34”; “35-44”; “45-54”; “55 and over”, whereas the latter was measured by categories ranging from “less than 1 year” to “more than 5 years. Gender was coded as a binary variable, where 0 = female, and 1 = male.
Data analysis
Subsequently, the Process Macro for SPSS by Hayes (2013) was utilized to test the proposed mediation and moderation effects as well as the total moderated mediation effect. The effect of LMX on employee engagement and the conditional indirect effect of LMX on intentions to turnover through employee engagement was analyzed for three conditions of the moderator, autonomy (at the mean and one standard deviation above and below the mean to represent high and low autonomy). In all data analyzes, the control variables gender, age, and tenure were included.
RESULTS
Descriptive statistics
TABLE 1
Means, Standard Deviations and Correlations
M SD 1 2 3 4 5 6 7 1. Intentions to turnover 2.47 0.93 (.88) 2. Employee engagement 4.11 0.56 -.58** (.84) 3. LMX 3.67 0.71 -.49** .38** (.89) 4. Autonomy 3.78 0.72 -.32** .37** .37** (.93) 5. Gender 0.49 0.50 .10 -.14 -.10 -.07 - 6. Age 3.17 0.92 -.11 .22* .03 .19 -.14 - 7. Tenure 3.71 1.41 .01 .05 -.06 -.01 .03 .31** -
Notes: N = 95; * p < .05; ** p < .01; Gender: 0 = ‘Female’, 1 = ‘Male’; Age: 1 = ‘Under 25’, 2 = ‘25-34’, 3 = ‘35-44’, 4 = ‘45-54’, 5 = ‘55 and over’; Tenure: 1 = ‘Less than 1 year’, 2 = ‘1 year to less than 2 years’, 3 = ‘2 years to less than 3 years’; 4 = ‘3 years to less than 5 years’, 5 = ‘More than 5 years’; Cronbach Alpha Coefficients in parentheses).
Hypothesis testing
Hypothesis 1 predicted that employee engagement would have a negative effect on intentions to turnover. Results indeed showed a significant negative association between employee engagement and intentions to turnover (B = -.97, t(90) = -6.61, p < .001). Hypothesis 1 is thus supported (see table 2).
TABLE 2
Employee Engagement and Intentions to Turnover Regression Analysis Dependent variable intentions to turnover
Predictor B SE t p Control variables Gender .03 .16 .15 .878 Age .01 .09 .13 .900 Tenure .02 .06 .41 .683 Study variables Employee engagement -.97* .15 -6.61 .000 R2 .34
Notes: N = 95; B = unstandardized regression coefficient; SE = standard error; * p < .001.
TABLE 3
LMX and Employee Engagement Regression Analysis Dependent variable employee engagement
Predictor B SE t p Control variables Gender -.09 .11 -.84 .402 Age .12 .06 1.96 .053 Tenure .01 .04 .12 .905 Study variables LMX .29* .07 3.89 .000 R2 .20
Notes: N = 95; B = unstandardized regression coefficient; SE = standard error; * p < .001.
TABLE 4
Employee Engagement Mediating Effect Analysis Dependent variable intentions to turnover (step 1)
Predictor B SE t p Control variables Gender .07 .17 .40 .693 Age -.09 .10 -.95 .343 Tenure .01 .06 .11 .911 Study variables LMX -.63* .12 -5.28 .000 R2 .25
Dependent variable employee engagement (step 2)
Predictor B SE t p Control variables Gender -.09 .11 -.84 .402 Age .12 .06 1.96 .053 Tenure .01 .04 .12 .905 Study variables LMX .29* .07 3.89 .000 R2 .20
Dependent variable intentions to turnover (step 3 & 4)
Predictor B SE t p Control variables Gender -.00 .15 -.01 .993 Age -.00 .09 -.02 .987 Tenure .01 .06 .19 .849 Study variables LMX -.41* .11 -3.58 .000 Employee engagement -.77* .15 -5.15 .000 R2 .42
Notes: N = 95; B = unstandardized regression coefficient; SE = standard error; * p < .001.
the stronger the relation between LMX and engagement. Thus, the results assume a negative impact of autonomy and do not provide support for the claim of hypothesis 4, which is therefore rejected.
