Abstract
Theories of perceived overqualification have tended to focus on employees’ job-related responses to account for effects on performance. We offer an alternative perspective and theorize that perceived overqualification could influence work performance through a relational mechanism. We propose that relational skills, in the form of interpersonal influence of overqualified employees, determine their tendency to experience social acceptance and, thus, engage in positive work-related behaviors. We tested this relational model across two studies using time-lagged, multisource data. In Study 1, the results indicated that for employees high on interpersonal influence, perceived overqualification was positively related to self-reported social acceptance, whereas for employees low on interpersonal influence, the relationship was negative. Social acceptance, in turn, was positively related to in-role job performance, interpersonal altruism, and team member proactivity evaluated by supervisors. In Study 2, we focused on peer-reported social acceptance and found that the indirect relationships between perceived overqualification and supervisor-reported behavioral outcomes via social acceptance were negative when interpersonal influence was low and nonsignificant when interpersonal influence was high. The implications of the general findings are discussed.
Overqualification, or the situation where employees possess qualifications such as education, experiences, and/or skills exceeding their job requirements (Erdogan, Bauer, Peiró, & Truxillo, 2011b), has become a prevalent phenomenon across different countries (Büchel & Mertens, 2004; Erdogan & Bauer, 2009; Sadava, O’Connor, & McCreary, 2000). Recently, considerable attention from organizational scholars has been devoted to studying employees who consider themselves overqualified (e.g., Feldman, Leana, & Bolino, 2002; Maynard, Joseph, & Maynard, 2006; Maynard & Parfyonova, 2013). For example, research has consistently found that employees who feel overqualified are more likely to develop negative organizational attitudes such as intention to quit and job dissatisfaction, experience poor well-being, and engage in counterproductive behaviors (Erdogan & Bauer, 2009; Johnson & Johnson, 1997, 2000; Liu, Luksyte, Zhou, Shi, & Wang, 2015; Luksyte, Spitzmueller, & Maynard, 2011). Moreover, perceived overqualification has been found to have implications for employees’ performance (Bolino & Feldman, 2000; Erdogan & Bauer, 2009; Hu, Erdogan, Bauer, Jiang, Liu, & Li, 2015).
To date, a dominant perspective to understand perceived overqualification focused on job-related responses due to the discrepancy between possessed qualifications and job requirements (Erdogan & Bauer, 2009). Researchers have contended that because perceived overqualification represents a perception of underutilization of abilities and skills, when overqualified employees compare their own qualifications with the required qualifications, they tend to experience a sense of deprivation (Feldman et al., 2002), injustice (Liu & Wang, 2012), and misfit (Maynard et al., 2006), leading to subsequent behavioral outcomes.
A neglected possibility in the extant overqualification literature is that the sense of superiority embedded in perceptions of overqualification can evoke a relational mechanism to influence employees’ performance at work. In brief, because perceived overqualification denotes surplus job capacity employees possess for carrying out their current jobs, overqualified employees may hold “a positive view of their job competence” (Zhang, Law, & Lin, 2016: 62) and sense of agency (Liu & Wang, 2012). Drawing on the literature on self-perceptions and social interactions (Baumeister, Campbell, Krueger, & Vohs, 2003; Paulhus, 1998; Swann, Chang-Schneider, & McClarty, 2007), we propose that the sense of superiority embedded in perceived overqualification can bring different social implications in interactions with others at work. On one hand, overqualified employees could be admired and respected by their peers because of the skills and resources they possess and the potential value they bring to a workgroup (Erdogan et al., 2011b). Alternatively, they may be rejected and isolated by their coworkers by actions conveying their felt superiority (cf. Tai, Narayanan, & McAllister, 2012). In other words, employee perceptions of overqualification are likely to result in different levels of social acceptance by coworkers and, thus, performance at work (Avery, McKay, & Wilson, 2007; Hodson, 1997; Kahn, 1990). These relational implications cannot be captured by the job-focused perspective and suggest a necessity to bring an alternative framework to understand the impact of overqualification on work outcomes.
Drawing on social influence theories (Higgins, Judge, & Ferris, 2003; Levy, Collins, & Nail, 1998), we suggest that the direction of such relational impact depends on employees’ interpersonal influence, or the capability of appropriately adapting and calibrating their behavior to elicit the desired responses from those around them (Ferris, Davidson, & Perrewe, 2005). High interpersonal influence should enable overqualified employees to display positive social behaviors in interactions with coworkers and be seen as competent and likeable, resulting in higher levels of social acceptance. High social acceptance, in turn, will motivate them to perform well and engage in more positive work behaviors. In contrast, employees who feel overqualified and are low on interpersonal influence are likely to communicate their sense of entitlement and felt superiority to coworkers, resulting in lower levels of social acceptance. Consequently, these employees may feel demotivated and engage in less positive work behaviors. To provide a stronger test of our model, we examine different types of work performance, including in-role performance and affiliation-oriented (interpersonal altruism; Organ, 1988) and change-oriented (team member proactivity; Griffin, Neal, & Parker, 2007) organizational citizenship behavior (MacKenzie, Podsakoff, & Podsakoff, 2011; Van Dyne & LePine, 1998). Figure 1 presents our proposed research model.
The Proposed Model of the Current Research
Our examination of the relational implications of perceived overqualification offers three specific contributions to the literature. First, to date, the nature of the relationship between perceived overqualification and employees’ relations with their coworkers has been neglected. This is an important omission to address as the scholarly opinion seems divided in this regard. For example, Sierra (2011) contends that feelings of overqualification may hamper relations with coworkers as a result of reduced cohesion and collaborative behaviors. However, Erdogan et al. (2011b) recognize the possibility that those who feel overqualified may serve as mentors to others, potentially contributing positively to relations with coworkers. By establishing social acceptance as a relational mediator, we answer the call for exploring different mechanisms behind the overqualification-performance relationship and the call for examining the social implications of overqualification (Erdogan et al., 2011b).
