“I Put In My Two Weeks Notice!” Employee Behavior and Wellness Prior to Exit

Abstract

Previous research has identified four distinct “proximal withdrawal states” among employees. That is, employees can generally be distinguished by their desire to stay or leave (e.g., “stayers” or “leavers”) and their control over this desire (e.g., “enthusiastic” versus “reluctant” staying or leaving). However, little is known about how these withdrawal states impact voluntary behavior and employee wellness. Thus, we examined these relationships in US (n = 516) and South Korea (n = 414). Latent profile analysis indicated that there are two profiles of stayers (i.e., embedded stayers and detached stayers) and two of seekers (i.e., dissatisfied seekers and script-driven seekers) in both samples. US and Korean profiles were consistent, except seekers differed by country based on planning, job seeking, and socio-emotional reasons. We also found that profiles differentially predicted work behaviors and wellness outcomes, such that embedded stayers showed most adaptive behavioral patterns, whereas script-driven seekers reported the least wellness.

Keywords

proximal withdrawal states theory latent profile analysis organizational citizenship behavior counterproductive work behavior wellness

Introduction

Employee turnover research has been important for both organizational researchers and practitioners. For example, the Society for Human Resource Management (SHRM) announced that 46% of HR managers regarded turnover among their top worries in 2016. Hiring and replacement expenses are estimated to cost 90%–200% of each exiting employee’s annual salary (Allen et al., 2010). Furthermore, personnel losses can disrupt a variety of productivity-related outcomes (e.g., job performance), decrease financial performance when departing staff take their talents and know-how to competitors, stymie workforce diversity, and demoralize other employees (Hom et al., 2017).

To better understand the employee turnover phenomena, proximal withdrawal states theory proposed that there are distinct employee withdrawal profiles among the population (Hom et al., 2012). Proximal withdrawal state is defined as a psychological state regarding (a) the employees’ preference for leaving versus staying, combined with (b) their volitional control over staying or leaving. Although proximal withdrawal states theory was developed to overcome the limitations of conventional turnover research (e.g., overemphasizing distal predictor-criterion relationships), little research addresses the consequences of dissatisfied or reluctant staying (e.g., Li et al., 2016; Liu & Raghuram, 2022). Furthermore, withdrawal research remains heavily focused on predicting turnover, to the neglect of investigations on the impact of withdrawal states on work behaviors and wellness before severing employment (Mai et al., 2016). Consequently, we test this phenomenon in the current study.

Specifically, we first attempt to extract and replicate withdrawal profiles (i.e., the empirical manifestation of a withdrawal state) in two countries (the US and South Korea). Although previous studies examined distinct withdrawal profiles, inferences are limited due to study designs that are cross-sectional (e.g., Woo & Allen, 2014) and retrospective (e.g., Li et al., 2016), as well as a lack of measurement comprehensiveness (e.g., using proxies such as turnover intentions; Liu & Raghuram, 2022). Then, we contribute to withdrawal profiling research by expanding the nomological network of withdrawal profiles by testing their behavioral and wellness outcomes. Finally, we investigate distinct employee withdrawal profiles in both collectivistic (i.e., South Korea) and individualistic (i.e., US) cultures (e.g., Oh et al., 2014; Ramesh & Gelfand, 2010). Given that the relationship between employee withdrawal and wellness has not been explored and that little is known about withdrawal states outside the US, this study can improve our understanding of the generalizability of withdrawal states, and the effect of these withdrawal profiles on voluntary behavior and employee wellness.

Proximal Withdrawal States Theory

Proximal withdrawal states theory differentiates employee withdrawal status into four patterns based on two domains, preference to stay or leave and control over staying or leaving (Hom et al., 2012). Combining these two dimensions, there are four categories, reflecting enthusiastic leavers (“I want to and can leave”), enthusiastic stayers (“I want to and can stay”), reluctant leavers (“I want to stay but must leave”), and reluctant stayers (“I want to leave but I cannot”). Although Hom et al. (2012) conceptually proposed additional withdrawal states (e.g., trapped stayers), empirical research show four types similar to those described above, including embedded stayers (“I want to stay and will stay because I love my job”), detached stayers (“I am not going to leave even though I sometimes fantasize about it”), script-driven seekers (“I will be leaving, and I have a clear plan”), and dissatisfied seekers (“I want to leave but I don’t have a tangible plan yet; ” Woo & Allen, 2014). The four types are slightly different from the original version of Hom et al.’s (2012) conceptualization (stayers versus leavers) because they reflect a mixture of intention to quit, job search behavior, and diverse reasons for staying/leaving (i.e., socio-emotional, economic, and external reasons) before exit. Nonetheless, this approach allows researchers to better capture employee withdrawal states via more comprehensive clustering indicators rather than focusing on retrospective categorization methods (Lee et al., 2017). However, this solution must be further replicated and validated using different sample characteristics and longitudinal design (Woo, 2019). Specifically, according to best practices in latent profile analysis, the replicability of a solution should be tested across different samples, contexts, and timepoints, as only 81.0% of cross-validations fully replicate the initial results (Spurk et al., 2020). Despite the importance of cross-validating the profile solutions, only 39.1% of studies have cross-validated a latent profile solution with another sample and only 6.5% with another time period (Spurk et al., 2020). Here, we test the validity of the four-profile solution in two samples, one of which uses a longitudinal design.

Hypothesis 1

Four latent profiles will be extracted in both US and South Korean employees that reflect embedded stayers, detached stayers, script-driven seekers, and dissatisfied seekers.