TABLE 5
Autonomy Moderating Effect Analysis Dependent variable employee engagement
Predictor B SE T p Control variables Gender -.07 .10 -.68 .497 Age .09 .06 1.49 .140 Tenure .02 .04 .54 .591 Study variables LMX .17* .08 2.11 .038 Autonomy .14 .08 1.82 .071 Interaction LMX*Autonomy -.18* .08 -2.38 .020 R2 .29 R2 change interaction .05
Relationship between LMX and employee engagement at low, average and high values of autonomy
Autonomy Effect SE t p 95% CI Low .30** .08 3.61 .001 [.134,.461] Average .17* .08 2.11 .038 [.010,.327] High .04 .11 .36 .721 [-.178,.256] Notes: N = 95; B = unstandardized regression coefficient; SE = standard error; CI = Confidence Interval; * p < .05; ** p < .001.
FIGURE 2
Relationship Between LMX and Employee Engagement for High, Average, and Low Level Autonomy (A)
Lastly, hypothesis 5 predicted a moderated mediation effect whereby the indirect relationship between LMX and intentions to turnover, as mediated by employee engagement, would be moderated by autonomy. To test the fifth hypothesis, model 7 of Hayes’ (2013) Process Macro was used. Firstly, table 6 presents the effects, standard errors, and significance value of the conditional indirect effects across low, average, and high levels of autonomy. It is demonstrated that the indirect effect in the low autonomy condition (indirect effect -.23, SE = .09, 95% CI [-.404,-.020]) was significant and stronger than the indirect effect in the average and high autonomy condition (average autonomy: indirect effect -.13, SE = .07, 95% CI [-.271,.007]; high autonomy: indirect effect -.03, SE = .09, 95% CI [-.230,.138]) which were both not significant. The finding of the low autonomy condition is consistent with the findings for hypothesis 4. That is, other than proposed, a negative influence of the moderator was found. Still, average and one standard deviation above average conditions of autonomy proved to be not significant. Moreover, a significant moderated mediation is denoted by a 95% confidence interval (CI) that does not include, or overlap, with zero (Hayes, 2015). Yet, the results show that the proposed overall moderated mediation is not significant (index: .14, SE = .09, 95% CI [-.089,.263]). Hence, in the total moderated mediation model zero is a valid option. In conclusion, this suggests that the indirect effect of LMX and intentions to turnover through employee engagement does not significantly differs based on the level of autonomy. Therefore, no support is found for hypothesis 5.
TABLE 6
Relationship between LMX and intentions to turnover through employee engagement at low, average and high values of autonomy
Autonomy Effect SE 95% CI Low -.23 .09 [-.404,-.020] Average -.13 .07 [-.271,.007] High -.03 .09 [-.230,.138]
Notes: N = 95; B = unstandardized regression coefficient; SE = standard error; CI = Confidence Interval; * p < .05; ** p < .001; based on 5000 bootstrap samples.
DISCUSSION
support the proposed predictive link between LMX and employee engagement. Hence, the relationship between supervisors and subordinates is to be regarded as a critical factor in engaging an organizations workforce. Furthermore, support was found for the hypothesis that employee engagement was negatively related to employees’ intentions to turnover. Accordingly, this study extends previous research in the areas of turnover, LMX and employee engagement by endorsing the engagement perspective whereby employee engagement mediates the relation between its antecedents and outcomes. The results of the current research provide support for this mediating engagement perspective, though a partial mediation. Finally, and other than expected, LMX was positively related to employee engagement under low levels of autonomy. Under high levels of autonomy, this connection was unrelated.
Theoretical implications
This research advances and contributes to existing literature in several ways. Firstly, and in line with a call for further research into antecedents of engagement (Wollard & Shuck, 2011), the results showed that leader-member exchange can be accounted for as an important predictor of employee engagement. Especially since thus far, surprisingly, scholars have rarely examined the role of middle managers as an antecedent of enhancing employee engagement (Bailey et al., 2017). The results of this study, thereby, do not merely add to our understanding of antecedents of the concept of engagement, they moreover confirm that supervisors play a significant role in this. So, job resources located at the level of interpersonal relations as put forward by the JD-R model (Bakker & Demerouti, 2007), can be held to affect staff’s engagement levels. In order to encourage staff to invest more energy and dedication into their jobs, middle managers in specific would do well to invest in the relationships with their direct reports.
Thus, this study contributes to research on engagement by providing additional empirical evidence for the relationship between employee engagement and employees’ intentions to turnover.
