1 Introduction

Overqualification is a form of underemployment wherein people have more skills, experience, knowledge, and abilities than required for a job (Erdogan et al. 2011b; Maynard et al. 2006). Overqualification is a pervasive organizational phenomenon, especially during times of economic downturn (Reingold 2009). For example, during the global 2008/09 financial crisis, one in five American workers and one in three Australian employees reported being overqualified (Luo 2010; Skills Australia 2009). Other countries experience comparable overqualification rates, ranging from 14 % in India to 31 % in Spain (O’Connell 2010). These figures are of particular concern given consistent evidence that overqualification is negatively related to subjective well-being at work (i.e., an employee’s evaluation of her/his work lives) (Bakker and Oerlemans 2011), such as job satisfaction, emotion at work and organizational affective commitment (Bolino and Feldman 2000; Feldman et al. 2002; Fine and Nevo 2008; Johnson and Johnson 2000a; Johnson et al. 2002).

Despite the prevalence of overqualification in the global marketplace, current research is somewhat limited regarding factors that can mitigate its negative impact on subjective well-being at work. Understanding boundary conditions in which overqualified people might experience higher subjective well-being at work is important because this knowledge may help organizations to more effectively manage these potentially excellent workers (e.g., Erdogan et al. 2011b). In one of the few studies exploring potential moderators of overqualification-outcomes links, Erdogan and Bauer (2009) showed that psychological empowerment attenuated the negative impact of overqualification on related outcomes. For overqualified employees, psychological empowerment likely reduces the felt deprivation stemming from the inability to utilize one’s skills (Crosby 1976), which in turn improves job satisfaction and reduces voluntary turnover.

Although Erdogan and Bauer’s (2009) study is informative, several questions require further investigation. Psychological empowerment refers to employees’ feelings about their job as reflected in four cognitive elements that can elicit intrinsic task motivation: meaningfulness, competence, self-determination, and impact (Spreitzer 1995; Thomas and Velthouse 1990). Psychological empowerment can arise from multiple sources, including individual differences such as self-esteem (e.g., Spreitzer 1995) and an array of contextual variables such as leadership (e.g., Avolio et al. 2004), job characteristics (e.g., Liden et al. 2000) and work practices (see Spreitzer 2007, for a review). Establishing the moderating role of empowerment is useful, but it remains an open question as to what steps an organization might take to enhance feelings of empowerment so as to improve subjective well-being of their overqualified employees. To shed light on this issue, we adopt a work design perspective to assess the moderating role of job autonomy in the overqualification-attitudes link. Job autonomy refers to individuals’ latitude to make decisions about their day-to-day work, such as the methods or the order of tasks employees use to accomplish their task-related activities (Hackman and Oldham 1976). Importantly, job autonomy is a core job design characteristic that is possible to increase through job enrichment (Chung and Ross 1977; Parker 1998; Parker and Wall 1998). In other words, job autonomy is a work feature that can be enhanced, thereby providing a potential point of leverage for reducing the negative effects of overqualification. As we elaborate shortly, job autonomy can have such effect because it enables an individual to engage in a regulatory process to adjust their feelings and behavior to actively improve the situation.

Given the increasing prevalence of overqualification across the world, it is also important to understand whether the same managerial practices are universally effective for influencing the psychological effects of overqualification. Given that effectiveness of managerial practices can be different across cultures (Eylon and Au 1999; Robert et al. 2000), we explore the potential role of national culture in shaping the effectiveness of job autonomy as a buffer of the negative effects of overqualification on subjective well-being at work. In this study, we focus national culture on the dimension of individualism–collectivism and use national scores on this dimension to classify employees from nations with individualistic cultures and employees from nations with collectivistic cultures. Employees in individualistic cultures tend to value independence. They place a high priority on pursuing and maximizing individual goals and strive to be unique, assert oneself, express one’s inner attributes, and focus on personal rights, autonomy and self-fulfillment (e.g., Hofstede 1991; Kagitcibasi 1997; Lu and Gilmour 2007; Markus and Kitayama 1991; Triandis et al. 1988). In contrast, people in collectivistic cultures are apt to focus on building harmonious relationships with others and value belonging, fitting in, restraining oneself, and promoting others’ goals and focus on common fate and values (e.g., Hofstede 1991; Kagitcibasi 1997; Lu and Gilmour 2007; Markus and Kitayama 1991; Triandis et al. 1988). Because of these culturally-specific values, we argue that job autonomy will function differently across cultures in assuaging the negative association between overqualification and subjective well-being at work.

