Capable Fish or Deficient Ponds? A Meta-Analysis of Consequences,Mechanisms,and Moderators of Perceived Overqualification

An Integrative Framework of POQ and Its Consequences

The prevailing view in the existing literature is that POQ ¹ is a negative experience (Erdogan & Bauer, 2021), typically involving perceptions of deficiency (Erdogan et al., 2011; Erdogan, Tomás, Valls, & Gracia, 2018) and misalignment with one’s job or organization (Luksyte & Spitzmueller, 2016; Maynard et al., 2006). Research from this viewpoint has identified several explanatory mechanisms linking POQ to adverse work outcomes, including perceptions about one’s lack of fit within the work environment (e.g., Luksyte et al., 2022), lack of internal interests and values about work (e.g., Simon, Bauer, Erdogan, & Shepherd, 2019), perceptions of relative deprivation (e.g., Lee et al., 2021), and strong negative affective states (e.g., Liu, Luksyte, Zhou, Shi, & Wang, 2015). However, a contrasting perspective has emerged in recent years, focusing on the surplus skills and qualifications that can accompany POQ (Erdogan & Bauer, 2021). This viewpoint acknowledges the additional capabilities employees who view themselves overqualified can leverage to manage their job demands and perform at high levels (e.g., Deng, Guan, Wu, Erdogan, Bauer, & Yao, 2018). Studies adopting this perspective have identified positive self-assessments, such as self-efficacy (Zhang et al., 2016), as the primary mechanisms linking POQ to beneficial work outcomes. Importantly, the interpretation of POQ (whether as job deficiency or surplus capability) that is most salient in employees’ minds can significantly impact their performance, work behavior, and workplace well-being.

Given these divergent perspectives, scholars have noted that current explanations for why POQ leads to disparate outcomes are limited by the fragmented approach in which explanatory mechanisms have typically been investigated in prior literature (Erdogan & Bauer, 2021). This is particularly noteworthy given the abundance of mediators that have been proposed as potential explanations for the negative consequences of POQ. We concur with Erdogan and Bauer (2021: 274) that, “the examination of mediators would benefit from taking a theory-based approach.” To this end, we draw from existing theory and research on work motivation (Hu et al., 2022; Parker et al., 2010) to provide a comprehensive, theory-driven framework for understanding why POQ leads to positive or negative consequences.

Grounded in traditional theories of motivation (e.g., Eccles et al., 1983; Vroom, 1964), Parker et al. (2010) introduced a theoretical model to explain the dominant underlying motivational states that influence employee behavior. Though it was initially developed to understand proactive behavior, the model has since been applied more broadly to explain a broad array of situations and work outcomes (e.g., Bharanitharan, Chen, Bahmannia, & Lowe, 2019; Hu et al., 2022; Pan & Lin, 2015; Schaubroeck et al., 2017). The first motivational pathway, labeled “reason-to” motivation, reflects the presence of a compelling reason or justification for action and involves questions such as “Why should I do this?” and “Is there value in doing this?” It includes employees’ autonomous motivation and internalized states such as identity and commitment. The second motivational pathway, labeled “energized-to” motivation, reflects the affective states that influence action and involves questions such as “Do I feel like doing this?”, though the influence of certain affective states may be below the level of conscious awareness. It includes employees’ affective states at work. The third motivational pathway, labeled “can-do” motivation, reflects employees’ beliefs regarding their capabilities for action and involves questions such as “Can I do this?” and “Is it feasible for me to do this?” It includes employees’ perceptions about their self-efficacy and control appraisals. These motivational states can play distinct roles and jointly affect employee work outcomes (Cai, Parker, Chen, & Lam, 2019).

We believe Parker et al.’s (2010) model of work motivation provides a compelling framework to examine the divergent consequences of POQ. As illustrated in Figure 1, we systematically organize the dominant mechanisms through which POQ relates to work outcomes according to three distinct explanations. For the reason-to pathway, POQ is perceived as a deficiency that inhibits employees from forming adequate rationales and justifications toward their work. Variables in this category include employees’ cognitive evaluations (i.e., person-environment fit and relative deprivation) and internalized psychological states (i.e., organizational commitment, organizational identification, and autonomous motivation). For the energized-to pathway, POQ is perceived as a deficiency that triggers aversive affective reactions at work that produce deleterious work outcomes. Variables in this category include employees’ affective reactions (i.e., positive affect, negative affect, and work engagement). For the can-do pathway, POQ is perceived as a capability that highlights employees’ beliefs that they possess the talents and resources to perform well and engage positively at work. Variables in this category include employees’ beliefs regarding competence (i.e., self-efficacy) and feasibility (i.e., control appraisal).

We focus on two broad categories of work outcomes in our meta-analytical review. The first is job performance, which includes the following: (a) in-role performance (also termed “task performance”), referring to activities at work that accomplish job duties and responsibilities (Motowidlo, 2003); (b) extra-role performance, involving discretionary actions that contribute to organizational effectiveness (Katz & Kahn, 1978); and (c) counterproductive work behaviors, encompassing voluntary behaviors that hinder an organization’s effectiveness and/or its members (Robinson & Bennett, 1997). This cluster of outcomes offers a comprehensive understanding of employees’ performance at work (Rotundo & Sackett, 2002) and has been widely examined in organizational research (e.g., Colquitt, Scott, & Lepine, 2007; Colquitt et al., 2013; Diehl, Richter, & Sarnecki, 2016; Martin, Guillaume, Thomas, Lee, & Epitropaki, 2016).

The second category is employee workplace well-being, which we conceptualize as job-related satisfaction and mental and physical health in the workplace ² (Danna & Griffin, 1999). Existing research suggests that POQ is largely detrimental to employee well-being (e.g., Erdogan et al., 2018; Harari et al., 2017; McKee-Ryan & Harvey, 2011). However, this may be because studies examining the relationship between POQ and well-being have predominantly viewed POQ as a deficiency (i.e., Erdogan et al., 2018) while overlooking the potential positive impact of POQ on employee well-being. With our proposed framework, we provide a more balanced view of the relationship between POQ and its outcomes, including well-being. In the following section, we develop a series of predictions regarding the effects of POQ on each pathway as well as its distal impact on the aforementioned outcomes.