Furthermore, we investigate whether withdrawal profiles are similar across national cultures. In terms of employee retention, the effects of embeddedness on voluntary turnover are higher in individualistic cultures than collectivistic (e.g., Ramesh & Gelfand, 2010). This is because in a collectivistic culture, relational factors (e.g., social norms and family-related values) are typically more important than organizational embeddedness (Moon & Lee, 2021). For instance, collectivists tend to suppress their personal feelings that they perceive may hurt in-group harmony (Oyserman et al., 2002). On the other hand, rational factors are more important than relational factors in Euro-American cultures (Oh et al., 2014). Given that the withdrawal and turnover literature overrepresent individualistic cultures (Ramesh & Gelfand, 2010), testing withdrawal profiles in South Korea advances understanding regarding cultural similarities and differences in withdrawal profiles. To clarify, because latent profile analysis is not sample dependent, we expect to replicate the four-profile structure in both samples. However, there may be differences between the samples regarding the specific characteristics of each profile.

Research Question 1

Are there different characteristics of South Korea and US withdrawal profiles? If so, what are the characteristics of each profile?

Outcomes of Withdrawal Profiles

Intent to leave may result in behavioral changes prior to employee exit. For example, employees who have higher intent to leave are more likely to display negative (i.e., tardiness and abuse against others; Mai et al., 2016) and less likely to display positive (i.e., helping others) work behaviors (Kiazad et al., 2015). Likewise, reluctant stayers are psychologically detached and show greater neglect behavior during staying (Boswell et al., 2017). These findings are consistent with how withdrawal states are theorized to reflect motivations regarding employee leaving and staying (Woo & Allen, 2014). Consequently, each withdrawal state may exhibit distinct levels of voluntary behaviors, such as organizational citizenship behaviors (OCBs; e.g., helping others; Dalal, 2005) and counterproductive work behaviors (CWBs; e.g., damaging company property; Spector & Fox, 2010).

Although Hom et al. (2012) proposed that different withdrawal types may exhibit varying levels of OCBs and CWBs, research has focused largely on the distinct effects of withdrawal states on job attitudes, task performance, and turnover (Li et al., 2016; Liu & Raghuram, 2022). For example, initial empirical testing of proximal withdrawal states demonstrated that preference was more relevant factors than perceived control when it comes to predicting employees’ attitudes (Li et al., 2016). More specifically, enthusiastic stayers and reluctant leavers were similarly high on job satisfaction, job embeddedness, affective commitment, and continuance commitment, whereas enthusiastic leavers and reluctant stayers had relatively low on those criteria. On the contrary, in terms of employees’ voluntary turnover, predict validity was heavily depending on employees’ control over their preference for leaving or staying (i.e., conventional turnover framework was not applied in the profiles of reluctant leavers and stayers; Li et al., 2016). Due to these mixed results, a broadened perspective may be necessary to frame the relationship between withdrawal states and voluntary behaviors.

In particular, embedded stayers and dissatisfied seekers may exhibit OCBs and CWBs at expected (i.e., inverse) levels. Specifically, embedded stayers are willing to enact more OCBs and fewer CWBs, as they are motivated to invest resources to gain more resources (Lee et al., 2004). Conversely, dissatisfied seekers may exhibit more CWBs and fewer OCBs because they are motivated to withdraw their resources at work (Kiazad et al., 2015). In addition, according to job embeddedness theory (Mitchell et al., 2001), highly embedded employees (e.g., engaged stayers or embedded engaged stayers) have more instrumental resources (e.g., organizational fit and links), which allow them to achieve better performance, motivate to helping others, and suppress CWBs (e.g., Kiazad et al., 2015).

Hypothesis 2

Embedded stayers will display more OCBs and fewer CWBs, and dissatisfied seekers will display fewer OCBs and more CWBs, relative to other profiles.

On the other hand, script-driven seekers and detached stayers are more complex than dissatisfied seekers and embedded stayers in that these profiles exhibit mixed preferences for leaving and staying, coupled with at least one reason for staying (e.g., high turnover intention with strong economic reasons for staying in the organization; Woo & Allen, 2014). As Hom et al. (2012) suggested, these withdrawal states may consequently enact atypical levels of voluntary behaviors (e.g., high enactment of both OCBs and CWBs). That is, OCB and CWB are typically negatively correlated (e.g., meta-analytic r = −.32, Dalal, 2005), but this is not likely the case for script-driven seekers and detached stayers. Instead, because of contractual conditions and disengaged job attitudes, detached stayers exhibit low CWBs and withdrawal behaviors (e.g., Benson et al., 2004), as well as low OCBs and job performance. Employees exhibit less OCBs or favorable discretionary behaviors when employees felt that their employer violated the terms of the employment contract (Robinson & Morrison, 1995). On the contrary, in order to maintain a good reputation, script-driven seekers may exhibit more OCBs than usual when they are preparing for leaving in the future (Hom et al., 2012). This is because employees with turnover intentions tend to cut back on their positive contextual performance before leaving as a coping strategy (Boswell et al., 2017)

At the same time, script-driven seekers are less motivated to engage in CWBs toward the organization, as they are leaving in order to follow their career paths, rather than because of image violation (negative shocks; Lee & Mitchell, 1994). On the contrary, according to expectancy theory (Vroom, 1964), when employees have high turnover intentions, they are more likely to focusing on short-term economic and instrumental exchanges and less likely to focusing on longer-term interpersonal relationships. For instance, Mai et al. (2016) found that turnover intentions predicted a strong transactional contract orientation and a weak relational contract orientation, which then predicted fewer OCBs and more deviant behaviors. As script-driven seekers regard themselves as progressing towards turnover, they may emotionally detach themselves from their organizations (Colarelli & Beehr, 1993). This psychological distance may drive script-driven seekers engage in deviant or unethical behaviors to restore this discrepancy.