Based on the model of culture fit in managerial practice (Aycan et al. 2000; Kanungo and Jaeger 1990; Mendonca and Kanungo 1994) and regulatory fit theory (Higgins 2005), we argue that job autonomy may alleviate the negative effects of overqualification on subjective well-being at work for employees in individualistic (as opposed to collectivistic) cultures. Specifically, employees in individualistic cultures tend to emphasize their uniqueness and their personal values (e.g., Hofstede 1991; Kagitcibasi 1997; Lu and Gilmour 2007; Markus and Kitayama 1991), and therefore will take advantage of job autonomy to pursue personal values and goals by, for example, crafting greater meaning into their work. As such, job autonomy will help employees in individualistic cultures to regulate the negative impact of overqualification on subjective well-being at work. In contrast, employees from collectivistic cultures tend to nurture and value their relationships with others more than maximizing individual goals (e.g., Hofstede 1991; Kagitcibasi 1997; Lu and Gilmour 2007; Markus and Kitayama 1991). Because the regulatory function of job autonomy is less congruent with such values and goals, we do not expect that job autonomy will mitigate the negative impact of overqualification on subjective well-being at work in collectivistic cultures. In sum, we propose a three-way interaction between overqualification, job autonomy, and national culture in shaping employees’ subjective well-being at work. To examine our hypothesis, data from the 5th European Working Conditions Survey (EWCS) (Gallup Europe 2012) were analyzed. We expect that job autonomy will mitigate negative impact of overqualification on subjective well-being at work for employees from nations with individualistic cultures, but not for employees from nations with collectivistic cultures.

Our investigation contributes to literature in two ways. First, we suggest that overqualification is exacerbated by poor job design; specifically, a lack of job autonomy. Overqualification has been conceptualized as a multi-dimensional construct consisting of two facets: mismatch (i.e., having more qualifications than is required) and no-growth (i.e., a lack of opportunities to utilize one’s skills; (Johnson and Johnson 1996). Rather than focusing on the aspect of personnel selection to recruit employees whose background matches the job (Fine and Nevo 2011), we emphasize the role of job autonomy in allowing for growth, despite the mismatched job. In particular, when overqualified employees have job autonomy, they can proactively craft their jobs and enhance the meaning of work, thereby resulting in more favorable subjective well-being at work. Compared to selecting new employees with job-matched backgrounds, a job design approach can help maximize the development of existing human capital in an organization, which is sometimes the only option open to organizations.

A second contribution of our study is that we offer insights into when job autonomy is most useful in shaping overqualified employees’ subjective well-being at work. As indicated by the job demands-control model (Karasek 1979), job autonomy can help individuals to deal with suboptimal work conditions, and thus positively influence one’s work experiences. However, as other scholars have argued (e.g., Parker and Sprigg 1999), this buffering role of autonomy does not apply to all individuals. In line with this perspetive, we suggest that the effect of job autonomy in mitigating the negative impact of overqualification on subjective well-being at work will only occurs for employees from individualistic cultures. As the business environment has become increasingly globalized (Gibson and McDaniel 2010), organizations need to learn how to effectively manage culturally diverse employees. Accordingly, it is important to unpack the possible differential effect of job autonomy across cultures. In the remainder of the introduction, we review literature on the relationship between overqualification and subjective well-being at work, and then develop our research hypotheses.

1.1 Overqualification and Subjective Well-Being at Work

There are several reasons regarding why overqualification is related negatively to subjective well-being at work (Anderson and Winefield 2011; Erdogan et al. 2011a; Johnson and Johnson 1999, 2000a, b). A person-job fit perspective (Kristof-Brown et al. 2005) suggests that overqualified people have negative work experiences because of a lack of similarity between the individual and their environment (i.e., supplementary misfit, Maynard et al. 2006). This type of person-job misfit will result in negative feelings such as a sense of deprivation. Based on relative deprivation theory (Crosby 1976), overqualification denotes a condition in which an individual has not obtained the expected job that they think they could have attained with their given qualifications. This can result in a feeling of being deprived of utilizing one’s skills or being under-valued, thus leading to negative work experiences such as job satisfaction or organizational commitment (Feldman et al. 2002). Empirical studies support these theoretical assertions. Specifically, overqualified employees tend to be dissatisfied with their jobs and experience lowered affective commitment (Bolino and Feldman 2000; Feldman et al. 2002; Fine and Nevo 2008; Johnson and Johnson 2000a, b; Johnson et al. 2002), are psychologically and physically more distressed (Johnson and Johnson 1996, 1997, 1999), are more likely to voluntarily quit their jobs (Erdogan and Bauer 2009; Erdogan et al. 2011b; Maynard et al. 2006), and have a greater tendency to engage in counterproductive work behaviors (Luksyte et al. 2011).

Despite the informative nature of these studies, they have predominantly focused on examining outcomes of overqualification and have largely ignored moderators of these relationships, with only a handful of studies addressing this gap. Regarding the latter, overqualified employees who value skill utilization and professional growth opportunities are more likely to engage in job search behaviors than those who do not endorse these values (Maynard and Parfyonova 2013). It has also been shown that the negative impact of overqualification on voluntary turnover and job satisfaction is attenuated for those with high levels of psychological empowerment (Erdogan and Bauer 2009) and that the effect of overqualification on psychological distress is weakened for individuals with high levels of emotional support (Johnson and Johnson 1997). Given the prevalence of overqualification, it is important to explore other moderators to enhance insight into ways that organizations might reduce its negative effects. As subjective well-being at work (e.g., job satisfaction) predicts key indicators of organizational effectiveness such as job performance and voluntary turnover (e.g., Harter et al. 2002), this issue is worthy of increased empirical attention.