Second, our research examines the role of an individual skill (i.e., interpersonal influence) in moderating the effects of overqualification, in contrast to the previous focus on contextual moderators such as empowerment (Erdogan & Bauer, 2009), job autonomy (Wu, Luksyte, & Parker, 2015), and peer overqualification (Hu et al., 2015). Our consideration of interpersonal influence adds to a comprehensive understanding of when perceived overqualification is beneficial or harmful for work performance, answering the call for “more research into the moderators in the overqualification-performance relationship” (Erdogan & Bauer, 2009: 563). Furthermore, this examination allows us to take a person-centric approach that views employees as active agents who can shape their own work experiences (Weiss & Rupp, 2011) and increase our understanding of how employees’ characteristics intersect with their feelings of overqualification in influencing their integration in their workplace and levels of effectiveness.
Finally, our research incorporates multiple behavioral outcomes that capture different aspects of work performance. With a few exceptions (Chen, 2009; Hu et al., 2015; Zhang et al., 2016), the majority of research on overqualification has examined its implications for in-role performance, largely ignoring extrarole performance (cf. Hu et al., 2015), or “behavior that is not strictly role prescribed but contributes to organizational effectiveness” (LePine & Van Dyne, 2001: 326). More research on how to stimulate extrarole behaviors is necessary, because “given the discretionary nature of extra task behaviors, this may be precisely the area in which the effects of overqualification (specially perceived overqualification) on performance are greatest” (Bashshur, Hernández, & Peiró, 2011: 196). Our research stands poised to bring additional insights in this neglected area of research.
Hypothesis Development
Overqualification can be measured both by objective standards, such as overeducation (e.g., Rubb, 2009), and by perceptual measures. The two measurement methods are distinct and may affect different outcomes (Maltarich, Reilly, & Nyberg, 2011). For example, objective overqualification ideally involves assessing the discrepancy between job demands and employee skills by an outside party and is regarded as a better predictor of job mobility and recruiter reactions to the particular individual (Maltarich et al., 2011). In contrast, subjective overqualification is regarded as a more proximal predictor of employees’ job-related cognitions and behaviors (Liu & Wang, 2012). Given our focus on employees’ experience of their social acceptance in the group, we follow previous research (e.g., Erdogan & Bauer, 2009; Feldman et al., 2002; Liu et al., 2015) and develop a relational model based on perceived overqualification (i.e., individuals’ assessment of the directional mismatch between their abilities and job opportunities to perform; Maynard et al., 2006).
Perceived Overqualification and Social Acceptance: The Moderating Role of Interpersonal Influence
Individuals’ self-perceptions can shape their social interactions and interpersonal relationships (Baumeister et al., 2003; Paulhus, 1998; Swann et al., 2007) as people are motivated to reveal their self-views and behave consistent with them (Korman, 1970). A positive self-perception can be a mixed blessing that induces both positive and negative impressions in the eyes of others (cf. Paulhus, 1998). Perceived overqualification, which inherently involves a positive self-perception regarding one’s skills, knowledge, and abilities, can influence an individual’s social interactions. Presumably, overqualified employees are capable of accomplishing their work more effectively and efficiently (e.g., Erdogan & Bauer, 2009; Holtom, Lee, & Tidd, 2002). As a result, overqualified employees may present themselves in a way that convinces their coworkers that they are competent group members who possess superior skills and knowledge that would benefit the group and, thus, enjoy high social acceptance. In contrast, the negative emotions and feelings associated with being overqualified (Liu et al., 2015) could negatively affect the interpersonal interactions at work. Overqualified employees may interact with coworkers in an abrasive manner that leads to the typically undesirable social image that they are conceited and act superior to others.
To understand when perceived overqualification leads to positive or negative relational consequences, we follow social influence theories (Ferris, Treadway, Perrewé, Brouer, Douglas, & Lux, 2007; Higgins et al., 2003; Levy et al., 1998) and propose that employees’ interpersonal influence is key to unpacking this puzzle. Interpersonal influence is a critical dimension of political and relational skills and refers to “a subtle and convincing personal style that exerts a powerful influence on those around” (Ferris, Treadway, et al., 2005: 129). Individuals high on interpersonal influence are capable of great flexibility in interpersonal interactions by appropriately adapting and calibrating their behavior to different situations and eliciting desired responses from others. They are motivated and able to appear pleasant and productive to others and are masters at controlling their environment (Munyon, Summers, Thompson, & Ferris, 2015). Interpersonal influence is particularly relevant to the potential social implications of overqualification because it is likely to shape employees’ image in the eyes of their coworkers and contribute to feelings of acceptance by coworkers.
When overqualified employees are high on interpersonal influence, they are more likely to utilize excessive job-related resources in an interpersonally beneficial way as they are aware of the possible gap between self-perception and other-perception and capable of reducing it (Ferris, Davidson, & Perrewe, 2005). Specifically, because people high on interpersonal influence are effective communicators with a convincing style (Ferris et al., 2007), they should be able to monitor their interactions more effectively and utilize their skills without acting superior to their peers when interacting with them. Moreover, although these employees may feel entitled and experience typical negative feelings associated with overqualification, such as anger (Liu et al., 2015), interpersonal influence enables them to calibrate their self-expression in interpersonal interactions, avoiding creating unfavorable social comparisons between them and coworkers. For example, they perhaps would not appear condescending to their coworkers by emphasizing their superior skills and demeaning coworkers. Thus, when interpersonal influence is high, overqualified employees are able to effectively leverage excessive qualifications as resources to create a desired social image among coworkers that they are likeable, making people enjoy associating with them.