Hypothesis 3

Script-driven seekers and detached stayers will display inconsistent voluntary behaviors: more OCBs with more CWBs, or less OCBs and less CWBs.

Furthermore, due to their perceived external and internal control over staying and leaving, employees may experience specific job demands from their workplace. When individuals psychologically detach themselves from organization, they are more “physically uninvolved in tasks, cognitively unvigilant, and emotionally disconnected from others” (Kahn, 1990, p. 702). The job search process may intensify detachment from the organization, and this detachment engenders adverse reactions directed toward the organizational members (e.g., employers; Boswell et al., 2017). Resultantly, if employees dissatisfied with the current employment situation remain in the current situation, they would experience the negative wellness (i.e., higher emotional exhaustion and lower engagement). In other words, employees who cannot achieve desired career changes may experience distinct cognitive and emotional processes (e.g., cognitive dissonance; Festinger, 1957), and such misfit between current status and desired status results in emotional exhaustion (Liu & Raghuram, 2022). Indeed, trapped stayers (i.e., employees embedded in workplace because of external reasons) demonstrate lower sleep quality and higher emotional exhaustion because they cannot control their high job insecurity (e.g., Allen et al., 2016). However, there are few studies investigating employee wellness prior to exit, even though researchers proposed a high cost of employees’ withdrawal state (e.g., Boswell et al., 2017). Thus, to extend the criterion domain prior to leaving, we examine whether withdrawal profiles will exhibit dissimilar behaviors and wellness.

Hypothesis 4

Employee withdrawal states differentiate levels of wellness status.

Method

Sample and Procedure

Data collection for this project received Institutional Review Board approval at Central Michigan University (#2021-298 and #2021-403). In order to improve generalizability and obtain a heterogeneous sample representing diverse job titles, industries, and organizational conditions, we used multiple recruitment strategies. First, participants in the US sample were recruited using Amazon’s Mechanical Turk (MTurk), which is an online data crowd-sourcing platform to gain human participation to complete various tasks. Although some researchers have raised concerns about the motivation of those who contribute to online survey panels, the heterogeneity of the online panel data was appealing in order to achieve a more generalizable effect reflecting withdrawal profiles across a variety of organizational cultures and contexts (e.g., Landers & Behrend, 2015). Empirical research shows that online panel data is of equal or higher psychometric quality than organizational convenience samples when adhering to best practices, and concerns regarding the quality of data from MTurk samples are overestimated and inappropriately applied (Landers & Behrend, 2015).

Via MTurk, a recruitment ad was shown. Participants who responded to the ad completed two waves of data collection with a 2-month time lag. Following the best practices recommendations, only full-time workers in US with at least 90% approval ratings from previous MTurk assignments were eligible to participate (Moon et al., 2023). Those who participated in Time 1 received $1.00, and those who participated Time 2 received $1.25. The Time 1 sample included a total of 700 working adults residing within US. Of these working adults, four were removed for failing to meet the minimum qualifications of satisfactory completion of surveys (i.e., those who answered two or more quality control items incorrectly; e.g., Landers & Behrend, 2015). Time 2 included 516 of the 696 working adults who completed the survey at Time 1, resulting in a response rate of approximately 74.1%. Profile extraction variables (i.e., withdrawal profiling scale) in US sample were measured in Time 1, and outcomes (i.e., voluntary behaviors and outcomes) were measured in Time 2. The US sample consisted of 258 (50.0%) males with a mean age of 41.48 years (SD = 11.80). The average organizational tenure of employees was 8.65 years, and employees are mostly Caucasian (74.8%). Participants were from diverse job industries including health care (13.2%), technology (13%), consumer services (e.g., retail, media; 11.4%), and finance (12.8%), among others.

For the South Korean sample, 300 participants were recruited and paid by SouthernPost corporation via online survey in South Korea. An additional 140 participants (unpaid) were recruited via snowball sampling using email and social apps (e.g., KakaoTalk). A total of 440 participants were recruited, but of these working adults, 26 were removed for failing to two or more quality control items. The final South Korean sample consisted of 193 (46.6%) males with the mean age of 39.53 years (SD = 8.74). The average organizational tenure of employees was 9.29 years, and employees represent professionals (76.3%), manufacturing and technology (6.1%), research and development (9.5%), and sales industry (4.1%). In this sample, all variables were measured at one time point.

Measures

All scales were previously validated. Descriptive statistics and coefficient alphas are reported in Table 1. For the South Korean survey, all questionnaire materials were translated into Korean and back-translated in line with best practices (e.g., using a separate translator and back translator; Brislin, 1970). For all measures except the withdrawal profile questionnaire, response anchors were 1 (strongly disagree) to 5 (strongly agree).

Table 1.

Means, Standard Deviations, and Correlations of All US and South Korean Variables.