POQ as Job Deficiency: Impacts on Reason-to and Energized-to States and Outcomes

By and large, people often assume that their job should align with their qualifications (Hu et al., 2015; Vaisey, 2006). For instance, it is generally accepted that those with higher education levels should have jobs with superior social status, better pay and benefits, and promising career prospects (Rose, 2005). The perception of overqualification suggests that employees feel that there is a deficiency in their job, such that their skills and qualifications are not fully utilized or valued in their current roles (Erdogan et al., 2011). We contend that, when interpreted as deficiency, POQ negatively impacts both employees’ rational justifications for their work (i.e., reason-to motivation) and their affective reactions to it (i.e., energized-to motivation), ultimately hindering their work performance and well-being.

Jobs that fail to fully utilize employees’ potential provide little reason or incentive for them to work hard or integrate into the work environment (Liu & Wang, 2012). Existing research suggests that overqualified employees may blame their organization for not providing suitable conditions to showcase their talents, leading employees to psychologically distance themselves from and become less attached to the organization (Fan, He, Dai, Jing, & Shang, 2023; Liu et al., 2015). Overqualified employees are less likely to develop a sense of ownership of the organization or a willingness to work toward collective goals (Ma et al., 2023; Wu, Weisman, Sung, Erdogan, & Bauer, 2022). Consistently, research has found POQ to predict a range of psychological states that reflect a lack of interest in or internalized justification toward their work, including reduced organizational commitment (Bolino & Feldman, 2000; Maynard et al., 2006) and identification (Zhao, Zhao, Zeng, & Bai, 2021), perceptions of lack of fit (Luksyte, Spitzmueller, & Maynard, 2011) and deprivation (Erdogan et al., 2018), and reduced autonomous motivation (Chambel, Carvalho, Lopes, & Cesário, 2021).

In addition to POQ’s effects on employees’ cognitive justifications and internalized psychological states, perceiving underutilization in one’s job may also trigger aversive affective reactions that impede employees’ willingness or ability to fully engage in their work. For example, overqualified employees may feel mistreated by their organization because they believe that the organization should provide opportunities and resources that align with their qualifications (Liu & Wang, 2012; Liu et al., 2015). They may develop feelings of anger and resentment toward their organization over their underutilization (Liu et al., 2015); experience painful social comparisons with their peers (Li, Liao, & Han, 2022); or feel frustrated, anxious, or bored at work (Andel, Pindek, & Arvan, 2022; Zhang, Yan, Wang, Qu, & Qian, 2022). Consistent with this view, research has found POQ to trigger a range of negative emotional states (Yu, Yang, Wang, Sun, & Hu, 2021) as well as a reduction in positive affect states (Allan, Rolniak, & Bouchard, 2020) and general work engagement (Dumani, 2015). Taken together, we hypothesize the following:

We further consider the potential for POQ to harm employee job performance and workplace well-being by way of its negative impact on their motivational states. Research has long established that motivational processes, which involve individuals’ cognitive evaluations and internalized states as well as their affective experiences, can significantly influence work outcomes (Kanfer, Frese, & Johnson, 2017). For employees who perceive themselves as overqualified, reduced reason-to and energized-to motivational states can make them less willing to engage and persist in effortful acts such as accomplishing tasks (e.g., Hu et al., 2015) and can even affect their health and psychological well-being (Erdogan et al., 2018). Accordingly, we propose that POQ negatively impacts employee work outcomes through two distinct motivational pathways.

First, POQ has a negative impact on employee work outcomes through reason-to motivational states. Decades of organizational research have emphasized the important role that employees’ cognitive rationales and justifications toward their work play in shaping their work outcomes (Kanfer et al., 2017). For instance, employees’ perceptions regarding how well their skills and abilities align with the organization’s needs (Greguras & Diefendorff, 2009), their psychological attachment to and embeddedness in the work environment (Meyer, Paunonen, Gellatly, Goffin, & Jackson, 1989), or their beliefs regarding how they compare to others both inside and outside of work (Zoogah, 2010) all stand to influence the amount of effort employees will devote to behaving productively and achieving positive work outcomes (Kanfer et al., 2017; Parker et al., 2010). Consistent with these arguments, employees who perceive they are overqualified may struggle to justify exerting sufficient effort toward their work, which can negatively influence their work outcomes. For example, research has found that through its effects on key psychological states, such as perceptions of person-environment fit, relative deprivation, and perceived autonomy, POQ can impact a range of work outcomes including performance and well-being (e.g., Hu et al., 2015; Luksyte et al., 2022; Simon et al., 2019).

Second, POQ negatively impacts employee work outcomes through energized-to motivational states. It is well understood in research that experiencing negative affective states, or failing to experience positive states and become engaged at work, can have profound consequences on employees’ well-being and their tendency to perform productively at work. Indeed, as with the emergence of negative cognitive appraisals, the experience of aversive affective reactions can impair individuals’ self-regulatory functioning (Lerner & Tiedens, 2006; Tiedens, 2001) and detract from the amount of effort employees put into their work (Kaplan, Bradley, Luchman, & Haynes, 2009); these are key processes that can thwart employees’ positive work behavior and well-being (e.g., Koopman, Lanaj, & Scott, 2016; Lian, Brown, Ferris, Liang, Keeping, & Morrison, 2014). Consistent with these arguments, research has found that POQ can produce negative emotional reactions that can ultimately harm their work outcomes (Peng, Yu, Peng, Zhang, & Xue, 2023). For example, through its effects on employees’ affective states, such as anger and envy, POQ is likely to lead to a range of detrimental outcomes, including reduced performance and increased CWBs (Li et al., 2022; Liu et al., 2015; Peng et al., 2023).