Hypothesis 5 proposed a moderated mediation model to signify a pathway linking LMX to intentions to turnover. Even though no significant evidence for this overall conceptualization was found, separate analyses of the mediating and moderating effects still resulted in several notable contributions.
Explicitly, it was demonstrated that employee engagement significantly partly mediated the relationship between LMX and intentions to turnover. Thereby confirming and adding to the growing body of literature on engagement that propose such mediation effects (e.g. Rich et al., 2010; Saks, 2006). This engagement perspective also signifies that in studying organizational and individual work-related outcomes, employee engagement is a fundamental concept to consider and embrace in research designs. This, in turn, also leads to another important contribution of this study. By means of a meta-analysis review, Dulebohn, Bommer, Liden, Brouer, and Ferris (2012) namely found that to date most LMX literature conceptualize LMX as acting as a direct precursor of outcome variables like job performance and organizational commitment. Including LMX in a model of intentions to turnover, nevertheless, showed that this relation is at least partly mediated by employee engagement. The latter, therefore, provides a valuable concept to include in LMX research endeavors.
the level of engagement of individuals that experience a high level of autonomy in performing their work is not dependent on the quality of the relationship with their direct manager. Clearly, when employees are given the opportunity to set their own goals and way of working, they implicitly demonstrate desirable behaviors beyond what was originally targeted (Ludwig & Frazier, 2012).
Practical implications
In addition to the theoretical contribution and implications, the results of this study also have some practical implications. To start with, results showed that LMX is a significant predictor of employee engagement and intentions to turnover. Hence, organizations that wish to retain staff and enhance employee engagement should focus on building and maintaining high quality relationships between supervisors and their subordinates. It stems from this that an important role herein is ascribed to middle management’s behavior and actions as they are lined up to directly manage the workforce. One way that organizations can stimulate high quality leader-member relationships is by encouraging and training middle managers in e.g. reward behavior, providing feedback, setting clear performance expectations (Dulebohn et al., 2012), active listening skills, and techniques for exchanging resources (Graen, Scandura, & Graen, 1986) towards direct reports. That way managers are better equipped to actively work on the quality of LMX relationships. This is consistent with the tendency that, in seeking competitive advantage, organizations should urge middle managers to deploy and focus more on soft skills in managing their employees (Ellinger, Ellinger, & Keller, 2003; Stoker, 2006).
provide valuable insights, a lot of managers fail to act upon its results (Sanchez, 2007). In the sense that nearly 80 percent of the effort involved in employee surveys is expended in the follow-up phase (Wiley, 2012), organization should see to it that follow-up actions are taken. Including HR in the roll-out but moreover in the follow-up phase would be advisable here, since middle managers might not consider it to be part of their core tasks (yet) and lack the knowledge and skills to facilitate such a process (Sanchez, 2007). Furthermore, putting KPI’s in place not just on survey outcomes, or merely on response rate, but rather on follow-up measures can make such an effort more fruitful and valuable. The insights of an engagement survey could, subsequently, be used to draw up job designs and measures to enhance the level of employee engagement.
Under low levels of job autonomy, LMX and employee engagement were found to be significantly, positively related. Thus, being a job design aspect, autonomy is a factor to consider by managers. This implies that in jobs that do not allow for high degrees of autonomy, the quality of LMX becomes a more decisive factor in engaging employees. This also goes for departments and jobs where a drop or decrease of autonomy is anticipated for, for example as a consequence of standardization and efficiency measures. Hence, for employees in those situations, supervisors should invest in and strive for paying more attention to interrelationship job resources like LMX. One resulting benefit of this could be, for example, that middle managers gain more insights in an employee’s thoughts and ambitions. This could then in turn be a trigger to job rotation whereby a better fit of employees and jobs can be achieved, leading workers to have more positive attitudes, intentions, and behaviors (Saks, 2006).
As a conclusion it can be said that if organizations pay little or no attention to LMX and autonomy in crafting and filling in the job design, they are at risk of losing highly qualified talents and good personnel. Next to the fact that losing good personnel as such is a negative event, more negative consequences result from it. That is, high turnover rates lead to high costs associated with recruitment, selection, and training of new employees (Kim & Koo, 2017; Reina et al., 2018). Putting effort into improving the quality of LMX is therefore important for employee retention as well as employee engagement.