	US (N = 516)										South Korea (N = 414)
Variable	M	SD	1	2	3	4	5	6	7	8	M	SD
1. Gender	1.52	.54	–	−.10^*	−.06	−.15^**	.05	−.19^**	−.02	.10^*	1.56	.55
2. Age	41.48	11.80	.01	–	.68^**	−.06	.01	−.09	.18^**	−.14^**	39.53	8.74
3. Tenure	8.65	7.70	−.03	.51^**	–	.01	.06	−.10^*	.18^**	−.06	9.29	8.00
4. Education	3.85	.88	−.15^**	−.10^*	−.04	–	.04	.08	.05	.01	3.08	.48
5. OCB	3.79	.61	.07	.13^**	.14^**	−.01	(.71/.72)	−.20^**	.32^**	−.03	3.50	.58
6. CWB	1.53	.58	−.03	−.24^**	−.08	.00	−.26^**	(.87/.77)	−.18^**	.07	1.96	.52
7. WE	3.38	.98	−.03	.14^**	.09^*	.04	.39^**	−.30^**	(.87/.79)	−.27^**	3.03	.71
8. EE	2.81	1.15	.10^*	−.12^**	−.06	−.02	−.27^**	.36^**	−.51^**	(.95/.89)	3.16	.86

Note. US means, standard deviations, and correlations are below the diagonal; South Korean means, standard deviations, and correlations are above the diagonal. Reliabilities are along the diagonal; US is before the slash; OCB = Organizational citizenship behaviors; CWB = Counterproductive work behavior; WE = Work engagement; EE = Emotional exhaustion. *p < .05. **p < .01.

Withdrawal profiling was measured by Woo and Allen’s (2014) 19-item seeker-stayer profiling scale. Example items are “How often do you think about leaving your organization?” (turnover intention) and “Are you planning to leave your organization in the near future?” (job search behavior; 1 = yes, 0 = no). This scale also asks participants to indicate the degree to which items such as “sense of obligation” (socio-emotional), “pay satisfaction” (economic reasons), and “family or relationship” (external reasons) affected their decisions to stay or leave. This scale was previously correlated with job satisfaction, motivation (e.g., normative and moral), and personality (e.g., trait affect; Woo & Allen, 2014).

OCB. OCB was measured using Henderson et al. (2020) scale. After psychometric scrutiny using both classical testing theory and item response theory (Henderson et al., 2020), this scale includes six items, such as “I helped others with heavy workloads.”

CWB. CWB was measured using an 8-item scale (e.g., “put little effort into your work”) used by Matthews and Ritter (2016), which they validated against mistreatment.

Emotional exhaustion. Five items (e.g., “I feel emotionally drained at work”) from the Maslach Burnout Inventory (Maslach et al., 1996) were used. This scale has received validation of its factor structure and measurement invariance (e.g., Bakker et al., 2002).

Work engagement. Employee’s work engagement was measured by three items from Schaufeli et al. (2019). This scale was validated against nine well-being indicators and six strains (Schaufeli et al., 2019). An example item includes “I am immersed in my work.”

Results

Identification of Employee Withdrawal Profiles

The purpose of latent profile analysis is to identify types, groups, or classes (i.e., profiles) of people based on personal or environmental characteristics (i.e., indicators). This profile is considered to be an unobserved categorical latent variable, evidenced by the pattern of indicators. Although there are concerns with categorizing people based on group membership (e.g., loss of power and granularity), the profile approach is beneficial because more indicators can be included than would be reasonable in alternative methods (e.g., a six-way multiple regression interaction). To advance our understanding of how employee withdrawal states are different across two countries, we use this person-centered approach. Previous studies on multiple job holding motivations and vocational research have similarly adopted person-centered approach and brought new insights into their literature (e.g., Campion & Csillag, 2022; Spurk et al., 2020). Thus, we followed best practices as described in previous research (e.g., Nylund et al., 2007; Spurk et al., 2020).

Table 1 presents descriptive statistics and correlations among the variables. To test the presence of multiple employee withdrawal profiles, we conducted latent profile analysis based on turnover intentions, job search behaviors, and reasons for staying or leaving. Table 2 provides the fit index for possible latent class structures. We selected the four-profile solution because it demonstrated the best fit statistics compared to other profile solutions. Although the fifth-profile solution had slightly lower fit indices (e.g., AIC and BIC) compared to the fourth-profile solution in both samples, the Lo et al. (2001) (LMR) test was nonsignificant. LMR is believed to be the most appropriate test (Nylund et al., 2007), so we retained the more parsimonious four-profile structure in the US and South Korean samples.

Table 2.

Fit Statistics for Withdrawal Profile Structures in US and South Korea.

No. of Profiles	AIC	BIC	SSA-BIC	LMR(p)	Entropy
US
2	8457.74	8544.08	8483.75	0	.89
3	8281.46	8404.15	8318.42	.0014	.76
4	8165.93	8324.97	8213.84	.0022	.77
5	8122.29	8317.68	8181.15	.0875	.77
6	8076.19	8308.19	8146.25	.548	.76
South Korea
2	4574.17	4650.66	4590.37	0	.86
3	4456.61	4565.31	4479.64	.3134	.82
4	4399.59	4540.49	4429.43	.0421	.75
5	4343.53	4516.64	4380.19	.2271	.79
6	4308.74	4514.06	4352.23	.0035	.82

Note. AIC = Akaike information criteria; BIC = Bayesian information criteria; SSA–BIC = Sample-size–adjusted BIC; LMR = Lo et al. (2001) test.