In stark contrast, when overqualified employees are low on interpersonal influence, they are less sensitive to the discrepancy between their self-view and others’ perception of them. Employees with low interpersonal influence tend to be less capable of effectively communicating and building a good relationship and rapport with others (Ferris, Davidson, & Perrewe, 2005; Ferris et al., 2007; Ferris, Treadway, et al., 2005). Consequently, when they also think of themselves as highly competent, their coworkers may instead see them as conceited. It is also possible that they may make their coworkers feel uncomfortable (perhaps accidentally) by communicating their feelings of entitlement to coworkers. As a result of their inability to effectively calibrate their styles and behaviors, the relationships with coworkers may be further plagued by ostensible revelation of their dissatisfaction, anger, and frustration. In short, overqualified employees low on interpersonal influence are less capable of leveraging the possession of surplus skills and knowledge. Instead, they may create an undesired social image among coworkers that they are arrogant and unlikeable, leading to social rejection. Our reasoning suggests that high perceived overqualification and low interpersonal influence interactively lead to low social acceptance.
Taken together, we propose that interpersonal influence shapes the effectiveness of social interactions of overqualified employees with their coworkers such that overqualification will be positively related to social acceptance among those high on interpersonal influence but overqualification will negatively relate to social acceptance among those low on interpersonal influence. Although no previous work has provided direct evidence for this proposition, social influence research has suggested that self-promotion behavior achieved the desired image of competent only for those with high interpersonal competence but led to the undesired image of conceited for those with low interpersonal competence (Turnley & Bolino, 2001), lending some indirect support to our research. On the basis of our reasoning, we suggest that the relationship between perceived overqualification and social acceptance will be contingent on the level of interpersonal influence.
Hypothesis 1: Interpersonal influence moderates the relationship between perceived overqualification and social acceptance such that this relationship is positive when interpersonal influence is high but negative when it is low.
Social Acceptance and Work Performance
Positive social relations with coworkers are a “rising motivational tide that lifts all boats toward higher levels of job dedication” (Chiaburu & Harrison, 2008: 1085). We further propose that social acceptance will promote a variety of work performance, including in-role performance, core task performance required by one’s job, and two forms of extrarole performance, interpersonal altruism and team member proactivity (MacKenzie et al., 2011; Van Dyne & LePine, 1998). Interpersonal altruism and team member proactivity are under the relative discretion of employees (Van Dyne & LePine, 1998) but are different in their ways to contribute at work. Interpersonal altruism is affiliation-oriented behavior that aims to maintain or enhance interpersonal relationships (MacKenzie et al., 2011), whereas team member proactivity is change-oriented behavior that aims to bring constructive changes to benefit a work unit and team members (Griffin et al., 2007).
There are good reasons to expect that higher social acceptance is related to better performance and more positive behaviors. In general, being accepted indicates high-quality relationships in the workplace, which can motivate employees to exert more effort into their tasks, reciprocate their coworkers, and contribute to the team (Stephens, Heaphy, & Dutton, 2012). Specifically, when overqualified employees are accepted or even admired by coworkers, they feel obliged to perform well and live up to the expectations of others. A sense of obligation is a strong driving force to make use of their abilities and skills, leading to better in-role job performance (e.g., Eisenberger, Armeli, Rexwinkel, Lynch, & Rhoades, 2001; Shanock & Eisenberger, 2006). At the same time, social acceptance may make employees believe that their teams and coworkers are supportive and are on their side and that they are one of the team. The norm of reciprocity (Gouldner, 1960) as well as social exchange theory (Blau, 1964) indicate that when individuals feel socially accepted and supported by their team members, they are most likely to be motivated to reciprocate their colleagues.
One likely way is to provide their coworkers with interpersonal help, or altruism, given the extra resources or bandwidth that being overqualified affords them (Erdogan & Bauer, 2009). This social exchange account has received extensive support in the literature on organizational citizenship behavior (e.g., Ang, Van Dyne, & Begley, 2003; Settoon, Bennett, & Liden, 1996). Employees may find other ways that have an even bigger impact to reciprocate. Proactive behaviors, such as job crafting (Wrzesniewski & Dutton, 2001), have been suggested to be a possible positive outcome of overqualification when employees are motivated to make contributions (Agut, Peiró, & Grau, 2009; Liu & Wang, 2012). It is highly likely that employees who feel accepted engage in proactive behavior that aims to benefit team members and the team. In line with our arguments, perceptions of team supportiveness have been found to be strongly related to team member proactivity (Griffin et al., 2007).
Hypothesis 2: Social acceptance is positively related to (a) in-role job performance, (b) altruism, and (c) team member proactivity.
A Moderated Mediation Model
Taken together, the above considerations sketch a complex picture of the relationships between perceived overqualification and work performance that suggests that social acceptance will mediate the associations between perceived overqualification and work performance and that the strength of these indirect relationships hinges on the level of interpersonal influence. Specifically, overqualified employees who possess high interpersonal influence are likely to be accepted by coworkers and experience high-quality social interactions. This experience, in turn, motivates employees to reciprocate their coworkers as well as the team to sustain positive social relationships by better performing their core tasks, helping coworkers, and engaging in team-focused proactive behavior. On the contrary, overqualified employees who do not have sufficient interpersonal influence are likely to suffer negative social interactions and be rejected by coworkers, which demotivates them to perform, to help, and to be proactive. In sum, employees’ interpersonal influence and, consequently, experienced relational standing are posited to play an important role in affecting their work performance.