Limitations and future research
focusing on creating a broader and sounder base of evidence for the triggering role of LMX on employee engagement and intentions to turnover across industries and countries. That way substantial and more generalizable conclusions can be drawn on their interrelatedness. Building on that direction, future research would furthermore be served well by discovering tactics and methods that managers in organizations can apply to improve the quality of their relationship with subordinates. Potential tactics managers could attain to might be found in areas such as rewards behavior, leader extraversion, trust, likeability (Dulebohn et al., 2012).
A second limitation is that this study relied on cross-sectional design which limits causal inference for the associations found. Futures studies are thus encouraged to adopt a longitudinal research design to address the issue of causality. Especially organizations that wish to train managers to enhance the quality of LMX, would provide for an interesting study environment. That is, a longitudinal design will not merely provide a sounder understanding of the concepts under study, it will moreover give insights into the effectiveness of such organizational trainings. Particularly since LMX relationships develop over time (Volmer et al., 2012).
Thirdly, the results and concurrent analyses are based on self-report data which could imply a limitation for the result’s validity. By using well established scales and survey items and by guaranteeing confidentiality and anonymity of responses this limitation was mitigated in part (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). Future research would nonetheless be served well by including supervisors and, potentially, peer workers in the research design, or by e.g. using observations in measuring LMX.
Lastly, the results of this paper showed that the relation between LMX and intentions to turnover is at least partly mediated by employee engagement. Another fruitful future LMX research endeavor on antecedents and consequence would therefore benefit from including employee engagement in the equation. This also goes for research on employee engagement, where including middle managers and their relationships with subordinates as a predictor of engagement is a worthwhile undertaking.
All the suggested directions for future research will shed more clarity on the validity of the conceptual model as tested in this study and thereby on the role of LMX and autonomy in creating employee engagement and intentions to turnover.
Conclusions
argued and showed that the relationship between middle managers and their subordinates is to be considered as a significant predictor of employee engagement and consequently intentions to turnover. The results contribute to the void of a lack of research into antecedents of employee engagement and indicates that organizations should pay attention to the quality of LMX. That way costs associated with turnover can be reduced and retention of talented professionals can be ameliorated. The implications also include the need for middle managers to consider and be trained in ways of improving the quality of LMX relationships. Especially in situations of lower job autonomy this seems to be a worthwhile path to pursue. In the end, as the old adage says: “people quit their bosses, not their jobs”.
REFERENCES
Agarwal, U. A., Datta, S., Blake-Beard, S., Bhargava, S. 2012. Linking LMX, innovative work behaviour and turnover intentions: The mediating role of work engagement. Career Development International, 17: 208-230.
Aon Hewitt. 2011. Managers: Your Strongest (or Weakest) Link in Driving Employee Engagement? Paper on the results of the annual Aon Hewitt Manager Survey.
Aubé, C., Rousseau, V., Morin, E.M. 2007. Perceived organizational support and organizational commitment: The moderating effect of locus of control and work autonomy. Journal of Managerial Psychology, 22(5): 479-495.
Avery, D.R., McKay, P.F., Wilson, D.C. 2007. Engaging the Aging Workforce: The Relationship Between Perceived Age Similarity, Satisfaction With Coworkers, and Employee Engagement. Journal of Applied Psychology, 92(6): 1542-1556.
Bailey, C., Madden, A., Alfes, K., Fletcher, L. 2017. The Meaning, Antecedents and Outcomes of Employee Engagement: A Narrative Synthesis. International Journal of Management Reviews, 19: 31-53.
Bakker, A.B., Demerouti, E. 2007. The job demands–resources model: State of the art. Journal of Managerial Psychology, 22: 309−328.
Belgio, E. 2018. Do Rewards and Recognition Drive Engagement? A Quantitative Study Assessing the Impact of Monetary and Nonmonetary Rewards on Employee Engagement. Dissertation Abstracts International Section A: Humanities and Social Sciences, 78(8). Byrne, Z.S., Hayes, T.L., Holcombe, K.J. 2017. Employee Engagement Using the Federal
Employee Viewpoint Survey. Public Personnel Management, 46(4): 368-390.
Cheng, J., Lu, K., Chang, Y., Johnstone, S. 2013. Voice behavior and work engagement: the moderating role of supervisor-attributed motives. Asia Pacific Journal of Human Resources, 51: 81-102.