1.2 Joint Moderating Effect of Job Autonomy and Culture

We propose that job autonomy will moderate the negative association between overqualification and subjective well-being at work, but this effect will only occur for employees from individualistic cultures. In making this argument we recognize that job autonomy can simultaneously have a positive main effect on subjective well-being at work. Motivation theories suggest that job autonomy can result in positive subjective well-being at work. The job characteristics model (Hackman and Oldham 1976) proposed that core job characteristics including autonomy generate a sense of responsibility for outcomes, meaningfulness, and knowledge of results, which in turn elicits intrinsic work motivation and job satisfaction, an indicator of subjective well-being at work. Meta-analytic evidence (Humphrey et al. 2007) shows job autonomy accounts for most variance in meaningfulness and resultant favorable job attitudes. Likewise, from the self-determination theory perspective (Ryan and Deci 2000), job autonomy fulfills the fundamental need for autonomy, which increases intrinsic work motivation and positive work experiences. These positive motivational effects of job autonomy should apply to all individuals. Consistent with this idea, job autonomy predicts subjective well-being at work such as job satisfaction across many cultures, including 42 nations with distinct cultural values (Au and Cheung 2004) as well as American, Australian, and Indian samples (DeCarlo and Agarwal 1999). We thus expect a main effect of job autonomy on subjective well-being at work in our study.

However, we propose that in addition to shaping subjective well-being at work directly, job autonomy can buffer the negative effects of overqualification because it provides an opportunity for individuals to actively deal with suboptimal work conditions, and thereby improve their subjective well-being at work. From this perspective, overqualification is a regulatory mismatch between a person and their situation, and employees with high job autonomy engage in a regulatory process to adjust their feelings and behavior to actively improve the situation. We further argue that only individuals in individualistic cultures will take up the opportunity provided by autonomy to actively improve the situation to achieve regulatory fit.

In particular, job autonomy is not only a motivational resource that boosts subjective well-being at work directly, it may represent a situation which is not constrained by formal rules and procedures (Meyer et al. 2010). As such, autonomy gives individuals the opportunity to regulate their feelings and behaviors to pursue goals based on their personal values (Sheldon and Elliot 1999). For example, job autonomy provides opportunities for employees to increase their meaning in their work via job crafting and other proactive behaviors (Wrzesniewski and Dutton 2001). Through this process, job autonomy can enhance employees’ sense of value at work and thus lower the feelings of relative deprivation and heighten the job satisfaction of overqualified employees. This proposed buffering role of autonomy is similar to that advocated in the demand-control model of strain (Karasek 1979): autonomy is argued to give individuals the opportunity to actively manage demands and thereby mitigate their stressful effects.

However, we suggest this moderating function of autonomy will only operate for some individuals because its effectiveness depends on whether individuals appreciate and engage in the opportunities afforded. Specifically, such a regulatory mechanism will only occur for employees from individualistic cultures because they tend to place a high priority on pursuing individual goals, with a strong focus on personal values, autonomy and self-fulfillment (e.g., Hofstede 1991; Kagitcibasi 1997; Lu and Gilmour 2007; Markus and Kitayama 1991; Triandis et al. 1988). In other words, job autonomy “is congruent with individualistic values, emphasizing freedom of choice and providing the opportunity to influence and to attribute the behavioral outcomes to oneself” (Erez 2010, p. 393). From the perspective of regulatory fit theory (Higgins 2005), which highlights the importance of pursuing goals in a manner that is consistent with one’s values, job autonomy ‘fits’ employees from individualistic cultures because it is a means to regulate feelings and behaviors to sustain the values of being a unique person and achieving self-fulfillment. This sense of fit can facilitate successful self-regulatory process and positive evaluations towards the activities (Förster et al. 1998; Higgins 2005). Employees from individualistic cultures tend to change the environment to fit their needs (Rothbaum et al. 1982; Weisz et al. 1984), such as initiating job changes to enhance person-job fit (Liu et al. 2013). Thus, having autonomy at work may motivate employees to actively improve their situation to achieve greater personal meaning when they are overqualified.

In contrast, people from collectivistic cultures tend to focus on building harmonious relationships with others and value belonging, fitting in, restraining oneself, and promoting others’ goals (e.g., Hofstede 1991; Kagitcibasi 1997; Lu and Gilmour 2007; Markus and Kitayama 1991; Triandis et al. 1988). We suggest that employees from collectivistic cultures are less likely to make use of job autonomy to actively craft their job content to achieve personal meaning when they are overqualified because job autonomy is less congruent with collectivistic values and characteristics. Specifically, employees are less likely to actively shape their jobs because they are more motivated to change themselves to fit the environment, rather than alter the situation to fit their needs (Rothbaum et al. 1982; Weisz et al. 1984). Yet, their desire to build harmonious relationships may prevent them from actively changing their jobs because these modifications may influence others’ work and bring resistance from and conflict with supervisor or colleagues. Because of these tendencies, having job autonomy may not help employees from collectivistic cultures to regulate their subjective well-being at work when they are overqualified as actively shaping the work situation is incongruent with their cultural values.

Other scholars have made similar argument about the moderating effect of autonomy on suboptimal work conditions and its differential application for some individuals but not others. For example, building on the job demands-control model (Karasek 1979), Parker and Sprigg (1999) found that the moderating effect of job autonomy on stressful job demands only applied to individuals high on proactive personality. The authors argued that individuals with a proactive personality “take advantage of autonomy afforded them to manage job demands and thereby limit the threat of demands to their psychological health” (p. 926). This suggests individual differences may play a role in how people utilize job autonomy in dealing with suboptimal work conditions. Similar to this idea but focusing on culture differences, we suggest that only individuals from an individualistic culture are likely to take advantage of autonomy to achieve more meaningful jobs when they feel overqualified. Our reasoning is also in line with the model of culture fit in managerial practice (Aycan et al. 2000; Kanungo and Jaeger 1990; Mendonca and Kanungo 1994), which posits that a managerial practice (e.g., job autonomy) is effective when its characteristics are consistent with the socio-cultural environment (e.g., individualistic cultures) (Aycan et al. 2000; Kanungo and Jaeger 1990; Mendonca and Kanungo 1994).