Taken together, we propose that POQ will impede employees’ reason-to and energized-to states, which ultimately detracts from their work performance and overall well-being. We therefore hypothesize the following:

POQ as Surplus Capability: Impacts on Can-Do States and Outcomes

While POQ may be seen as job deficiency and, therefore, lead to detrimental effects on employee outcomes, a growing body of literature suggests that POQ can also yield beneficial consequences in the workplace (e.g., Lee et al., 2021; Lin et al., 2017; Zhang et al., 2016). This view highlights the positive self-assessments overqualified employees may form regarding their surplus skills and qualifications (Erdogan & Bauer, 2021; Lee et al., 2021). Such positive self-assessments can shape overqualified employees’ beliefs about their ability to successfully manage job demands and complete their tasks (Van Dijk et al., 2020; Zhang et al., 2016). Accordingly, we posit that when interpreted as capability, POQ facilitates employees’ beliefs in their abilities and control over their work (i.e., can-do motivational states), ultimately facilitating positive work outcomes.

Although viewing oneself as overqualified for the job might provide little reason or energy for employees to engage productively in their work, it is likely to instill in them the understanding that they are certainly capable of performing at a high level. Prior research, for instance, has found a positive relationship between POQ and role-breadth self-efficacy (Zhang et al., 2016), as well as task mastery (Lee et al., 2021). In line with this prior work and research on work motivation (Parker et al., 2010), we focus on two sets of psychological states that reflect employees’ can-do assessments and beliefs about their competence in relation to their work. These are beliefs about their ability to execute behavior required to successfully finish work tasks (i.e., self-efficacy) and beliefs that their behavior will result in desired outcomes (i.e., control appraisal). We therefore hypothesize the following:

Through its impact on employees’ can-do states, POQ should have a beneficial impact on work outcomes. A wide range of studies have suggested that individuals’ belief in their ability is a critical factor for achieving high levels of job performance and work-related well-being (Judge & Bono, 2001; Nielsen, Nielsen, Ogbonnaya, Känsälä, Saari, & Isaksson, 2017; Wood & Bandura, 1989). Indeed, a belief that one can succeed and achieve desirable outcomes can prompt individuals to set more challenging goals at work (Locke & Latham, 1990) and persistently pursue more effective and efficient behavioral strategies (Lent, Brown, & Larkin, 1987). For overqualified employees, the emergence of such can-do states may enhance their tendency to utilize their surplus knowledge or abilities to carry out job-related activities more effectively (Erdogan & Bauer, 2021).

In particular, prior research suggests that a heightened belief about their abilities may increase overqualified employees’ task-related efforts (Erdogan & Bauer, 2009) and provide them with the additional bandwidth to put forth effort beyond their required in-role duties (Deng et al., 2018). Because they possess skills and expertise that exceed what is needed in their job, overqualified employees may also sense that they have more control over the environment (Lin et al., 2017; Zhang, Wang, Qian, & Parker, 2021), which can allow them to deploy effective coping strategies to anticipate and overcome performance shortfalls (Carver & Scheier, 2001; Zimmerman, 2000). A strong sense of control may also help overqualified employees refrain from engaging in counterproductive behavior. Prior research, for instance, suggests that without effective ways to cope with work demands, employees may lash out in ways harmful to individual and organizational performance (Krischer, Penney, & Hunter, 2010). Moreover, a series of studies have found that merely feeling in control over one’s own behavior can provide individuals with the necessary willpower to avoid maladaptive behavior (e.g., Chow, Hui, & Lau, 2015; Job, Dweck, & Walton, 2010), including CWBs (e.g., Debusscher, Hofmans, & De Fruyt, 2016). Furthermore, a heightened level of perceived competence and control implies that employees hold positive self-views (Deng et al., 2018), which should be beneficial to their overall well-being at work (Nielsen et al., 2017).

Taken together, we propose that POQ will activate employees’ can-do states, thereby exerting a beneficial impact on their work performance and overall well-being. We, therefore, hypothesize the following:

Moderators of POQ and Its Consequences

As described previously, POQ can be interpreted as job deficiency or surplus capability, which can differentially influence work outcomes. As a result, it is important to further understand where and for whom such disparate consequences occur. In the present research, we focus on cultural characteristics (i.e., cultural flexibility and individualism), the economic environment (GDP per capita), and sociodemographic characteristics (i.e., age and gender). Notably, we go beyond prior studies by leveraging an advanced culture framework—that is, the Minkov-Hofstede two-dimensional cultural model—to comprehensively examine the cross-cultural differences that may shape the relationship between POQ and its outcomes (Minkov & Kaasa, 2022; Minkov et al., 2017, 2018). In addition, we provide one of the first comprehensive examinations of how the broader economic context (i.e., GDP per capita) influences the relationships between POQ and work outcomes. Finally, our investigation into the role of gender and age goes beyond their mere relationships with POQ itself, focusing instead on how such sociodemographic variables shape overqualified employees’ motivational states and subsequent work outcomes.

Post Hoc Analysis for Methodological Moderators

Methodological issues have been recognized as an important, yet largely ignored, aspect in POQ research (Feldman, 1996; Liu & Wang, 2012; McKee-Ryan & Harvey, 2011). Investigating whether the existing findings in the POQ literature vary due to methodological issues is an essential undertaking to advance the field. Indeed, researchers have emphasized the need for improved methodological rigor in POQ research (McKee-Ryan & Harvey, 2011). Therefore, employing a targeted post hoc approach (Hollenbeck & Wright, 2017), we further examine potential variations that might arise from methodological moderators, specifically the measures used to assess POQ and the study design employed.