Christian, M.S., Garza, A.S., Slaughter, J.E. 2011. Work engagement: A quantitative review and test of its relations with and contextual performance. Personnel Psychology, 64: 89-136.
Colarelli, S.M. 1984. Methods of communication and mediating processes in realistic job previews. Journal of Applied Psychology, 69: 633-42.
Dalal, R.S., Baysinger, M., Brummel, B.J., LeBreton, J.M. 2012. The Relative Importance of Employee Engagement, Other Job Attitudes, and Trait Affect as Predictors of Job Performance. Journal of Applied Social Psychology, 42: 295-325.
Dansereau, F., Graen, G.B., Haga, W. 1975. A vertical dyad linkage approach to leadership in formal organizations. Organizational Behavior and Human Performance, 13: 46-78. DeConinck, J.B. 2009. The effect of leader–member exchange on turnover among retail buyers.
Journal of Business Research, 62(11): 1081-1086.
Dulebohn, J.H., Bommer, W.H., Liden, R.C., Brouer, R.L., Ferris, G.R. 2012. A Meta-Analysis of Antecedents and Consequences of Leader-Member Exchange: Integrating the Past With an Eye Toward the Future. Journal of Management, 38(6): 1715-1759.
Ellinger, A.D., Ellinger, A.E., Keller, S.B. 2003. Supervisory coaching behavior, employee satisfaction, and warehouse employee performance: a dyadic perspective in the distribution industry. Human Resource Development Quarterly, 14(4): 435-458.
Farr-Wharton, R., Brunetto, Y. & Shacklock, K. 2011. Professionals’ supervisor–subordinate relationships, autonomy and commitment in Australia: a leader–member exchange theory perspective. The International Journal of Human Resource Management. 22(17): 3496-3512.
Graen, G., Uhl-Bien, M. 1995. Relationship-Based Approach to Leadership: Development of Leader–Member Exchange (LMX) Theory of Theory Over 25 Years: Applying a Multi-Level Multi-Domain Perspective. Leadership Quarterly, 6(2): 219-247.
Graen, G. B., Scandura, T. A., & Graen, M. R. 1986. A field experimental test of the moderating effects of growth need strength on productivity. Journal of Applied Psychology, 71(3), 484-491.
Gruman, J.A., Saks, A.M. 2011. Performance management and employee engagement. Human Resource Management Review, 21: 123-136.
Hackman, J. R., Oldham, G. R. 1975. Development of the Job Diagnostic Survey. Journal of Applied Psychology, 60: 159-170.
Hallberg, U., Schaufeli, W.B. 2006. ‘Same same’ but different? Can work engagement be discriminated from job involvement and organizational commitment? European Psychologist, 11(2): 119-127.
Harter, J.K., Schmidt, F.L., Hayes, T.L. 2002. Business-unit-level relationship between employee satisfaction, employee engagement, and business outcomes: A meta-analysis. Journal of Applied Psychology, 87: 268-279.
Hayes, A. F. 2015. An index and test of linear moderated mediation. Multivariate Behavioral Research, 50: 1-22.
Hayes, A. F. 2013. Model templates for PROCESS for SPSS and SAS. Retrieved June 2018. Hom, P.W., Ellis, A., Roberson, L. 2008. Challenging conventional wisdom about who quits:
Revelations about employee turnover from corporate America. Journal of Applied Psychology, 93: 1-34.
Kahn, W. 1990. Psychological conditions of personal engagement and disengagement at work. Academy of Management Journal, 33: 692-724.
Ketter, P. 2008. What’s the big deal about employee engagement? T+D, 62(2): 44-49.
Kim, M.S., Koo, D.W. 2017. Linking LMX, engagement, innovative behavior, and job performance in hotel employees. International Journal of Contemporary Hospitality Management, 29(12): 3044-3062.
Konrad, A.M. 2006. Engaging employees through high-involvement work practices. Ivey Business Journal, #9B06TB08: 1-6.
Ludwig, T.D., Frazier, C.B. 2012. Employee engagement and organizational behavior management. Journal of Organizational Behavior Management, 32(1): 75-82.
Macey, W.H., Schneider, B. 2008. The meaning of employee engagement. Industrial and Organizational Psychology, 1: 3-30.
Maslach, C., Schaufeli, W.B., Leiter, M.P. 2001. Job burnout. Annual Review of Psychology, 52: 397-422.