To examine this mechanism as a whole, we therefore specify a first-stage moderated mediation model (Preacher, Rucker, & Hayes, 2007), which suggests that interpersonal influence will moderate the mediation effect of social acceptance on the association between perceived overqualification and work performance. This model integrates a relational mediator (i.e., social acceptance), a relational moderator (i.e., interpersonal influence), and relational outcomes (i.e., altruism and team member proactivity) into an overarching framework and well represents our proposed relational perspective of overqualification. Therefore, we hypothesize the following:
Hypothesis 3: The indirect effects of perceived overqualification on (a) in-role job performance, (b) altruism, and (c) team member proactivity via social acceptance are moderated by interpersonal influence such that these indirect associations are positive when interpersonal influence is high but negative when it is low.
We tested our model in two studies, with two separate samples from China. Overqualification is a global phenomenon that influences both developed and developing countries (Erdogan & Bauer, 2009). China is a country where overqualification is particularly prevalent, with over 80% of employees reporting that they feel overqualified (Randstad, 2012). Perceived overqualification has been studied in Chinese samples in previous research and shown to have important implications for employee attitudes and behaviors (e.g., Hu et al., 2015; Liu et al., 2015; Yang, Guan, Lai, She, & Lockwood, 2015; Zhang et al., 2016). Participants in both samples worked in teams, which makes these appropriate contexts to study the relational implications of perceived overqualification. In Study 1, we test our model using two sources of data (i.e., employees and their immediate supervisors) with social acceptance reported by employees themselves. In Study 2, we utilized three sources of data from employees, coworkers, and supervisors with social acceptance evaluated by coworkers.
Study 1
Method
Sample and Procedure
To test our hypotheses, we conducted a study in a state-owned information and technology (IT) company in Beijing, China. Employees were engineers who worked collaboratively in teams to set up and maintain IT systems for corporate clients. We collected data via surveys at three time points from both employees and their immediate supervisors to minimize common method variance concerns (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). At Time 1, we distributed a survey to all 290 employees working in the company and assessed perceived overqualification, interpersonal influence, and control variables. One month later at Time 2, employees completed a survey in which social acceptance was measured. One month after the second survey, at Time 3, we obtained in-role job performance, interpersonal altruism, and team member proactivity from the direct supervisor of each respondent. Participants were assured of confidentiality, and they were told that the data collected were to be used only for research purposes. Completed questionnaires were returned directly to research assistants on site.
A total of 194 complete, matched, and usable sets of questionnaires were obtained out of the 290 distributed questionnaires, yielding an overall response rate of 67%. The 194 employees were nested in 50 supervisors, 98% were male, 3% had a vocational school education (equivalent to high school), 25% had an associate degree, 68% received a college education, and 4% received a postgraduate education. The average age of respondents was 29 years old (SD = 4), and the average tenure was 5.65 years (SD = 4.35).
Measures
The three survey versions were translated into Chinese (i.e., Mandarin) using a back-translation procedure (Brislin, 1970). All measures were established scales and were evaluated on a 7-point scale ranging from 1 (strongly disagree) to 7 (strongly agree) unless otherwise specified.
Perceived overqualification
We used the nine-item measure by Maynard et al. (2006) to assess perceived overqualification. Sample items included “My job requires less education than I have” and “I have a lot of knowledge that I do not need in order to do my job” (α = .87).
Interpersonal influence
We used the four-item scale of the Political Skill Inventory (PSI) developed by Ferris, Treadway, et al. (2005) to measure interpersonal influence. This scale was chosen because it directly relates to the ability to appropriately adapt and calibrate one’s behavior when interacting with different people and, thus, accurately captures what we intended to measure. It is common practice in the literature to use only one or two dimensions of the PSI that correspond to the research interest (e.g., Baer, 2012; Brockner, Spreitzer, Mishra, Hochwarter, Pepper, & Weinberg, 2004; Ng & Feldman, 2010, 2011; Thompson, 2005). Research has also focused specifically on interpersonal influence to operationalize work-related relationship building skill and communication effectiveness (Fuller, Simmering, Marler, Cox, Bennett, & Cheramie, 2011). Sample items of interpersonal influence included “I am able to make most people feel comfortable and at ease around me” and “I am good at getting people to like me” (α = .86).
Social acceptance
We captured focal employees’ perceptions regarding the degree to which they are socially accepted by their coworkers using the eight-item scale of popularity, which is defined as “being generally accepted by one’s peers” (Scott & Judge, 2009: 21). Sample items were “I am liked by my coworkers” and “I am accepted by my coworkers” (α = .94).
In-role job performance
In-role job performance was measured using three items from Ashford and Black (1996). These three items focus mainly on one’s overall performance and have been validated in previous research (Wu, Parker, & de Jong, 2014). Sample items were “The overall performance of this employee is good” and “This employee has high-quality work performance” (α = .92).
Interpersonal altruism
Four items from Smith, Organ, and Near (1983) were used on a 7-point scale ranging from 1 (never) to 7 (always). These items focus on providing interpersonal help to coworkers. Sample items included “This employee helps coworkers who have been absent” and “This employee helps coworkers who have heavy workloads” (α = .92).
Team member proactivity
A three-item scale developed by Griffin et al. (2007) was utilized. These items were assessed on a 7-point scale (1 = never to 7 = always). Sample items included “This employee suggests ways to make the team more effective” and “This employee develops new and improved methods to help the team perform better” (α = .96).