Literature Search and Inclusion Criteria

We took several steps to systematically identify eligible studies for this meta-analysis. First, we searched online databases—including Web of Science, Google Scholar, EBSCO Host Web, and ProQuest Dissertations & Theses Global—for relevant articles. We used search terms relevant to our focal variable, POQ. These search terms included “perceived overqualification,” “overqualification,” “overqualified,” “overqualified employee,” “subjective underemployment, ³ ” “perceived underemployment,” “overeducation,” and “skill underutilization.” Second, we searched for articles published in top-tier journals of management and applied psychology, including Journal of Applied Psychology, Academy of Management Journal, Administrative Science Quarterly, Journal of Management, Personnel Psychology, Organizational Behavior and Human Decision Processes, Journal of Personality and Social Psychology, Organization Science, Journal of Organizational Behavior, Leadership Quarterly, Human Resource Management, Journal of Vocational Behavior, Journal of Occupational and Organizational Psychology, and Management and Organization Review. Third, we included articles about POQ from the reference lists of prior review articles (i.e., Erdogan & Bauer, 2021; Harari et al., 2017; Liu & Wang, 2012; Mckee-Ryan & Harvey, 2011; Yang & Li, 2021). Finally, we solicited conference articles from the annual meetings of the Academy of Management and the Society for Industrial and Organizational Psychology for unpublished manuscripts. We completed the literature search on January 31, 2024, and identified 1,033 articles in total. ⁴

Next, we established several criteria to determine whether to include each of the identified articles and studies reported. First, the included articles must be empirical and quantitative research. Therefore, nonempirical pieces such as review articles or conceptual articles were not included (e.g., Erdogan & Bauer, 2021; Liu & Wang, 2012). Second, to be included in our meta-analysis, a study must examine the relationship between POQ and at least one mediating or outcome variable. Third, included studies must report correlations and sample size. This inclusion rule ensured that we had sufficient information to calculate meta-analytic effect sizes (Schmidt & Hunter, 2015). In addition, if a research article involved more than one sample (e.g., two studies with nonoverlapping samples), we coded them separately. However, if we deem that the same sample was used repeatedly (e.g., Duan et al., 2022; Xia, Xu, &Wu, 2019), we only included the more recent version to avoid repetitive inclusion of the same sample and biased estimation of parameters (Lyubykh et al., 2022). In sum, we included 251 independent samples from 226 articles, with a total sample size of 87,229 employees. Among these 251 independent samples, 85.25% are from published academic journals (219 samples), 10.36% are from doctoral or master dissertations (26 samples), and 2.39% are from unpublished conference papers (6 samples). Additionally, for samples that offered no information about the country background (16 samples) or included participants from several countries (14 samples), the cultural orientation dimensions were not coded. Furthermore, 61 samples provided no information about age composition, and 17 samples did not provide information regarding gender composition. These samples were included in the analyses of the main effects but not in the analyses of respective moderating effects. We depict the literature search and inclusion processes with a flowchart in Figure 2 and provide the full list of studies included in our meta-analyses in online supplementary materials.

OPEN IN VIEWER

Figure 2

A Flowchart of the Literature Search, Inclusion, and Meta-Analytic Procedures

Coding Procedures

We followed the meta-analytic coding procedures recommended by Lipsey and Wilson (2001). Specifically, our author team first created a coding sheet, including study information (e.g., authors, year of publication, journal names/conference paper/dissertation), variables measured, correlation coefficients, reliability coefficients (i.e., Cronbach’s alphas), sample size, the country where the study was conducted, the cultural value scores and GDP per capita of the country where the sample was collected (described in detail later), sample mean age, sample gender composition, measures of POQ, and study designs (i.e., cross-sectional vs. longitudinal, single source vs. multiple sources). Second, two authors and a trained research assistant (a graduate student majoring in management) coded all selected studies. Each study was coded by two coders (i.e., an author and the trained research assistant) independently to ensure coding reliability. Of the 19,712 cells, 18,859 were coded identically across two coders (agreement rate = 95.67%). The remaining discrepancies were resolved by the whole author team via discussions until the agreement reached 100%. Third, the author team discussed and integrated variables with similar measures into aggregated categories based on the conceptualization of and the items used for each measure (see online supplementary material for details). This is a valid approach to testing higher-level constructs in a meta-analysis (Schmidt & Hunter, 2015) and has been widely used in prior research (e.g., Chung, Zhan, Noe, & Jiang, 2022; Crawford, LePine, & Rich, 2010; Jiang, Lepak, Hu, & Baer, 2012). We explicate the definitions and measures of our focal variables in the following sections.

Meta-Analytic Procedures

Following Schmidt and Hunter’s (2015) psychometric approach, we conducted the meta-analysis based on the random effect model. First, we calculated the sample-weighted mean correlation for each relationship (r). Second, we corrected measurement errors and estimated the true score correlations (ρ) by utilizing the reliability coefficients (Schmidt & Hunter, 2015). For variables for which there was no information on reliability coefficients available, we calculated the sample size weighted average reliability from other studies (Chung et al., 2022; see Table 1 for the reliabilities of our measures). Third, we computed the 95% confidence intervals (CIs) around the true score correlations. A 95% CI that excludes zero indicates a significant relationship. Moreover, we also calculated the 80% credibility intervals (CVs), which provides information about the distribution of effect size.

OPEN IN VIEWER

Table 1

Weighted Mean Reliabilities of Focal Variables

Variables	k	Weighted mean reliability
POQ	251	.84
Reason-to motivational states
Overall	87	.85
Autonomous motivation	27	.89
Organizational commitment	23	.80
Organizational identification	8	.87
Person-environment fit	13	.90
Relative deprivation	16	.81
Energized-to motivational states
Overall	61	.89
Negative affect	39	.89
Positive affect	10	.90
Work engagement	12	.88
Can-do motivational states
Overall	40	.84
Self-efficacy	34	.85
Control appraisal	6	.81
Work outcomes
In-role performance	34	.83
Extra-role performance	87	.87
Counterproductive work behavior	51	.86
Workplace well-being	91	.89

Note: k = total number of effect sizes included in analysis.

To verify the existence of potential moderators, we first calculated the percentage of the variance of observed correlations attributable to sampling error and measurement error (%Var). A percentage lower than 75% indicates the potential existence of moderators (Schmidt & Hunter, 2015). To examine moderating effects, we used a random-effect meta-regression procedure for analyses (Field & Gillett, 2010). For categorical moderators, we further conducted subgroup analysis to detect the effect size differences among groups (Schmidt & Hunter, 2015).

Next, we used MASEM (Viswesvaran & Ones, 1995) to test the mediation effects. We conducted the analyses in Mplus 8.5 (Muthén & Muthén, 1998-2017), building an 8 × 8 meta-analytic correlation matrix based on the estimation of true score correlations. In addition to the mediation paths shown in Figure 2, we also controlled for the direct effects of POQ on the four work outcome variables. We allowed the three mediators and four outcomes to freely covary, respectively. We calculated the harmonic mean of the sample sizes for MASEM, following the suggestions of Viswesvaran and Ones (1995). We then conducted Monte Carlo simulation procedures with 20,000 repetitions and calculated the 95% CIs using the online R tool (http://www.quantpsy.org/medmc/medmc.htm; Selig & Preacher, 2008) to test the indirect effects.