Hypothesis

There will be a three-way interaction between overqualification, national culture, and job autonomy in predicting subjective well-being at work. The mitigating effect of job autonomy on the negative association between overqualification and subjective well-being at work will depend on national culture. Specifically, overqualification will have a weaker negative association with subjective well-being at work for those with higher rather than lower job autonomy; and this interaction between autonomy and overqualification will be observed in individualistic, but not collectivistic, cultures.

2 Method

2.1 Participants

We used the data from the 5th European Working Conditions Survey (EWCS) collected in 2010 (Gallup Europe 2012). We included a total of 7,310 employees from nine European countries that represent distinct cultures based on individualism–collectivism distinction. We selected four countries (3,405 respondents) with the highest individualism/collectivism scores and five countries (3,905 respondents) with the lowest individualism/collectivism scores to classify them as two groups. The four nations high in individualism were Sweden, Denmark, Netherlands, and United Kingdom. The five nations low in individualism were Portugal, Turkey, Estonia, Lithuania, and Latvia. In the total sample, 52 % were women with the mean age of 41.61 years old (SD = 12.25). The average organizational tenure was 9 years (M = 9.14, SD = 9.39) with 91.8 % participants having worked more than a year in the same organization. Regarding education background, it was distributed as follows: 7.3 % participants completed primary education (ISCED 1) or less, 22.2 % completed lower secondary education (ISCED 2), 29.6 % participants completed upper secondary education (ISCED 3), 6.1 % participants completed post-secondary education (ISCED 4) and 34.9 % participants completed tertiary education—first level (ISCED 5) and advanced level (ISCED 6). The two cultural groups did not differ in the proportion of gender. The range of age was similar in the two cultural groups (16–80 in high individualism group; 16–79 in low individualism group), but the mean age in high individualism group (M = 43.34, SD = 12.00) was higher than that in low individualism group (M = 40.11, SD = 12.27). Participants in the high individualism group are more educated and had more years of organizational experience (M = 9.98, SD = 9.93) than participants in the lower individualism group (M = 8.40, SD = 8.82).

2.2 Measures

2.2.1 Overqualification

We measured overqualification with a single item, wherein the participants indicated the extent to which their skills were comparable with their job requirements. The participants were asked: “Which of the following alternatives would best describe your skills in your own work?” Three options were provided in the survey and based on their responses, we categorized respondents as underqualified if they chose “I need further training to cope well with my duties,” adequately qualified if they chose “My duties correspond well with my present skills,” and overqualified if their answer was “I have the skills to cope with more demanding duties.” We then created dummy variables representing the comparison between underqualification and match, and the comparison between overqualification and match, which we used in subsequent analyses. The use of one-item measure may be justifiable here because the overqualification measure assesses a “self-reported fact” for which it is more acceptable to measure with a one-item scale (cf. Wanous and Hudy 2001; Wanous et al. 1997). Further, based on data reported in Luksyte et al.’s (2011) study on overqualification, a single item very similar to that used in the current study (“I have job skills that are not required for this job”) had significant negative association with outcomes such as person-job fit (i.e., needs-supplies fit; r = −.17, p < .05), burnout (i.e., cynicism; r = .14, p < .05), psychological contract (i.e., balanced; r = −.14, p < .05) and counterproductive work behaviors (r = .13, p = .06); showing a pattern very similar to that obtained using the full 9-item overqualification scale developed by Maynard et al. (2006). This finding suggests a single-item assessment of overqualification used in this research is reliable and valid.

2.2.2 Job Autonomy

We used self-reported measure of job autonomy as perceived job autonomy is highly correlated with objective job autonomy (see Fried and Ferris 1987). We measured job autonomy with six items. Three items used dichotomous response scale ‘yes’ and ‘no’ (i.e., “Are you able to choose or change your order/method/speed of tasks?”). Three items (i.e., “You are able to apply your own ideas in your work,” “You have influence over the choice of your working partners,” and “You can take your break when you wish”) used a five-point Likert scale ranging from 1 (almost always) to 5 (almost never). We used these six items to maximize the scope of job autonomy in different aspects. Supporting the validity of this measure, the exploratory factor analysis showed that these items were influenced by a single factor with loadings ranging from .39 to .76. Because these items had different response scales, we used a factor score derived from regression method (DiStefano et al. 2009) to indicate the level of job autonomy for each participant. We tested a multilevel confirmatory factor analysis (CFA) model in which the six items were influenced by a single latent factor at individual level, and variances of each item across nations were acknowledged at a national level. We examined this model with limited information maximum likelihood (MUML) estimator in Mplus (Muthén and Muthén 1998–2012) because we have unequal sample sizes across nations and only focus on a random-intercept model, which is suitable for using MUML estimator. Also, the MUML estimator has computational and convergence advantage over full information maximum likelihood (FIML) estimator and has similar estimation to FIML estimator (Bovaird 2007). This model fit the data well (MUML-χ 2 = 828.95, df = 24; CFI = .92; TLI = .90; RMSEA = .07; Individual-level SRMR = .06; National-level SRMR = .51). The composite reliability (Fornell and Larcker 1981) of these items derived from the multilevel CFA model was.75.