These four profiles in both the US and South Korean samples resembled the profiles from previous studies (see Figures 1 and 2; Woo & Allen, 2014). The first group (n = 172 for US and n = 127 for South Korea) was characterized by the lowest level of turnover intention and job search behavior and had lower scores for all three reasons to stay (i.e., socio-emotional, economic, and external). Thus, we labeled them “detached stayers.” The second group (n = 193 for US and n = 157 for South Korea) had similar withdrawal profiles with first group, but they exhibited the highest scores on all reasons for staying. Therefore, we labeled this group “embedded stayers.” The third group (n = 105 for US and n = 96 for South Korea) showed the highest level of turnover intentions and job search behaviors, but also moderate scores for reasons to stay. We label this group as “script-driven seekers” because they were distinguished by their high likelihood of having an alternative offer in hand. The fourth group (n = 46 for US and n = 34 for South Korea) showed the second highest turnover intentions and job search behaviors among the groups, but were lower in all three reasons to stay than the other groups. Therefore, we label this group “dissatisfied seekers.” Although the profile compositions differed somewhat between the US and South Korea, the general structure replicated in both samples, supporting Hypothesis 1. This also provides insight into Research Question 1, showing that the profiles mostly generalized across cultures, with the exception that socio-emotional reasons for staying among dissatisfied seekers in South Korea were prominently stronger than in the US.

To compare the behavioral and well-being outcomes of withdrawal states, we use ANOVA to compare the profiles (Table 3). When it comes to voluntary behaviors, embedded stayers had highest OCBs (M_US = 4.02; M_{South Korea} = 3.73) among groups, and relatively low CWBs (M_US = 1.35; M_{South Korea} = 1.91). In terms of OCBs, script-driven seekers displayed similar amounts of OCBs (M_US = 3.71; M_{South Korea} = 3.59) relative to embedded stayers, but higher levels of CWB (M_US = 1.82; M_{South Korea} = 1.98). In general, dissatisfied seekers had generally higher CWBs (M_US = 1.67; M_{South Korea} = 2.49) and lowest OCBs (M_US = 3.37; M_{South Korea} = 3.16). Overall, these results support Hypothesis 2 and 3.

Table 3.

Validation of Mean and Significance by Four Profile Types for Distal Outcomes in US and South Korea.

	1. Embedded Stayers	2. Detached Stayers	3. Dissatisfied Seekers	4. Script-Driven Seekers	Overall Test	Significance Test
US
OCB	4.02	3.69	3.37	3.71	X² (df = 3) = 55.938^**	1 = 4>2 = 3
CWB	1.35	1.51	1.67	1.82	X² (df = 3) = 20.139^**	4 = 3>1 = 2
Emot. Exhaustion	2.64	2.35	3.20	3.69	X² (df = 3) = 79.242^**	4>3>1 = 2
Work Engagement	3.87	3.52	2.83	2.75	X² (df = 3) = 81.590^**	1>2>3 = 4
South Korea
OCB	3.73	3.24	3.16	3.59	X² (df = 3) = 33.985^**	1 = 4>2 = 3
CWB	1.91	1.86	2.49	1.98	X² (df = 3) = 195.362^**	3>1 = 2 = 4
Emot. Exhaustion	3.09	2.90	2.94	3.75	X² (df = 3) = 52.616^**	4>1 = 2 = 3
Work Engagement	3.19	2.87	2.90	2.98	X² (df = 3) = 4.032^*	1>2 = 3

Note. OCB = Organizational citizenship behaviors; CWB = Counterproductive work behavior. Emot. = Emotional.

In terms of wellness, script-driven seekers reported the poor wellness (i.e., highest emotional exhaustion) compared with other profiles in both countries (M_US = 3.69; M_{South Korea} = 3.75). Dissatisfied seekers reported relatively higher emotional exhaustion (M_US = 3.20; M_{South Korea} = 2.94). When it comes to work engagement, embedded stayers had highest scores in both samples (M_US = 3.87; M_{South Korea} = 3.19). Detached stayers had relatively high levels of wellness (i.e., emotional exhaustion: M_US = 2.35; M_{South Korea} = 2.90; work engagement: M_US = 3.52; M_{South Korea} = 2.87) compared to seekers. These results provide the evidence supporting Hypothesis 4, as withdrawal profiles relate to different levels of wellness.

Discussion

Based on proximal withdrawal states theory, this study investigates distinct employee withdrawal states and their differences in behavioral and wellness outcomes. We identified the same profile solution, including two classes of stayers and two classes of job seekers, in two countries. We also found that there are cultural differences of withdrawal profiles among two countries, such that the socio-emotional reasons for staying of dissatisfied seekers in South Korea were higher than in the US. Finally, we examined the outcomes of employee withdrawal profiles, which can expand an organization’s ability to identify employees who may engage in OCB or CWB, or who are at risk for burnout, even if they do not exit. Specifically, embedded stayers showed high OCBs in both countries, whereas dissatisfied seekers had high CWBs and low OCBs across the two countries. When it comes to employee wellness, script-driven seekers were particularly emotionally exhausted, whereas detached stayers showed higher levels of both emotional exhaustion and work engagement.