Control variables
Because social acceptance was self-reported, to take into account the issue of authenticity in responses (Ellingson, Smith, & Sackett, 2001), we controlled for social desirability bias, which has been shown to affect self-perceptions (e.g., Kernis, 2003). We measured social desirability in responding using the five-item scale developed by Hays, Hayashi, and Stewart (1989) to reduce the time burden on respondents. Sample items included “I am always courteous even to people who are disagreeable” and “There have been occasions when I took advantage of someone.” The reliability coefficient for this measure employed in our study was .60, which is similar to reliability estimates of .62 to .66 found in previous research utilizing this measure (e.g., Williams, Pillai, Deptula, & Lowe, 2012). Following previous research on overqualification, we controlled for education and tenure in our analyses to ensure that observed overqualification effects do not simply serve as a proxy for education or tenure (e.g., Erdogan & Bauer, 2009; Luksyte et al., 2011). However, there was no change in the significance of the results if these demographic variables were not controlled for.
Data Analysis
Our data structure was nested (i.e., multiple subordinates reported to the same supervisor), therefore violating the assumption of independence of observations (Bickel, 2007). Because of this, we used multilevel methods—random intercept models—to test our hypotheses. The relationships were examined at the individual level, while taking into account the possible effect from the supervisor level. These analyses produce estimates comparable to unstandardized regression coefficients. All independent variables were mean centered prior to the analyses (Aiken & West, 1991). Selig and Preacher’s (2008) Monte Carlo method, which has been recommended by Preacher, Zyphur, and Zhang (2010), was used to estimate 95% confidence intervals (CIs) for the hypothesized mediated relationships to determine their significance. The use of such CIs is considered superior to traditional methods (e.g., the Sobel test) in examining (conditional) indirect relationships because it ameliorates power problems introduced by nonnormal sampling distributions of an indirect effect (Preacher et al., 2010). Following previous research using the same multilevel analytical method (e.g., Lam, Huang, & Chan, 2015; Walter, Lam, Van der Vegt, Huang, & Miao, 2015), we also reported R12, an indicator of proportions of explained variance in random intercept models (Bickel, 2007: 133). This statistic is comparable to the traditional effect size indicator (i.e., R2) in ordinary regression analysis and can be interpreted in a similar way (Bickel, 2007).
Results and Discussion of Study 1
Means, standard deviations, and correlations are presented in Table 1. We performed a series of confirmatory factor analyses to establish the discriminant validity of our measurement model. Because the subject-to-item ratio was below the recommended ratio of 10:1 (Bandalos, 2002), we randomly formed three parcels for two constructs with the most items (i.e., perceived overqualification and social acceptance) to make the ratio closer to the ideal one. As the measures of perceived overqualification and social acceptance we used are unidimensional in nature (Maynard et al., 2006; Scott & Judge, 2009) and the composite scores rather than individual items were used in hypotheses testing, this practice is not likely to cause biased estimates of relationships. The fit statistics of the hypothesized seven-factor model indicated acceptable fit: χ 2 = 440.30, p < .01; df = 254; comparative fit index (CFI) = .94; standardized root mean square residual (SRMR) = .07; root mean square error of approximation (RMSEA) = .06. This seven-factor model was significantly better than a six-factor model in which interpersonal influence and social acceptance were combined into one factor (Δχ2 = 261.81, p < .01; Δdf = 6) and a five-factor model in which all three behavioral outcome variables were combined into one factor (Δχ2 = 663.08, p < .01; Δdf = 11).
Correlations, Means, and Standard Deviations for Variables (Study 1)
p < .05.
p < .01.
Hypothesis 1 predicts that interpersonal influence moderates the relationship between perceived overqualification and social acceptance. As shown in Table 2, the cross product of perceived overqualification and interpersonal influence was positively associated with social acceptance (b = 0.16, p < .01) after considering the control variables and main effects. Figure 2 illustrates the form of this interaction by plotting the simple slopes at conditional values of interpersonal influence at 1 SD above and below its mean (Aiken & West, 1991). As predicted, the simple slope of the relationship between perceived overqualification and social acceptance was significant and positive under the condition of high interpersonal influence (b = 0.15, p < .05), whereas the simple slope was significant and negative under the condition of low interpersonal influence (b = −0.15, p < .05). Hence, Hypothesis 1 was supported.
Results for Moderation and Moderated Mediation Hypotheses (Study 1)
Note: The table shows unstandardized coefficients (with standard errors in parentheses) unless otherwise noted. CI = confidence interval.
p < .01.
The Interaction of Perceived Overqualification and Interpersonal Influence on Social Acceptance (Study 1)
Hypotheses 2a, 2b, and 2c predict that social acceptance is positively related to in-role job performance, interpersonal altruism, and team member proactivity. As Table 2 shows, after taking into account the effects of the control variable and the predictor (i.e., overqualification), we found that the associations of social acceptance with job performance (b = 0.38, p < .01), interpersonal altruism (b = 0.25, p < .01), and team member proactivity (b = 0.33, p < .01) were all positive and significant. Therefore, these hypotheses received support.
Hypotheses 3a, 3b, and 3c predict interpersonal skill to moderate the indirect effects of overqualification on work performance as transmitted by social acceptance. The results based on the Monte Carlo method showed that the indirect relationship between overqualification and job performance was positive and significant (effect = .06, 95% CI = [.01, .12]) when interpersonal skill was high and negative and significant (effect = −.06, 95% CI = [−.12, −.01]) when interpersonal skill was low, supporting Hypothesis 3a. The results also showed that the indirect relationship between overqualification and interpersonal altruism was positive and significant (effect = .04, 95% CI = [.01, .08]) when interpersonal skill was high and negative and significant (effect = −.04, 95% CI = [−.08, −.01]) when interpersonal skill was low, supporting Hypothesis 3b. Finally, Hypothesis 3c was supported with the results showing that the indirect relationship between overqualification and team member proactivity was positive and significant (effect = .05, 95% CI = [.01, .11]) when interpersonal skill was high and negative and significant (effect = −.05, 95% CI = [−.11, −.01]) when interpersonal skill was low.