2.2.3 National Culture Classification (Individualistic/Collectivistic Cultures)

We used a dummy variable to identify two groups of countries. Scores of individualism at a national level were obtained from Suh et al. (1998) which incorporates a score based on Hofstede’s IBM study (Hofstede 1980) and a score based on Triandis’ conceptualization (see Suh et al. 1998 for more detail). The first group consists of Sweden, Denmark, Netherlands, and the United Kingdom (the average individualism score was 8.18 out of 10) and the second group consists of Portugal, Turkey, Estonia, Lithuania, and Latvia (the average individualism score was 3.94 out of 10). We did not use individualism score for each country to examine the culture effect because not all countries in the survey had individualism scores in Suh et al.’s (1998) report. Also, individualism scores reported by Suh et al.’s research (1998) were collected a number of years ago. As countries’ levels of individualism may change over time due to social and economic fluctuations, it remains an open question whether the scores still can describe the current differences between countries at an interval level. However, we are more confident to use the scores to describe the current differences between countries at an ordinal level as rank ordering of countries on individualism is relatively stable (Hofstede 2001). Therefore, we treated individualism scores reported by Suh et al.’s research (1998) as an ordinal scale. To ensure our measure of individualism is valid to reflect the differences across countries, we argue that selecting typical countries with high and low individualism is a suitable approach. We used this approach because rank ordering of nations on individualism is relatively stable (Hofstede 2001) and known-group validity can be warranted when typical, contrasting individualism/collectivism countries are used in analyses. Also, the approach using country-of-origin of subjects as proxy of cultural effects is a general practice in cross-cultural studies (e.g., Bochner and Hesketh 1994; Earley and Mosakowski 1996; Oetzel et al. 2001; Robert et al. 2000).

2.2.4 Subjective Well-Being at Work

We measure subjective well-being at work using available items in the database. In line with Feldman et al.’s (2002) approach, we used items relating to various aspects, including satisfaction with payment, career advancement, attitudes towards organization, and work itself. They jointly reflect employees’ experiences and feelings towards their jobs. The five items were (a) “Are you very satisfied, satisfied, not very satisfied or not at all satisfied with working conditions in your main paid job?” (with four response categories as mentioned in the item); (b) “I am well paid for the work I do;” (c) “My job offers good prospects for career advancement;” (d) “I feel myself ‘at home’ in this organization” (the b–d items used the same response options ranging from 1 (strongly disagree) to 5 (strongly agree)); and (e) “I have the feeling of doing useful work” with response options ranging from 1 (almost always) to 5 (almost never). Exploratory factor analysis showed that these items were influenced by a single factor with loadings ranging from .42 to .71. Similarly, we used a factor score derived from regression method (DiStefano et al. 2009) to indicate the level of subjective well-being at work for each participant. We tested a multilevel CFA model in which the five items were influenced by a single latent factor at individual level, and variances of each items across nations were acknowledged at a national level. This model fit the data well (MUML-χ 2 = 408.39, df = 15; CFI = .93; TLI = .90; RMSEA = .06; Individual-level SRMR = .04; National-level SRMR = .54). The composite reliability of these five items derived from the multilevel CFA model was .68.

2.2.5 Control Variables

We controlled for basic demographic information (i.e., gender, age, education) and tenure in analyses predicting subjective well-being at work at individual level because of their potential to influence feeling and attitudes towards jobs, as evidenced by multiple meta-analyses (e.g., Ng and Feldman 2010).

3 Results

Table 1 presents descriptive statistics among research variables without considering the multilevel structure (i.e., individuals within nine countries). In the sample, 12.3 % participants classified themselves as underqualified; 57 % participants classified themselves as adequately qualified; and 30.8 % participants classified themselves as overqualified. Distributions are similar in both culture groups. Because individuals were nested within countries, we conducted multilevel modeling analyses to test the research hypothesis (Hox 2002).

Table 1 Descriptive statistics among research variables
Full size table

We conducted a series of two-level multilevel models (i.e., country as the second level and individual as the first level) using MIXED procedure in SPSS (Peugh and Enders 2005). Multilevel analysis was justifiable because the intra-class correlation coefficient (ICC(1)) of subjective well-being at work was 0.20, suggesting that 20 % of variance in subjective well-being at work is explained by the fact that people belonged to different nations. We modeled the control variables (i.e., gender, age, tenure, and education), underqualification, overqualification, and job autonomy as level-1 predictors, whereas individualism–collectivism was modeled as level-2 predictors. To obtain unbiased estimates of the hypothesized relationships, it is advisable to use centered scores for the predictor variables in multilevel analysis (Hofmann and Gavin 1998). As such, we centered continuous predictors at level-1 (i.e., age, tenure, education, and job autonomy) to the group-mean as group-mean centering method is more desirable to examine cross-level interactions (Enders and Tofighi 2007). We did not center other variables because they were dummy coded variables. Table 2 presents the results of multilevel modeling analyses in which the models included control variables, main effects, two-way then the three-way interaction effects.