Theoretical Implications

Although proximal withdrawal states theory initially indicated that different subtypes of withdrawal states may emerge across organization and cultures, we found that four withdrawal profiles consistently emerged across countries. There were slightly different degrees of endorsement for reasons to stay or leave from dissatisfied seekers between US and South Korean employees, but the four-profile solutions were otherwise similar. For example, although there were slightly different outcome scores among the two countries (e.g., OCB of embedded stayers in US [4.02] and South Korea [3.73]), statistical differences of values among withdrawal profiles were similar across the two countries. However, cross-cultural differences between South Korea and the US may still influence other outcomes of withdrawal states. For instance, because of collectivistic values, South Korean employees may have trouble viewing themselves as detached stayers or dissatisfied seekers when their cultural values conflict with their preference to stay in their organization (e.g., normative obligation to the boss may conflict with poor job satisfaction as a reason to leave). That is, although few cross-cultural differences between withdrawal profile solutions may exist, employees’ withdrawal states, when conflicting with their self-identities and values, may result in different behaviors. Additionally, future research should validate this profile solution in other countries, particularly those that have vastly different labor laws or norms. For example, in Germany, certain employees must give as much as seven months’ notice before leaving their job. Detached stayers might be particularly demoralized knowing they must later wait out a seven-month notice. It is also possible that profile configurations change over time in a long notice period.

This research also suggests that organizational theories can be further enriched by considering the link between distinct withdrawal profiles in the organization, behavior, and wellness status. For voluntary behaviors from distinct withdrawal states, we contributed to the literature of OCB and CWB relationship because there were distinct patterns of these behaviors. Previous meta-analytic studies have found small but positive relationships between OCB and CWB (e.g., Dalal, 2005). For those reasons, researchers proposed that the same employee might display both OCB and CWB via certain processes (e.g., moral licensing and moral cleansing theories; Yam et al., 2017) or due to methodological issues (e.g., Spector & Fox, 2010). By identifying distinct withdrawal profiles among employees, we also found that some withdrawal profiles (i.e., script-driven seekers and detached stayers) show independent levels of OCBs and CWBs. A possible explanation is that script-driven seekers are motivated to show OCBs to increase their visibility and better employability in the next organization before they finally quit. Thus, their OCBs are sort of impression management tactics to earn stronger reference letters or avoid “burning their bridges” (Hom et al., 2012). In contrast, high CWBs from both dissatisfied seekers and script-driven seekers can possibly be explained as an emotional process to “passively and indirectly cope with the emotion” (Spector & Fox, 2002, p. 274). For instance, employees experiencing emotional exhaustion or negative feelings might restore their energy by engaging in CWB to increase feelings of control. In addition, if dissatisfied seekers perceive unfairness, they might retaliate against their organization via CWB (Spector & Fox, 2002).

Furthermore, we explored how distinct withdrawal states relate to employee wellness. Across two samples, embedded stayers showed better wellness, whereas script-driven seekers consistently demonstrated worse wellness. We suspect that because script-driven seekers are relatively close to their progress to organizational exit than other profiles, they experience highest cognitive dissonance status between their desired status and current status. Thus, they are more likely to feel trapped by organization. As a result, their wellness status is somewhat lowered until the impediment for leaving is removed. In addition, if script-driven seekers perceive themselves as trapped stayers due to external reasons (e.g., family related), they experience more stress because they cannot control their situation (e.g., Allen et al., 2016). Therefore, by expanding the wellness domain of withdrawal states theory, we contribute to nomological network of withdrawal literature.

Practical Implications

This research has the potential to serve as a catalyst for practical advances in employee withdrawal and retention management. For one, profiling research might address the questions that remain among different employee withdrawal types in the organization. First, existence of distinct withdrawal profiles supports the ineffectiveness of a one-size-fits-all employee retention management. Based on distinct withdrawal profiles and outcomes among employees, organization can initiate target-oriented interventions, such as separate performance management for each group. In this regard, our study also has important implications for career counseling practice. The results show that both detached stayers in US and South Korea report lower work engagement and OCBs. For example, it is helpful for vocational counselors to focus on these characteristics and to initiate needs exploration and motivation building training for detached stayers when warranted. Also, career counselors could benefit from tracking the withdrawal profiles of employees through additional mentoring or occasional career development meetings.

Our findings that dissatisfied seekers exhibited fewer OCBs and more CWBs than other profiles imply that not all retention efforts are helpful. Moreover, script-driven seekers demonstrated more OCBs and CWBs than other profiles, whereas detached stayers were generally low on both behaviors. As these distinct progressions toward turnover lead to dissimilar behavioral patterns, managers may consider changing their withdrawal profiles or reasons for staying/leaving. For example, for those who are disgruntled with companies’ compensation policy, organization can initiate stock option program for stayers to increase their motivation (Larcker et al., 2010). Thus, managers should identify the true motives for their staying and leaving via carrying out regular stay interviews for further performance management as well as retention management (Grant, 2022). In addition, companies may also utilize machine learning technique to build up their customized withdrawal profiles based on their accumulated exit interviews and performance records (Sajjadiani et al., 2019). This big-data approach allows companies to identify future withdrawal profiles as well as provide information about personnel decisions (e.g., hiring and promotion).

From a health viewpoint, the wellness status difference among withdrawal profiles provides a vehicle for probing the loosely coupled but important connections among employee withdrawal profiles and their wellness. Specifically, script-driven seekers had worst wellness, such that companies may choose to target health interventions towards this group as soon as they can. Furthermore, if script-driven seekers’ main reason for leaving is health-related issue, organization can benefit from health intervention for their retention as well.