Results of Study 1 provide support for our hypotheses that perceived overqualification influences various performance outcomes through a relational mechanism and employees’ interpersonal influence shapes the direction of such influence. Nevertheless, Study 1 was not without limitations. First, social acceptance was reported by employees themselves, and evaluations from coworkers would provide stronger evidence for the relational implications of perceived overqualification. Second, data in Study 1 were from one company in a single industry, and it is important to explore the generalizability of these findings. To bolster confidence in our theoretical model, we conducted a second study in which coworker ratings of social acceptance were employed and data from a different company in another industry were collected.
Study 2
A total of 216 employees out of 448 of a private language training company in a southern city of China were randomly selected to participate in this study. These employees worked in teams and were from different departments, including marketing, training and development, financing, and the human resources department. Three sets of electronic questionnaires were prepared for employees, their coworkers, and direct supervisors, respectively, and there was a 2-week time lag between the administration of each survey to help guard against common method variance (Podsakoff et al., 2003). At Time 1, employee questionnaires were distributed in which perceived overqualification, interpersonal influence, and control variables were measured. At Time 2, coworkers rated employees’ social acceptance. Each employee was rated by a different coworker who worked closely with him or her in the same team. At Time 3, employees’ immediate supervisors evaluated their in-role job performance, interpersonal altruism, and team member proactivity. Most supervisors evaluated multiple employees who directly reported to them. Participants were assured of confidentiality.
After matching the employee, coworker, and supervisor surveys, we obtained 204 sets of complete questionnaires with a response rate of 94%. We explained the purpose of the research to senior managers of the company before the data collection, and they were highly interested. They encouraged the employees to cooperate with us at the beginning and sent several e-mails during the data collection processes to remind them about the study. This full cooperation and assistance from the management resulted in a high response rate.
The 204 employees were nested in 31 supervisors, 83% were female, 11% had a high school education, 38% had an associate degree, 40% received a college education, and 11% received a postgraduate education. The average age of respondents was 26 years old (SD = 3), and the average tenure was 3.90 years (SD = 2.43).
Measures
As in Study 1, questionnaires were translated into Mandarin Chinese following the same procedure (Brislin, 1970), and all measures were evaluated on a 7-point scale ranging from 1 (strongly disagree) to 7 (strongly agree).
Study Variables
Except perceived overqualification, all variables were measured using the same scales as those in Study 1. In order to ensure a high response rate, we utilized a four-item scale from Johnson and Johnson (1996) to measure perceived overqualification. This scale has been frequently used in previous research and has been shown to have good validity (e.g., Erdogan & Bauer, 2009; Hu et al., 2015; Johnson & Johnson, 1997). All the measures showed acceptable reliabilities (α) in our data: .78 for perceived overqualification, .85 for interpersonal influence, .94 for social acceptance, .95 for in-role job performance, .92 for interpersonal altruism, and .96 for team member proactivity.
Control Variables
Consistent with Study 1, tenure and education were controlled for in analyses. As in Study 1, the significance of the results remained the same if these demographic variables were excluded, but we report the results with these controls.
Results and Discussion of Study 2
Table 3 presents means, standard deviations, and correlations among the study variables. We performed confirmatory factor analyses to establish the discriminant validity of our measurement model. Again, we randomly formed three parcels for social acceptance, the construct with the largest number of items, to maintain an appropriate subject-to-item ratio (Bandalos, 2002). All the other items were not parceled. The fit statistics of the hypothesized six-factor model indicated acceptable fit (χ2 = 248.80, p < .01; df = 174; CFI = .98; SRMR = .04; RMSEA = .05), significantly better than a five-factor model combining interpersonal influence and social acceptance (Δχ2 = 626.02, p < .01; Δdf = 5) and a four-factor model combining all three outcomes (Δχ 2 = 836.86, p < .01; Δdf = 9).
Correlations, Means, and Standard Deviations for Variables (Study 2)
p < .05.
p < .01.
Hypotheses testing results are presented in Table 4. As shown, the interaction between perceived overqualification and interpersonal influence on social acceptance was positive and significant (b = 0.15, p < .01) after the effects of tenure and education were controlled, consistent with Study 1. Figure 3 illustrates the interaction pattern (Aiken & West, 1991). As predicted, the simple slope was significant and negative under the condition of low interpersonal influence (b = −0.28, p < .01); however, the simple slope of the relationship between perceived overqualification and social acceptance was not significant under the condition of high interpersonal influence (b = 0, n.s.). Therefore, Hypothesis 1 was partially supported.
Results for Moderation and Moderated Mediation Hypotheses (Study 2)
Note: The table shows unstandardized coefficients (with standard errors in parentheses) unless otherwise noted. CI = confidence interval.
p < .05.
p < .01.
The Interaction of Perceived Overqualification and Interpersonal Influence on Social Acceptance (Study 2)
Furthermore, as Table 2 shows, social acceptance was positively and significantly related to job performance (b = 0.38, p < .01), interpersonal altruism (b = 0.26, p < .01), and team member proactivity (b = 0.26, p < .01). Therefore, Hypotheses 2a, 2b, and 2c received support. Finally, we tested the indirect effects of perceived overqualification on the three behavioral outcomes when interpersonal influence is low. Consistent with our expectation, the results showed that the indirect associations of overqualification with all three types of performance were negative and significant (effect = −.11, 95% CI = [−.19, −.04] for in-role job performance; effect = −.07, 95% CI = [−.14, −.02] for interpersonal altruism; and effect = −.07, 95% CI = [−.14, −.02] for team member proactivity) when interpersonal skill was low. However, the indirect effects of perceived overqualification under high levels of interpersonal influence were not significant for all three behavioral outcomes. Overall, the negative indirect effects proposed in Hypotheses 3a, 3b, and 3c were supported, whereas the positive indirect effects were not.