Table 2 Results of multilevel analysis with estimates of fixed effects
Full size table

In addition to these analyses to assess whether there was a three-way interaction, we investigated the conditional two-way interaction between job autonomy and overqualification in predicting subjective well-being at work in individualistic or collectivistic cultures so that we could more readily test the more specific form of the interaction that we predicted. We also further explored the nature of the interaction using simple slope tests (Preacher et al. 2006).

As hypothesized, after the entry of control variables, main effects, and two-way interaction effects, there was a significant three-way interaction in predicting subjective well-being at work (β = .09, p < .05)2. More specifically, we proposed that job autonomy buffers the negative impact of overqualification on subjective well-being at work in individualistic cultures only. Consistent with this prediction, the conditional two-way interaction term of overqualification and job autonomy was significant for employees from individualistic cultures (β = .12, p < .01), but it was non-significant for employees from collectivistic cultures (β = .02, p = .41). Thus, as expected, job autonomy mitigated the relationship between overqualification and subjective well-being at work for employees from individualistic cultures only.

We further explored these findings using simple slope tests as shown in Figs. 1 and 2. Figure 1 depicts the two-way interaction for employees from individualistic cultures. There was a negative relationship between overqualification and subjective well-being at work for people from individualistic countries for low levels of job autonomy (β = −.25, p < .01); whereas there was no significant association between overqualification and subjective well-being at work for high levels of job autonomy (β = −.05, p = .22). As illustrated in Fig. 2, for collectivistic cultures, overqualification was significantly negatively associated with subjective well-being at work for both low (β = −.16, p < .01) and high (β = −.12, p < .01) levels of job autonomy. Yet, this link was stronger for low than for high levels of job autonomy (−.16 vs. −.12). Although the value of pseudo-R 2 is lower (.001) for the model including the three-way interaction, our finding reveals practical significance of the effect as we discussed below. Moreover, excluding control variables (gender, age, tenure, and education) from the model did not influence the results of the three-way interaction in predicting subjective well-being at work.

Fig. 1
figure 1

Plot of interaction effect of perceived overqualification, job autonomy on subjective well-being at work for nations with individualistic cultures. Factor score of subjective well-being at work was used as the dependent variable

Full size image
Fig. 2
figure 2

Plot of interaction effect of perceived overqualification, job autonomy on subjective well-being at work for nations with collectivistic cultures. Factor score of subjective well-being at work was used as the dependent variable

Full size image

4 Discussion

Considering the prevalence of overqualification across the globe and the increasing globalization of the modern marketplace, we examined what organizations can do to mitigate negative consequences of overqualification on subjective well-being at work in different countries with distinct cultural characteristics. Consistent with job design literature (Hackman and Oldham 1976), the model of culture fit in managerial practice (Aycan et al. 2000; Kanungo and Jaeger 1990; Mendonca and Kanungo 1994) and regulatory fit theory (Higgins 2005), the results of a large nationally representative sample suggest that job autonomy might be a useful strategy to mitigate negative consequences of overqualification on subjective well-being at work for employees from individualistic cultures, or those who tend to focus on their uniqueness, personal goals and self-actualization. For these workers, overqualification had a negative association with subjective well-being at work for low levels of job autonomy but no such association existed for employees with high job autonomy. Notably, job autonomy does not appear to yield the same effect for subjective well-being at work for overqualified employees from collectivistic nations, or those who consider more about social relationships with others, common fate with important others and social harmony. For employees in collectivist nations, overqualification was negatively associated with subjective well-being at work regardless of the level of autonomy. These findings have important theoretical and practical implications.

From a theoretical perspective, our study highlights the potentially beneficial role of job design (Hackman and Oldham 1976; Wall et al. 2002) in managing the psychological effects of overqualification, especially in individualistic cultures. Although we did not assess the underpinning mechanisms, our findings are consistent with arguments about the beneficial function of job autonomy in buffering the negative impact of overqualification on subjective well-being at work. In particular, autonomy provides an unconstrained environment that allows and prompts some individuals to take action according to their personal values. As such, employees with high job autonomy might be motivated and/or able to proactively craft their job content to achieve more meaning, which can counteract some of the negative effects of overqualification on subjective well-being at work. Our perspective complements the person-job fit literature (Kristof-Brown et al. 2005) by offering actionable steps regarding how organizations can potentially maximize human capital of their overqualified employees.

If further research shows that, as we have argued, job design is important in mitigating the effects of overqualification in individualistic cultures because it promotes crafting and proactivity, job design as a strategy for addressing over-compensation could have advantages beyond promoting subjective well-being at work of overqualified individuals. Scholars have argued that employee proactivity is increasingly important for success in complex and uncertain work environments (Griffin et al. 2007). As such, organizations can potentially utilize the outstanding credentials of overqualified employees to achieve constructive change and other benefits of job autonomy. We therefore recommend further research to investigate whether overqualified people make use of autonomy to think or behave more proactively, such as engaging in job-role negotiation (Parker and Collins 2010), developing broader and more flexible work orientations (Parker et al. 1997), or cognitively re-framing their jobs (Wrzesniewski and Dutton 2001). Past studies have shown that overqualified employees proactively search for more suitable employment outside of their employing organization (Feldman et al. 2002; Wald 2005). Yet, whether they use other strategies to reduce person-job misfit, for example, by proactively shaping their job content within the organization, has been underexplored. Future research will advance our knowledge in this area by explicitly studying proactive work behaviors of overqualified employees.