Limitations and Future Research Directions

A limitation of this research is we could not assess the link between latent profiles and later turnover behavior. Given that script-driven seekers were more likely to have an alternative job offer in hand relative to other profiles, they readily leave their organization with specific destination. However, because dissatisfied seekers often impulsively quit their job when they experience negative organizational “shock” events (e.g., conflict with coworkers), it is hard to predict exact turnover speed among profiles (Hom et al., 2012). Future research using survival analysis may better predict departure speeds.

In terms of research design, although our data were collected at two time points in the US sample, we did not use a true longitudinal design. For more robust testing, we encourage future research that includes multiple time points and diverse sample characteristics to validate our profile extractions. Also, we exclusively rely on the self-report measures. Because survey respondents were informed that their data are confidential and researchers were unconnected with their organizations, self-reports can be the best way of assessing their voluntary behaviors (especially CWB; Berry et al., 2012) and wellness status (Johnson & Spector, 2007). In addition, our additional 140 South Korean employees were recruited via snowball sampling technique, suggesting that they may not represent the general Korean workers. However, this 140 were not the majority of total sample size in our data, and previous meta-analytic research demonstrated that correlations from snowball samples were only moderately lower than from non-snowball samples (Wheeler et al., 2012). Nevertheless, we encourage future research use probability samples with various measurement techniques, such as company records and other-reported methods.

Additionally, future research may consider testing profiles with multiple time frames, to investigate withdrawal transition over time. For example, script-driven seekers may become detached stayers if they experience repetitive frustration from the job search process, especially in bad economic situations. This approach allows researchers to understand why employees become “stuck” in the organization, and how organizations can intervene in the transition process. Finally, future researchers might explore how withdrawal states result in employees become alumni (i.e., those who would not consider returning to their former organization) versus boomerangs (i.e., those who left the organization but would consider returning to their former organization). As previous research found that leaving due to personal shocks or external reasons (other than dissatisfaction) may allow employees to remain open about coming back to a previous job (Shipp et al., 2014), dissatisfied seekers may become alumni, whereas script-driven seekers may “boomerang” in the long term.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the SouthernPost, Korean Psychological Association (SouthernPost Research Funding Awards), and the Office of Research and Graduate Studies at Central Michigan University (Student Endeavors Grant).

Data Availability

The data that support the findings of this study are available from the corresponding author, Young-Kook Moon, upon reasonable request.

ORCID iD

Young-Kook Moon

Author Biographies

Young-Kook Moon is a doctoral graduate student in the Psychology Department at the Central Michigan University. His research program involves employee turnover, occupational health psychology, and individual differences. He loves watching basketball game, dramas, and movies.

Dr. Kimberly E. O’Brien received her PhD in Industrial/Organizational Psychology from the University of South Florida in 2008 and is now a professor at Central Michigan University. Her research program involves organizational citizenship behavior, counterproductive work behavior, mentoring, and job stress. She loves her family, all animals, dance, and any type of art.

References

Allen

D. G.

Bryant

P. C.

Vardaman

J. M.

(2010). Retaining talent: Replacing misconceptions with evidence-based strategies. Academy of Management Perspectives, 24(2), 48–64. https://doi.org/10.5465/amp.2010.51827775

Allen

D. G.

Peltokorpi

Rubenstein

A. L.

(2016). When “embedded” means “stuck”: Moderating effects of job embeddedness in adverse work environments. The Journal of Applied Psychology, 101(12), 1670–1686. https://doi.org/10.1037/apl0000134

Bakker

A. B.

Demerouti

Schaufeli

W. B.

(2002). Validation of the Maslach burnout inventory-general survey: An internet study. Anxiety, Stress and Coping, 15(3), 245–260. https://doi.org/10.1080/1061580021000020716

Benson

G. S.

Finegold

Mohrman

S. A.

(2004). You paid for the skills, now keep them: Tuition reimbursement and voluntary turnover. Academy of Management Journal, 47(3), 315–331. https://doi.org/10.5465/20159584

Berry

C. M.

Carpenter

N. C.

Barratt

C. L.

(2012). Do other-reports of counterproductive work behavior provide an incremental contribution over self-reports? A meta-analytic comparison. The Journal of Applied Psychology, 97(3), 613–636. https://doi.org/10.1037/a0026739

Boswell

W. R.

Gardner

R. G.

Wang

(2017). Is retention necessarily a win? Outcomes of searching and staying. Journal of Vocational Behavior, 98, 163–172. https://doi.org/10.1016/j.jvb.2016.11.006

Brislin

R. W.

(1970). Back-translation for cross-cultural research. Journal of Cross-Cultural Psychology, 1(3), 185–216. https://doi.org/10.1177/135910457000100301

Campion

E. D.

Csillag

(2022). Multiple jobholding motivations and experiences: A typology and latent profile analysis. The Journal of Applied Psychology, 107(8), 1261–1287. https://doi.org/10.1037/apl0000920

Colarelli

S. M.

Beehr

T. A.

(1993). Selection out: firings, layoffs, and retirement. In Schmitt

Borman

(Eds), Personnel selection in organizations (pp. 341–384). Jossey-Bass.

10.

Dalal

R. S.

(2005). A meta-analysis of the relationship between organizational citizenship behavior and counterproductive work behavior. The Journal of Applied Psychology, 90(6), 1241–1255. https://doi.org/10.1037/0021-9010.90.6.1241

11.

Festinger

(1957). A theory of cognitive dissonance. Stanford University Press. 2. https://doi.org/10.1515/9781503620766

12.

Grant

A. M.

(2022). The not-so-great resignation. https://podcasts.apple.com/in/podcast/the-not-so-great resignation/id1346314086?i=1000557193263

13.