Taken together, Study 2 generally replicated the findings of Study 1 in terms of the significant interaction effect between perceived overqualification and interpersonal influence, the effects of social acceptance on all three outcomes, and the conditional indirect effects of perceived overqualification on these behavioral outcomes under low interpersonal influence. It provides a more stringent examination of the uniqueness of the relational mechanism (i.e., social acceptance) in channeling the interactive effect between perceived overqualification and interpersonal influence on different forms of performance.
However, the beneficial effects of perceived overqualification under high level of interpersonal influence were not replicated. This difference might be caused by differences in research contexts across the two studies. Specifically, Study 1 was conducted in a state-owned company, whereas Study 2 was in a private one. Seniority, or how long an employee has worked in an organization, plays an important role in resource allocation in the state-owned sector. Overqualified employees may not be seen as particularly threatening to coworkers in such a context, and those who master interpersonal skills may induce positive responses from their coworkers. However, in a private company, with reward allocation entirely merit based, overqualified employees are likely to leverage their skills and abilities to gain more resources and, thus, be considered a threat by their coworkers. Therefore, overqualification may have a stronger negative main effect on social acceptance, which is mitigated but not reversed by interpersonal influence. Another difference between the two studies was that the measurement of social acceptance was from employees’ own perspective in Study 1 but from coworkers’ perspective in Study 2. It is plausible that overqualified employees may have created negative reactions on coworkers without being aware of them, explaining the more negative implications on coworker rated social acceptance. Of course, these arguments are speculative and the interaction between overqualification and interpersonal influence merits further scrutiny in future research.
General Discussion
Theoretical Implications
Our study provides a novel interpersonal mechanism for understanding the effects of perceived overqualification. Although multiple theoretical frameworks have been posited to understand the impact of perceived overqualification on work behavior (Bashshur et al., 2011), they ignore the reality that overqualified employees do not exist in a social vacuum but are embedded in relational contexts surrounded by their coworkers. Erdogan, Bauer, Peiró, and Truxillo identified the examination of additional mediators of overqualification as an important unresolved issue and hinted at the relevance of a relational perspective by stating that “examining overqualification by paying simultaneous attention to one’s coworkers seems important” (2011a: 264). To date, little theory and research have attended to the relational mechanism directly related to coworkers of overqualified employees. We take a step toward filling this gap and address their call by establishing social acceptance from coworkers as a relational underpinning. Our findings suggest that overqualified employees are not merely influenced by comparisons between what they actually have and they deserve and the associated emotional and cognitive responses. Instead, how they interact with coworkers has a significant impact on their perceived social acceptance in teams, which largely determines their performance. Our attention to the relational mechanism thus broadens the existing knowledge of the social process through which perceived overqualification influences employees’ behaviors.
Second, our study extends the current understanding of the boundary conditions under which perceived overqualification is beneficial or harmful. It demonstrates interpersonal influence as a key moderator to qualify the effects of perceived overqualification in relation to performance outcomes. An important insight from the recent development in the overqualification literature is that overqualification may lead to positive or negative outcomes, and the direction of its effects is shaped by boundary conditions (Liu & Wang, 2012). This is an important point suggesting that an examination of the simple effects of perceived overqualification may be misleading. As a case in point, in our study, perceived overqualification was not correlated with any of the outcomes of interest. Researchers have been unanimous on the importance of the examination of the potential moderators (e.g., Erdogan et al., 2011b; Liu et al., 2015; Liu & Wang, 2012; McKee-Ryan & Harvey, 2011), and a few studies have endeavored to uncover factors that can shape the effects of overqualification, such as empowerment (Erdogan & Bauer, 2009), emotional support (Johnson & Johnson, 1997), and peer overqualification (Hu et al., 2015). These studies are generally clustered around how the attitudes and behaviors of overqualified employees are influenced by the context in which overqualification occurs. An exception that deviates from this approach is Liu et al. (2015), who examined how employees with different levels of justice sensitivity responded differently to overqualification. Extending their work, we view overqualified employees as self-regulating agents who can proactively shape their interactions with coworkers using interpersonal influence. Thus, our study provides knowledge on ways of turning surplus skills of overqualified employees into productivity and answers the important question of “how can individuals and organizations make the best of a potentially bad situation” (Erdogan et al., 2011b: 230).
Finally, our study offers additional evidence for the effects of perceived overqualification on positive work behaviors beyond in-role job performance. In a recent review, Bashshur et al. concluded that “the relationship between overqualification and extra task behaviors remains largely unexplored” (2011: 196), and some research has started to expand the consequences of overqualification to other types of positive work behaviors (e.g., Chen, 2009; Zhang et al., 2016). By linking perceived overqualification with altruism and team member proactivity, we add new evidence to the existing literature. Chen (2009) found an association of perceived overqualification with citizenship behavior toward organizations but not with citizenship behavior toward individuals. Following a relational perspective, our research reveals that overqualification may exert a positive indirect influence on interpersonal altruism through enhanced social acceptance when employees’ interpersonal influence is high (Study 1) and a negative indirect effect through reduced social acceptance when interpersonal influence is low (Study 1 and Study 2). Zhang et al. (2016) found that employees with certain types of goal orientation are motivated to be more proactive because of self-efficacy. Given the importance of proactive behavior for organizational success (Campbell, 2000; Frese & Fay, 2001), more knowledge is needed regarding how to ignite such behavior from overqualified employees. Our Study 1 reveals that, under certain conditions (e.g., high interpersonal influence), overqualified employees will experience high social acceptance and utilize their excessive skills and proactively expand their work roles to contribute to their team.