Menguc, B., Auh, S., Fisher, M., Haddad, A. 2013. To be engaged or not to be engaged: the antecedents and consequences of service employee engagement. Journal of Business Research, 66: 2163-2170.
Morgeson, F.P., Humphrey, S.E. 2006. The Work Design Questionnaire (WDQ): Developing and Validating a Comprehensive Measure for Assessing Job Design and the Nature of Work. Journal of Applied Psychology, 91(6): 1321-1339.
Pitt-Catsouphes, M., Matz-Costa, C. 2008. The multi-generational workforce: workplace flexibility and engagement. Community, Work & Family, 11(2): 215-229.
Podsakoff, P.M., MacKenzie, S.B., Lee, J., Podsakoff, N.P. 2003. Common method biases in behavioral research: a critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5): 879-903.
Reina, C.S., Rogers, K.M., Peterson, S.J., Byron, K., Hom, P.W. 2018. Quitting the Boss? The Role of Manager Influence Tactics and Employee Emotional Engagement in Voluntary Turnover. Journal of Leadership & Organizational Studies, 25(1): 5-18.
Rich, B.L., Lepine, J.A., Crawford, E.R. 2010. Job engagement: Antecedents and effects on job performance. Academy of Management Journal, 53: 617-635.
Saks, A.M. 2006. Antecedents and consequences of employee engagement. Journal of Managerial Psychology, 21: 600-619.
Sanchez, P.M. 2007. The employee survey: more than asking questions. Journal of Business Strategy, 28(2): 48-56.
Sarker, S.J., Crossman, A., Chinmeteepituck, P. 2003. The relationships of age and length of service with job satisfaction: an examination of hotel employees in Thailand. Journal of Managerial Psychology, 18(7): 745-758.
Shuck, B., Reio, T.G, Rocco, T.S. 2011. Employee engagement: an examination of antecedent and outcome variables. Human Resource Development International, 14(4): 427-445. Stoker, J.I. 2006. Leading middle management: consequences of organisational changes for
tasks and behaviours of middle managers. Journal of General Management, 32(1): 31-42.
Thornhill, A., Saunders, M.N.K. 1998. What if line managers don’t realize they’re responsible for HR? Personnel Review, 27(6): 460-476.
Volmer, J., Spurk, D., Niessen, C. 2012. Leader-member exchange (LMX), job autonomy, and creative work involvement. The Leadership Quarterly, 23(3): 456-465.
Wiley, J.W. 2012. Six things you need to know about strategic employee surveys. People & Strategy, 35(1): 16-23.
Wollard, K.K., Shuck, B. 2011. Antecedents to employee engagement: A structured review of the literature. Advances in Developing Human Resources, 13(4): 429-446.
APPENDIX A: SURVEY MEASURES AND ITEMS
Employee Engagement
Rich et al. (2010)
1. I am enthusiastic in my job. 2. I feel energetic at my job. 3. I am interested in my job. 4. I am proud of my job. 5. I feel positive about my job. 6. I am excited about my job.
Items measured on a scale ranging from (1) strongly disagree to (5) strongly agree.
Intentions to Turnover
Colarelli, S.M. (1984)
1. I frequently think of quitting my job.
2. I am planning to search for a new job during the next 12 months.
3. If I have my own way, I will be working for this organization one year from now (R). Leiter, Spence Laschinger, Day, and Gilin Oore (2011)
4. I have been actively looking for other jobs.
Items measured on a scale ranging from (1) strongly disagree to (5) strongly agree.
Leader Member Exchange
DeConinck, J.B. (2009)
1. My immediate supervisor understands my problems and needs. 2. My immediate supervisor recognizes my potential.
3. My immediate supervisor has enough confidence in me that he or she would defend and justify my decisions if I were not present to do so.
4. I usually know where I stand with my immediate supervisor.
5. My working relationship with my immediate supervisor is extremely effective.
Autonomy
Morgeson and Humphrey (2006)
1. The job gives me a chance to use my personal initiative or judgment in carrying out the work.
2. The job allows me to make a lot of decisions on my own.
3. The job provides me with significant autonomy in making decisions.
4. The job allows me to make decisions about what methods I use to complete my work. 5. The job gives me considerable opportunity for independence and freedom in how I do
the work.
6. The job allows me to decide on my own how to go about doing my work.