Notably, our results showed that job autonomy can shape subjective well-being at work of overqualified employees in two ways. For employees from both individualistic and collectivistic cultures, job autonomy had a positive main effect on subjective well-being at work. For employees in individualistic cultures, job autonomy additionally had a moderating function, mitigating the negative impact of overqualification on subjective well-being at work. Our findings support our reasoning that job autonomy has functions, such as inducing intrinsic motivation and need fulfillment, which are universal for bringing about better subjective well-being at work, but also simultaneously plays a role in promoting autonomous regulation that is only beneficial to those from individualistic cultures. Regarding the latter, we suggested that individualism values support norms and beliefs to adopt this regulatory approach to help employees to deal with suboptimal work conditions such as overqualification. Our results might extend both job design theories (e.g., Parker and Wall 1998) and cross-cultural theories (e.g., Hofstede 1991; Kagitcibasi 1997; Lu and Gilmour 2007; Markus and Kitayama 1991; Triandis et al. 1988) by highlighting that although job autonomy provides several functions to shape individuals’ subjective well-being at work, these functions differ in their effectiveness across cultural boundaries.

Some readers might question why the cultural values of individualism–collectivism influenced the moderating power of job autonomy on the relationship between overqualification and subjective well-being at work, but not the link between overqualification and subjective well-being at work and the link between job autonomy and subjective well-being at work (the two-way interactions).Footnote 1 First, we did not expect that overqualification will have differential negative impact for employees from the two different cultures. This is because overqualification reflects a suboptimal status that an individual holds in a given job market, and its negative implications such as depriving chances for utilizing skills and obtaining expected rewards and appreciation, apply to employees who cannot find an expected job, irrespective of their cultural background. Past studies using samples from countries high in individualism such as USA (e.g., Feldman et al. 2002), or samples from countries high in collectivism such as Turkey (e.g., Erdogan and Bauer 2009) reported similar negative relationships between overqualification and job satisfaction, an indicator of subjective well-being at work. Conversely, one might predict that job autonomy would have a stronger effect on subjective well-being at work of employees from individualistic than collectivistic cultures. Job autonomy has functions that are universal to all employees in shaping one’s subjective well-being at work. Because of its universal appeal, job autonomy is likely to have a positive impact on subjective well-being at work across different cultures. We believe this finding reflects our earlier arguments about regulatory fit, such that culture fit most matters when some self-regulation is required. Overqualification might represent not just a person-job misfit (Maynard et al. 2006), but a regulatory mismatch that requires an individual to engage in a regulatory process to adjust their feelings and behavior to change themselves or the situation. As such, we detected culture fit effects for overqualification in combination with job autonomy, but not for autonomy alone.

From a practical perspective, these results have implications for organizations striving to more effectively manage their overqualified employees across countries with different cultures. First, rather than focusing on recruiting people with matching qualifications, which might not be a feasible strategy in many situations, we suggest that overqualification effects can be mitigated by job redesign or increasing autonomy at work in particular. That is, managers can potentially improve subjective well-being of their overqualified people by offering them more job autonomy. Job autonomy can be increased in several ways at different levels, such as tasks, supervisors, and organizations. In addition to having opportunities to decide when and how to complete tasks (Hackman and Oldham 1976), from a leadership perspective (Yukl et al. 2002), supervisors can delegate decision making to their subordinates. From an organizational design perspective (Daft 2008), organizations can find ways to eliminate bureaucracy so that employees can have more rooms to decide how to do their work. In other words, there is more than one way for organizations to enhance employees’ job autonomy and organizations can use different approaches according to feasibility in their settings.

However, job autonomy proves useful only for employees from individualistic cultures, which highlights a second practical implication. In particular, managers should consider cultural background of their overqualified employees when implementing job redesign strategies. Because job redesign relates to changes in task structure and relations between different jobs, it takes time and effort to be implemented. Accordingly, knowing when and how jobs need to be re-designed for overqualified employees is practically important as organizations strive to know the best ways to arrange their resources to enhance effectiveness. Our findings appear to suggest that enhancing job autonomy in nations high in collectivism may be costly, yet it will unlikely yield same return on investment in managing overqualification as it does in individualistic cultures. As such, our study is consistent with assertions regarding the importance of national context: “not only must we include an immediate social context, we must deal with the international and cultural aspects of the social world. More than ever, understanding employee action requires knowledge of how action is related to the environment in which it is embedded” (Gibson and McDaniel 2010, p. 459). Context (i.e., culture here) therefore provides valuable information to guide managers how to use the existing practices to deal with the encountered issues such as overqualification. As employees’ work value is shaped by national culture, it is reasonable for organizations to consider individualism–collectivism at individual level and based on such information to decide whether increasing job autonomy would be an effective way to overcome negative impact of overqualification.