In addition to MASEM, we also conducted supplemental analyses using FIMASEM (Yu et al., 2016) as a robustness check. Yu et al. noted that the traditional MASEM assumes homogeneity of effect sizes, which may be inconsistent with the heterogeneity assumption of the random effect model, and proposed FIMASEM as an alternative. Specifically, the FIMASEM model bootstrapped the effect sizes (ρ) and their heterogeneity (SD_ρ) and estimated the structural equation model for each bootstrapped matrix. This procedure results in a distribution for each path coefficient. The 80% credibility interval (CV_β) is then calculated for each path coefficient based on its distribution. This 80% CV_β represents the range within which 80% of the population parameters fall. We also calculated the percentage of effect sizes above or below zero in the population. We conducted the FIMASEM with 5,000 bootstrap iterations using the online web-based application developed by Yu et al. (https://mgmt.shinyapps.io/masem/).

Meta-Analytic Correlations

The Effects of POQ

We first examined the relationships between POQ and three types of motivational states. The results of meta-analytic correlations (see Table 2) show that POQ is negatively related to reason-to motivational states (ρ = −.30, 95% CI = [−.35, −.26]), supporting Hypothesis 1a. Specifically, POQ is negatively related to autonomous motivation (ρ = −.27, 95% CI = [−.35, −.19]), organizational commitment (ρ = −.39, 95% CI = [−.45, −.32]), organizational identification (ρ = −.13, 95% CI = [−.22, −.04]), and person-environment fit (ρ = −.25, 95% CI = [−.38, −.13]) and positively related to relative deprivation (ρ = .40, 95% CI = [.31, .48]). We also found that POQ is negatively related to energized-to motivational states (ρ = −.25, 95% CI = [−.30, −.19]), supporting Hypothesis 1b. Specifically, POQ has a significant positive relationship with negative affect (ρ = .29, 95% CI = [.23, .35]) and negative relationship with work engagement (ρ = −.19, 95% CI = [−.29, −.09]) but is not significantly related to positive affect (ρ = −.13, 95% CI = [−.27, .01]). Further, POQ is positively related to can-do motivational states (ρ = .15, 95% CI = [.07, .23]), supporting Hypothesis 3. Specifically, POQ is positively related to both self-efficacy (ρ = .15, 95% CI = [.06, .25]) and control appraisal (ρ = .15, 95% CI = [.09, .20]).

OPEN IN VIEWER

Table 2

Meta-Analytic Correlations Between POQ, Motivational States, and Work Outcomes

Relationships	k	N	r	SDr	ρ	SDρ	95% CI	80% CV	%Var
POQ → Motivational states
Reason-to motivational states a
Overall	87	31,221	−.26	.18	−.30	.20	[−.35, −.26]	[−.56, −.05]	7.92%
Autonomous motivation	27	11,094	−.22	.18	−.27	.20	[−.35, −.19]	[−.52, −.01]	7.16%
Organizational commitment	22	6,486	−.32	.13	−.39	.14	[−.45, −.32]	[−.57, −.21]	16.62%
Organizational identification	8	2,672	−.11	.11	−.13	.12	[−.22, −.04]	[−.28, .02]	24.39%
Person-environment fit	13	4,748	−.23	.20	−.25	.22	[−.38, −.13]	[−.54, .03]	5.93%
Relative deprivation	17	6,221	.33	.16	.40	.17	[.31, .48]	[.18, .62]	9.68%
Energized-to motivational states b
Overall	61	21,345	−.21	.18	−.25	.20	[−.30, −.19]	[−.50, .01]	8.10%
Negative affect	39	14,073	.25	.17	.29	.19	[.23, .35]	[.04, .53]	8.31%
Positive affect	10	2,808	−.12	.20	−.13	.22	[−.27, .01]	[−.40, .15]	8.48%
Work engagement	12	4,464	−.16	.16	−.19	.17	[−.29, −.09]	[−.40, .02]	11.44%
Can-do motivational states
Overall	40	10,974	.13	.22	.15	.25	[.07, .23]	[−.16, .47]	7.74%
Self-efficacy	34	9,104	.13	.24	.15	.27	[.06, .25]	[−.19, .50]	6.69%
Control appraisal	6	1,870	.12	.06	.15	.04	[.09, .20]	[.09, .20]	72.37%
POQ → Work outcomes
In-role performance	34	11,127	−.002	.16	.002	.18	[−.06, .06]	[−.22, .23]	12.66%
Extra-role performance	87	28,058	−.01	.24	−.01	.28	[−.07, .05]	[−.37, .35]	5.20%
Counterproductive work behavior	51	17,740	.19	.19	.22	.21	[.16, .28]	[−.05, .50]	7.27%
Workplace well-being	91	38,038	−.27	.13	−.31	.14	[−.34, −.28]	[−.49, −.13]	12.44%
Motivational states → Work outcomes
Reason-to motivational states
In-role performance	12	3,543	.22	.09	.26	.08	[.20, .31]	[.16, .36]	41.83%
Extra-role performance	19	6,030	.24	.13	.29	.13	[.22, .35]	[.12, .46]	18.12%
Counterproductive work behavior	15	4,551	−.18	.24	−.21	.28	[−.35, −.06]	[−.56, .15]	5.18%
Workplace well-being	32	13,893	.49	.21	.56	.23	[.48, .65]	[.27, .86]	3.22%
Energized-to motivational states
In-role performance	8	3,256	.33	.09	.39	.09	[.31, .46]	[.27, .51]	23.10%
Extra-role performance	20	8,508	.23	.38	.26	.13	[.20, .32]	[.09, .43]	13.70%
Counterproductive work behavior	19	7,206	−.30	.16	−.34	.18	[−.43, −.26]	[−.57, −.12]	8.48%
Workplace well-being	11	5,244	.37	.21	.41	.23	[.27, .55]	[.11, .71]	3.63%
Can-do motivational states
In-role performance	5	1,610	.41	.11	.52	.16	[.37, .67]	[.31, .73]	11.77%
Extra-role performance	12	3,043	.38	.22	.44	.26	[.29, .59]	[.10, .77]	5.45%
Counterproductive work behavior	4	1,305	−.14	.16	−.16	.18	[−.34, .03]	[−.39, .07]	10.69%
Workplace well-being	6	2,097	.20	.17	.24	.18	[.09, .39]	[.01, .48]	10.09%