Practical Implications
This study has important implications for practice. It suggests that perceived overqualification has positive indirect effects on in-role job performance, interpersonal altruism, and team member proactivity under certain circumstances. It informs how organizations can reap the potential benefits of overqualified employees. Managers may benefit from the knowledge that overqualification translates into higher effectiveness and more positive behaviors among those who are high on interpersonal influence. Overqualified employees who are lacking this key interpersonal skill may alienate their colleagues, which could hamper their own performance and demoralize them to engage in extrarole behaviors.
Organizations should take active measures to foster interpersonal influence among employees. First, organizations can include interpersonal influence as a selection criterion when interviewing a job applicant who is apparently overqualified. Doing this may largely increase the chance of hiring a good performer and a good organizational citizen. Moreover, certain organizational training programs could be provided to employees to elevate their awareness of the importance of good coworker relationships and equip them with necessary techniques to acquire relevant abilities (Bedwell, Fiore, & Salas, 2014). In addition, organizations can benefit from building a climate that emphasizes interpersonal harmony (Leung, Brew, Zhang, & Zhang, 2011) and high-quality relationships (Stephens et al., 2012). Overqualified employees exposed to this climate are likely to be accepted by coworkers and, thus, motivated to contribute to the team.
Potential Limitations and Future Directions
Despite several methodological strengths (e.g., multisource data and a time-lagged design), this study has some potential limitations. First, like most research that relies on correlational design, we are not able to make strong inferences of causality on the basis of the current data. For example, it is possible that employees feel accepted because they are good performers or behave helpfully. The use of a time-lagged design and the theoretical considerations between these two variables to some extent mitigate the concern about this issue. Moreover, longitudinal research in educational psychology has confirmed the effects of social acceptance on children’s school performance (Flook, Repetti, & Ullman, 2005; Wentzel & Caldwell, 1997) and helping behavior (e.g., Bowers, Woods, Carlyon, & Friman, 2000; Ervin, Miller, & Friman, 1996). Longitudinal and experimental research in the field of management has also supported the causal influence of interpersonal relationships on performance among working employees (Jehn & Mannix, 2001; Jehn & Shah, 1997). Nevertheless, our model should be examined with longitudinal and experimental designs in the work context to confirm the directions of causality.
Second, our two studies utilized data collected from two organizations within one cultural background (i.e., China). It is possible that some organizational or cultural characteristics could play a role in influencing our findings. Cultures characterized by high collectivism, such as China (Hofstede, 2001), are more concerned about the quality of interpersonal interactions. Therefore, the relational model we develop might be more prominent in such cultures. Although empirical research has been done in a Chinese context and has shown that collectivism does not significantly distort the effects of overqualification (Hu et al., 2015), caution is needed when generalizing our findings to other cultural contexts.
Third, consistent with previous researchers (e.g., Erdogan & Bauer, 2009; Maynard et al., 2006), we measured overqualification from employees’ own perspective. However, overqualification may be measured more objectively utilizing expert ratings of job qualifications and employee characteristics (Maltarich et al., 2011). Because our interest was on employees’ subjective experiences and how they translate into employees’ feelings of acceptance by the group, we chose to focus on self-reported overqualification, which is a better predictor of attitudinal variables (Erdogan et al., 2011b). An interesting extension of this work would be an examination of objective overqualification. The relationship between objective and subjective overqualification is currently unknown and it is important to investigate whether the relationship between overqualification and outcomes is dependent on the type of measurement. Maltarich et al. (2011) recommend using theory as the basis for measurement choices, and we would expect outcomes such as job mobility or recruiter reactions to be related to objective overqualification.
Beyond addressing limitations, the present investigation offers several directions for future research. First, we have examined how overqualified employees with high interpersonal influence can be motivated to be good performers and good organizational citizens via enhanced social acceptance. Our theoretical arguments may enable researchers to more broadly explore the consequences associated with this relational mechanism. For example, social acceptance may be the mediator of the relationship between perceived overqualification and organizational deviance behavior. Another possibility is career-related outcomes. Supervisors may recommend an employee who enjoys a high level of social acceptance for promotion because he or she could make a good team leader, thereby providing an ultimate remedy to overqualification. Moreover, well-being of overqualified employees has been an important concern for researchers, which is also a likely consequence of social acceptance. Future research examining these directions may enhance the generalizability of the present model and move closer toward an overarching relational framework for perceived overqualification.
Consistent with prior research (e.g., Erdogan & Bauer, 2009; Liu et al., 2015), the present model of overqualification is conceptualized at the interindividual level. That is, we argue that overqualified employees who are high/low on interpersonal influence are likely to experience higher/lower social acceptance and, thus, exhibit higher (lower) in-role job performance, altruism, and team member proactivity. While this conceptualization makes important contributions to the literature, it is worthwhile to scrutinize the framework at the team level. Research could examine whether and when the proportion of overqualified employees in a team can influence critical relational factors such as team cohesion and trust, which affect team effectiveness. Doing so will contribute to a deeper understanding of the macrolevel dynamics of the model proposed in the current study.
Finally, we theorized about processes through which perceived overqualification and interpersonal influence interactively influence social acceptance without measuring any specific mechanisms. Future research is encouraged to further explore this relationship by directly measuring possible social interactions involved in employee-coworker exchanges (e.g., information exchange and advice giving).
Conclusion
All in all, building on a relational perspective, the present study provides novel insights into the mechanism through which perceived overqualification is related to work performance. It addresses “unresolved issues” regarding additional mediators, moderators, and outcomes of perceived overqualification (Erdogan et al., 2011b) and points to important research directions that can further expand our knowledge of the effects of this construct.
Footnotes
Acknowledgements
This article was accepted under the editorship of Patrick M. Wright.