Nevertheless, the conventional job design perspective does not include job design factors that will shape psychological states to enhance the collective self such as maintaining harmonious relationships with others. Thus it provides a somewhat limited theoretical guidance for identifying job design factors that can reduce negative impact of overqualification for employees from collectivistic cultures. As people in collectivistic cultures concern more about their relationships with others, it is possible that more social or relational job design variables, such as interdependence, might be more meaningful moderators of the effects of overqualification in collectivist cultures. For example, relational work design in which individuals are connected with their beneficiaries (Grant 2007) might prove particularly useful in mitigating the effects of overqualification in collectivistic countries as it is a way that people in collectivistic cultures can sustain their values as being a person who can contribute to others or the society. Future research is required to examine these theoretical assertions. Altogether, as overqualification becomes a global issue, it is important to understand how organizations can minimize its negative impact on subjective well-being for employees in different cultures.

5 Limitations and Future Research

Despite the strengths of our research (a nationally representative sample), we note several limitations. First, although the use of archival data offers researchers a variety of benefits (e.g., representative sample), it also imposes challenges (Wang 2009). One of them concerns lack of control over how variables were measured. In the present study, we measured overqualification with only one item. Although this item captures the essence of the construct, by asking people to indicate whether they have more skills to do their jobs, we encourage future researchers to replicate the findings with multi-item scales of overqualification (e.g., Maynard et al. 2006). Also, we did not empirically examine the proposed mechanisms (e.g., increasing job meaning via job crafting) via which that job autonomy can mitigate the negative impact of overqualification for employees from individualistic cultures. The proposed mechanisms should be further examined as it helps not only to validate our propositions but also understand how an overqualified employee can do to deal with the suboptimal condition when having job autonomy at work.

Another limitation pertains to a somewhat small effect size of the three-way interaction (indicated by the pseudo-R 2 value) obtained in this study. Notably, small effect sizes are common in non-experimental studies (Champoux and Peters 1987). In addition, previous research showed that small effect sizes can still be of substantial practical importance (Abelson 1985). As such, we argue that our results are valuable in practical terms (Aguinis et al. 2010). In particular, our findings are noteworthy because we examined our research hypothesis in a large, nationally representative sample consisting of countries with distinct cultural values. Accordingly, we have a representative sample to detect culturally specific phenomena and a large sample to detect a small effect, which is more difficult to obtain with smaller samples. Moreover, we suggest this small effect is practically important because it provides actionable steps regarding management of overqualified people across culturally diverse countries. As job redesign has its own cost (Kelly 1980), our finding thus suggests that enhancing job autonomy is more rewarding to alleviate negative impact on subjective well-being at work owing to overqualification when it is applied in individualistic culture.

Further, our findings are based on European countries and thus their generalizability to other Western and non-Western countries may be limited. Although we have selected nations that represent highly individualistic and collectivistic cultures, future research will strengthen our results by replicating them in other countries that differ on common cultural values (Hofstede 2001). Further, we used nations as proxies for cultural values of individualism/collectivism and did not measure these values directly. Although this approach allowed us to analyze the proposed model across nations, it also opened up a possibility that other contextual factors (e.g., economic, political regime) might explain the present findings. Future research should include direct measures of individualism/collectivism as well as other cultural values when replicating the findings of the current research.

From a research design perspective, we used an archival data set in which all the data were measured via self-reports, suggesting that common method variance could have influenced the results (Podsakoff et al. 2003). It is important to mention, however, that common method variance cannot account for interactions (Evans 1985; Lai et al. in press; Siemsen et al. 2010). The cross-sectional research design also means we cannot rule out reverse causality as a potential explanation of our findings. For example, dissatisfaction with one’s job may lead one to report being overqualified. However, it seems unlikely that such a relationship would be stronger in the absence of job autonomy only in individualistic cultures, which was the more precise pattern of findings we reported here. We therefore believe the direction of causality we proposed is a more likely account of the findings. Nevertheless, we acknowledge that our focus in understanding the relationship between overqualification, subjective well-being at work, and job autonomy was in terms of a mechanical interaction rather than a dynamic interaction, which occurs over time (Endler and Parker 1992). It is possible that there are rather complex reciprocal relationships between overqualification, subjective well-being at work, and job autonomy that we did not capture. For example, lower job autonomy might lead to lower subjective well-being at work, which may lead an employee to over- or under-report true levels of their overqualification, and reinforce their negative job perception and attitudes in turn. Such a dynamic interaction process should be further examined in longitudinal studies.

Based on the results of our study, we offer possible avenues for future research beyond the methodological improvements highlighted above. One question is whether the beneficial role of job autonomy in improving work experience of overqualified people applies to other organizationally valued outcomes such as job performance and voluntary turnover. It is relatively well-known that overqualified employees tend to voluntarily quit their jobs as soon as they find a more suitable employment (e.g., Erdogan and Bauer 2009). Extrapolating from the current findings, it stands to reason that job autonomy likely improves the person-job fit of overqualified people. This improved match might increase retention rates of these workers; it could also discourage them from engaging in counterproductive work behaviors (Luksyte et al. 2011). Further, we showed that the effectiveness of job autonomy depends on cultural background of overqualified employees. What are other job design features that may prove useful in influencing work outcomes of overqualified workers whose cultural values differ from those of the current research? Most likely these interventions will also depend on personal characteristics of culturally diverse overqualified people. Future research should extend the current findings by incorporating personality, other cultural dimensions (e.g., power distance), and a wider range of job design characteristics into examining the quality of employment of overqualified workers.