Notes. k = number of correlations; N = total sample size; r = mean sample size weighted observed correlation; SD_r = standard deviation of the observed correlation; ρ = mean sample size weighted corrected correlation; SD_ρ = standard deviation of the corrected correlations; 95% CI = 95% confidence interval around the corrected correlation; 80% CV = 80% credibility interval around the corrected correlation; %Var = the percentage of observed variance accounted for by sampling error and measurement error.

a

The correlations between POQ and relative deprivation are reverse-coded in the relationship between POQ and reason-to motivational states.

b

The correlations between POQ and negative affect are reverse-coded in the relationship between POQ and energized-to motivational states.

In addition, consistent with prior mixed findings on the relationship between employees’ POQ and their behavioral outcomes (Erdogan & Bauer, 2021), our meta-analytical results showed that the relationship between POQ and in-role performance (ρ = .002, 95% CI = [−.06, .06]) and the relationship between POQ and extra-role performance (ρ = −.01, 95% CI = [−.07, .05]) are both nonsignificant. However, our results showed that the relationship between POQ and counterproductive work behavior is significantly positive (ρ = .22, 95% CI = [.16, .28]) and the relationship between POQ and workplace well-being is significantly negative (ρ = −.31, 95% CI = [−.34, −.28]). Next, we move to test the countervailing mechanisms between POQ and employees’ work outcomes.

Meta-Analytic Structural Equation Model

To test the mediation effects, we used MASEM (Viswesvaran & Ones, 1995) to build a model that simultaneously included POQ, the three motivational mechanisms, and the four work outcomes. The correlation matrix for the model is shown in Table 3. The estimated path coefficients are depicted in Figure 3 and Table 4. All path coefficients are significant in expected directions, except the relationship between reason-to motivational states and counterproductive work behavior (β = −.01, p = .52) . We then conducted Monte Carlo simulation procedures with 20,000 repetitions (Selig & Preacher, 2008) to test the indirect effects. The results showed that reason-to motivational states mediate the relationship between POQ and (a) in-role performance (indirect effect = −.02, 95% CI = [−.02, −.01]), (b) extra-role performance (indirect effect = −.05, 95% CI = [−.06, −.04]), and (c) workplace well-being (indirect effect = −.13, 95% CI = [−.14, −.11]) but not (d) counterproductive work behavior (indirect effect = .003, 95% CI = [−.01, .01]), thus partly supporting Hypothesis 2a.

OPEN IN VIEWER

Table 3

Correlation Matrix of POQ, Motivational States, and Work Outcomes

Variable	1	2	3	4	5	6	7	8
1. POQ
2. Reason-to motivational states (r, ρ)	−.26, −.30
SD ρ	.20
k (N)	87(31,221)
95% CI	[−.35, −.26]
3. Energized-to motivational states (r, ρ)	−.21, −.25	.41, .48
SD ρ	.20	.15
k (N)	61(21,345)	16(5,158)
95% CI	[−.30, −.19]	[.41, .56]
4. Can-do motivational states (r, ρ)	.13, .15	.19, .24	.27, .32
SD ρ	.25	.15	.16
k (N)	40(10,974)	7(2,193)	10(2,771)
95% CI	[.07, .23]	[.12, .36]	[.21, .42]
5. In-role performance (r, ρ)	−.002, .002	.22, .26	.33, .39	.41, .52
SD ρ	.18	.08	.09	.16
k (N)	34(11,127)	12(3,543)	8(3,256)	5(1,610)
95% CI	[−.06, .06]	[.20, .31]	[.31, .46]	[.37, .67]
6. Extra-role performance (r, ρ)	−.01, −.01	.24, .29	.23, .26	.38, .44	.49, .58
SD ρ	.28	.13	.13	.26	.15
k (N)	87(28,058)	19(6,030)	20(8,508)	12(3,043)	13(4,625)
95% CI	[−.07, .05]	[.22, .35]	[.20, .32]	[.29, .59]	[.50, .67]
7. Counterproductive work behavior (r, ρ)	.19, .22	−.18, −.21	−.30, −.34	−.14, −.16	−.20, −.24 a	−.10, −.12
SD ρ	.21	.28	.18	.18	.11	.29
k (N)	51(17,740)	15(4,551)	19(7,206)	4(1,305)	13(3,150)	15(4,984)
95% CI	[.16, .28]	[−.35, −.06]	[−.43, −.26]	[−.34, .03]	[−.31, −.17]	[−.27, .03]
8. Workplace well-being (r, ρ)	−.27, −.31	.49, .56	.37, .41	.20, .24	.18, .21	.14, .16	−.39, −.45
SD ρ	.14	.23	.23	.18	.16	.23	.24
k (N)	91(38,038)	32(13,893)	11(5,244)	6(2,097)	10(3,526)	8(3,721)	9(4,576)
95% CI	[−.34, −.28]	[.48, .65]	[.27, .55]	[.09, .39]	[.11, .32]	[.005, .32]	[−.61, −.29]

Notes. r = mean sample size weighted correlation; ρ = mean sample size weighted corrected correlation; k = the number of independent samples; SD_ρ = standard deviation of the corrected correlations; N = the total sample size; 95% CI = 95% confidence interval around the mean sample size weighted corrected correlation.

a

Correlation derived from Mackey, McAllister, Ellen, and Carson (2021).

OPEN IN VIEWER

Figure 3

MASEM and FIMASEM Results for the Relationships Between POQ, Motivational States, and Work Outcomes

OPEN IN VIEWER

Table 4

Results of MASEM and FIMASEM

Path	Model 1: MASEM			Model 2: FIMASEM
	β	SE	p	β	SD β	80% CV β	80% CV β width	%β above 0	%β below 0
POQ → Motivational states
Reason-to motivational state	−.30**	.02	.000	−.29	.18	[−.52, −.07]	.59	5.31%	94.69%
Energized-to motivational state	−.25**	.02	.000	−.25	.18	[−.48, −.01]	.49	9.45%	90.55%
Can-do motivational state	.15**	.02	.000	.13	.23	[−.17, .42]	.59	71.10%	28.90%
POQ → Work outcomes
In-role performance	.01	.01	.513	.01	.37	[−.40, .42]	.82	53.18%	46.82%
Extra-role performance	−.002	.02	.890	.01	.44	[−.54, .54]	1.08	51.29%	48.71%
Counterproductive work behavior	.17**	.02	.000	.18	.41	[−.27, .63]	.90	70.27%	29.73%
Workplace well-being	−.17**	.01	.000	−.19	.33	[−.55, .16]	.71	23.36%	76.64%
Reason-to motivational states→ Work outcomes
In-role performance	.05**	.02	.001	.04	.27	[−.25, .31]	.57	61.60%	38.40%
Extra-role performance	.17**	.02	.000	.17	.31	[−.20, .53]	.73	74.39%	25.61%
Counterproductive work behavior	−.01	.02	.516	−.03	.47	[−.60, .54]	1.14	46.37%	53.63%
Workplace well-being	.42**	.01	.000	.39	.39	[−.09, .85]	.95	85.68%	14.32%
Energized-to motivational states→ Work outcomes
In-role performance	.23**	.02	.000	.24	.28	[−.07, .54]	.61	85.27%	14.73%
Extra-role performance	.06**	.02	.001	.07	.34	[−.32, .43]	.75	60.93%	39.07%
Counterproductive work behavior	−.26**	.02	.000	−.26	.40	[−.72, .20]	.92	21.93%	78.07%
Workplace well-being	.13**	.02	.000	.16	.43	[−.37, .67]	1.04	66.39%	33.61%
Can-do motivational states → Work outcomes
In-role performance	.43**	.01	.000	.49	.30	[.16, .81]	.97	95.99%	4.01%
Extra-role performance	.38**	.02	.000	.40	.40	[−.07, .87]	.94	86.33%	13.67%
Counterproductive work behavior	−.10**	.02	.000	−.11	.36	[−.50, .27]	.77	35.84%	64.16%
Workplace well-being	.13**	.01	.000	.12	.33	[−.26, .50]	.76	66.26%	33.74%

Notes. MASEM = meta-analytic structural equation model; β in Model 1 = path coefficients of MASEM; SE = standard error of β; FIMASEM = full information meta-analytic structural equations model;‾β in Model 2 = mean path coefficients with 5,000 bootstrap iterations; SDβ = standard deviation of path coefficients with 5,000 bootstrap iterations; 80% CVβ = 80% credibility interval of the path coefficient; 80% CVβ width = the width of 80% credibility interval of the path coefficient; %β above 0 = percentage of estimated path coefficients above zero; %β below 0 = percentage of estimated path coefficients below zero; N = 4,255 (harmonic mean).

**

p < .01 (two-tailed).

Our results also showed that energized-to motivational states mediate the negative relationships between POQ and (a) in-role performance (indirect effect = −.06, 95% CI = [−.07, −.05]), (b) extra-role performance (indirect effect = −.02, 95% CI = [−.02, −.01]), (c) workplace well-being (indirect effect = −.03, 95% CI = [−.04, −.02]) and the positive relationship between POQ and (d) counterproductive work behavior (indirect effect = .07, 95% CI = [.05, .08]), supporting Hypothesis 2b. We further found that can-do motivational states mediate the positive indirect relationship between POQ and (a) in-role performance (indirect effect = .06, 95% CI = [.05, .08]), (b) extra-role performance (indirect effect = .06, 95% CI = [.05, .07]), (c) workplace well-being (indirect effect = .02, 95% CI = [.01, .03]) and the negative indirect relationship between POQ and (d) counterproductive work behavior (indirect effect = −.02, 95% CI = [−.02, −.01]), supporting Hypothesis 4.

Moderation Effects

Regarding the relationships between POQ and the three motivation mechanisms, the percentage of variance attributable to sampling and measurement error was far smaller than the threshold of 75% (7.92% for reason-to motivational states, 8.10% for energized-to motivational states, and 7.74% for can-do motivational states), indicating the existence of potential moderators of these relationships (see Table 4). Therefore, we moved on to test the moderating effects (see Table 5 for detailed results).

OPEN IN VIEWER

Table 5

The Moderation Effects between POQ and Motivational States

Path	k	N	b	se	95% CI	p
POQ → Reason-to motivational states
National culture
Flexibility	70	25,536	.001	.000	[.000, .001]	.142
Individualism			−.002**	.001	[−.003, −.001]	.002
Economic context
GDP per capita	69	25,058	−.05**	.01	[−.07, −.02]	.000
Sample demographics
Age	64	24,075	−.001	.003	[−.01, .01]	.680
Gender			−.46**	.17	[−.80, −.13]	.007
POQ → Energized-to motivational states
National culture
Flexibility	50	16,766	.003**	.001	[.001, .004]	.003
Individualism			.002	.002	[−.001, .01]	.183
Economic context
GDP per capita	51	18,159	−.03*	.02	[−.06, −.004]	.026
Sample demographics
Age	46	16,603	−.01**	.004	[−.02, .003]	.009
Gender			−.05	.16	[−.38, .28]	.743
POQ → Can-do motivational states
National culture
Flexibility	36	9,452	.002*	.001	[.000, .004]	.045
Individualism			.000	.001	[−.002, .003]	.784
Economic context
GDP per capita	38	10,163	−.03	.03	[−.09, .03]	.341
Sample demographics
Age	25	7,233	−.01	.01	[−.03, .01]	.259
Gender			.01	.26	[−.53, .55]	.959

Notes. Gender represents the percentage of females in samples. k = the number of correlations; N = the total sample size; b= regression coefficients of moderation effects; SE = the standard error of the regression coefficients; 95% CI = 95% confidence intervals (CIs) around the coefficients.

*

p < .05 (two-tailed).

**

p < .01.

Theoretical Implications

Practical Implications

This study also holds significant implications for practitioners. First, our research indicates that POQ has both positive and negative implications for employee work outcomes. Managers need to recognize that overqualification is not solely negative. On the one hand, managers have good reason to be concerned about their overqualified employees, as such perceptions could lead to undesirable motivational and behavioral outcomes, such as more negative affect and counterproductive work behavior. On the other hand, managers should also acknowledge that overqualified employees can bring additional benefits to their organization and make valuable contributions to their work by leveraging their surplus skills and qualifications (Wu et al., 2022). By highlighting the competing motivational mechanisms that determine which outcomes of POQ are likely to manifest, we provide managers with valuable insights for personnel selection and human resources practices.

Further, our research reveals that in certain cultural contexts (i.e., characterized by high flexibility or low individualism), the negative effects of POQ through reason-to and energized-to motivational states are likely to be alleviated, while the positive effect of POQ via can-do motivational states are more likely to emerge. This suggests that cultural background acts as an important boundary condition shaping the implications of POQ. Practitioners can use these findings to gain a better understanding of the advantages and disadvantages of overqualified employees in a global setting. In addition, while managers may hardly have control over the national culture where their overqualified employees are situated, they can apply the lessons learned to their organization’s culture to mitigate the harm from POQ and maximize its potential benefits. Managers might consider adjusting their human resources practices to help sensitize overqualified employees to their underlying competence and provide them with a sense of control over their work. For example, managers can offer overqualified employees more developmental opportunities, help them craft alternative career plans, and design their work to allow for greater task autonomy and impact on others, which may help those employees interpret their situation in a more positive light. These practices may help maximize the potential benefits of overqualified employees while minimizing potential negative consequences associated with POQ.

Limitations and Future Research

We acknowledge the limitations of this research, which also highlights opportunities for future research. First, as with other meta-analytic studies, the present study runs the potential risk of “mixing apples with oranges” concerning grouping similar variables together to form higher-order constructs (Cortina, 2003; Oh, 2020). However, we have taken ways to cautiously address this issue by providing clear definitions for all variables included in our model and systematically examining whether the measures in our samples align with these definitions (details can be found in the online supplementary material). We are confident that our theoretical framework and findings can serve as a sound roadmap for future studies.

Second, although we aimed to test a series of key moderators in our meta-analytical investigation, we were limited in moderators we could include due to the lack of empirical studies on certain factors. As a result, there exist compelling moderators that could enhance our understanding of how organizations can effectively manage overqualified employees that we were unable to examine in this study. For instance, certain leadership behaviors may bring about the more positive side of POQ (Wu et al., 2022). In addition, scholars have suggested that overqualified employees’ attributions regarding the reasons for their overqualification may significantly impact their reactions (Liu & Wang, 2012; Lin et al., 2017). Employees who attribute their overqualification to external factors such as a competitive labor market might be more likely to accept their underemployed status and view their surplus qualifications as valuable resources that can help them be recognized as “rising stars” within the organization. By contrast, if employees attribute their POQ as a result of injustice, such as unfair selection or promotion procedures, they may react more negatively. In addition, overqualified employees with different stable characteristics, such as personality traits, are likely to react differently (e.g., justice sensitivity, Liu et al., 2015; proactive personality, Wu & Chi, 2020). We encourage future studies to delve deeper into how organizational policies, human resources management tools, leadership behaviors, and the individual characteristics of overqualified employees influence their motivational states and work outcomes.

Third, while we operationalized the cultural dimensions of flexibility and individualism at the national level, it is worth noting that these dimensions have also been studied within different contexts and levels of analysis, including the firm, team, and individual levels. Therefore, we encourage future studies to take a multilevel perspective (Kozlowski & Klein, 2000) and investigate how organizational culture or climate (Lumpkin, Brigham, & Moss, 2010), team climates (Bell, 2007), and individual values (Strauss, Griffin, & Parker, 2012) impact overqualified employees’ motivational states and work outcomes.

Fourth, while our study contributes to the understanding of how sociodemographic characteristics shape employees’ reactions to their POQ, further research is necessary to comprehend the complex role that sociodemographic factors play in these relationships (Feldman, 1996; McKee-Ryan & Harvey, 2011). On the one hand, research has observed that POQ and employee affective reactions undergo changes during the socialization process (Simon et al., 2019). Therefore, other demographic characteristics such as job tenure, organizational tenure, and team tenure may also influence the consequences of POQ. On the other hand, although we empirically tested how age and gender moderate the effects of POQ on its consequences, we did not explore potential theoretical mechanisms where competing explanations may coexist. We encourage future studies to delve into these mechanisms, such as exploring mediated moderation effects (e.g., Wang, Burlacu, Truxillo, James, & Yao, 2015), to gain a better understanding of how sociodemographic characteristics interact with POQ to influence employees’ reactions.

Fifth, consistent with prior research, our study focused on employees’ perception of overqualification as compared to objective overqualification. However, it is important to acknowledge that employees who perceive themselves as possessing higher qualifications than required may not necessarily possess surplus capabilities (Erdogan & Bauer, 2021). Prior research has shown only a weak-to-moderate correlation between POQ and objective overqualification, ranging from .16 to .54 (Arvan, Pindek, Andel, & Spector, 2019; Harari et al., 2017; Lin et al., 2017). Moreover, feelings of overqualification can be influenced by various factors, such as personality traits, interpersonal interactions, and job characteristics (Bolino & Feldman, 2000; Erdogan & Bauer, 2021; Harari et al., 2017). For example, individuals high in trait narcissism, characterized by an inflated sense of self-importance, often perceive themselves in an overly positive manner (Martin, Côté, & Woodruff, 2016). Such individuals may hold exaggerated views of their own knowledge, skills, and abilities, leading them to believe that they possess excessive qualifications beyond the job requirements (Cui, 2023; Harari et al., 2017). As our meta-analysis primarily focuses on the consequences of POQ, we encourage future studies to explore these antecedents to enrich the existing literature.