Henderson

A. A.

Foster

G. C.

Matthews

R. A.

Zickar

M. J.

(2020). A psychometric assessment of OCB: Clarifying the distinction between OCB and CWB and developing a revised OCB measure. Journal of Business and Psychology, 35(6), 697–712. https://doi.org/10.1007/s10869-019-09653-8

14.

Hom

P. W.

Lee

T. W.

Shaw

J. D.

Hausknecht

J. P.

(2017). One hundred years of employee turnover theory and research. The Journal of Applied Psychology, 102(3), 530–545. https://doi.org/10.1037/apl0000103

15.

Hom

P. W.

Mitchell

T. R.

Lee

T. W.

Griffeth

R. W.

(2012). Reviewing employee turnover: Focusing on proximal withdrawal states and an expanded criterion. Psychological Bulletin, 138(5), 831–858. https://doi.org/10.1037/a0027983

16.

Johnson

H. A. M.

Spector

P. E.

(2007). Service with a smile: Do emotional intelligence, gender, and autonomy moderate the emotional labor process? Journal of Occupational Health Psychology, 12(4), 319–333. https://doi.org/10.1037/1076-8998.12.4.319

17.

Kahn

W. A.

(1990). Psychological conditions of personal engagement and disengagement at work. Academy of Management Journal, 33(4), 692–724. https://doi.org/10.5465/256287

18.

Kiazad

Holtom

B. C.

Hom

P. W.

Newman

(2015). Job embeddedness: A multifoci theoretical extension. The Journal of Applied Psychology, 100(3), 641–659. https://doi.org/10.1037/a0038919

19.

Landers

R. N.

Behrend

T. S.

(2015). An inconvenient truth: Arbitrary distinctions between organizational, Mechanical Turk, and other convenience samples. Industrial and Organizational Psychology, 8(2), 142–164. https://doi.org/10.1017/iop.2015.13

20.

Larcker

D. F.

McCall

Tayan

(2010). Equity on demand: Netflix approach to compensation, Case studies. Stanford Graduate School of Business.

21.

Lee

T. W.

Hom

P. W.

Eberly

M. B.

Mitchell

T. R.

Mitchell

T. R.

(2017). On the next decade of research in voluntary employee turnover. Academy of Management Perspectives, 31(3), 201–221. https://doi.org/10.5465/amp.2016.0123

22.

Lee

T. W.

Mitchell

T. R.

(1994). An alternative approach: The unfolding model of voluntary employee turnover. Academy of Management Review, 19(1), 51–89. https://doi.org/10.5465/amr.1994.9410122008

23.

Lee

T. W.

Mitchell

T. R.

Sablynski

C. J.

Burton

J. P.

Holtom

B. C.

(2004). The effects of job embeddedness on organizational citizenship, job performance, volitional absences, and voluntary turnover. Academy of Management Journal, 47(5), 711–722. https://doi.org/10.5465/20159613

24.

J. J.

Lee

T. W.

Mitchell

T. R.

Hom

P. W.

Griffeth

R. W.

(2016). The effects of proximal withdrawal states on job attitudes, job searching, intent to leave, and employee turnover. The Journal of Applied Psychology, 101(10), 1436–1456. https://doi.org/10.1037/apl0000147

25.

Liu

Raghuram

(2022). The effects of latent withdrawal profiles on employee turnover, destinations and job performance. Human Resource Management Journal, 32(2), 384–405. https://doi.org/10.1111/1748-8583.12362

26.

Mendell

N. R.

Rubin

D. B.

(2001). Testing the number of components in a normal mixture. Biometrika, 88(3), 767–778. https://doi.org/10.1093/biomet/88.3.767

27.

Mai

K. M.

Ellis

A. P. J.

Christian

J. S.

Porter

C. O. L. H.

(2016). Examining the effects of turnover intentions on organizational citizenship behaviors and deviance behaviors: A psychological contract approach. The Journal of Applied Psychology, 101(8), 1067–1081. https://doi.org/10.1037/apl0000115

28.

Maslach

Jackson

S. E.

Leiter

M. P.

(1996). MBI manual (3rd ed.). Consulting Psychologists Press.

29.

Matthews

R. A.

Ritter

K. J.

(2016). A concise, content valid, gender invariant measure of workplace incivility. Journal of Occupational Health Psychology, 21(3), 352–365. https://doi.org/10.1037/ocp0000017

30.

Mitchell

T. R.

Holtom

B. C.

Lee

T. W.

Sablynski

C. J.

Erez

(2001). Why people stay: Using job embeddedness to predict voluntary turnover. Academy of Management Journal, 44(6), 1102–1121. https://doi.org/10.5465/3069391

31.

Moon

Lee

(2022). Expansion of job embeddedness: Scale validation research of family embeddedness. Korean Journal of Industrial and Organizational Psychology, 34(4), 723–750. https://doi.org/10.24230/kjiop.v34i4.723-750

32.

Moon

Y. K.

O’Brien

K. E.

Mann

K. J.

(2023). The role of extraversion in the great resignation: A burnout-quitting process during the pandemic. Personality and Individual Differences, 205, 112074. https://doi.org/10.1016/j.paid.2022.112074

33.

Nylund

K. L.

Asparouhov

Muthén

B. O.

(2007). Deciding on the number of classes in latent class analysis and growth mixture modeling: A Monte Carlo simulation study. Structural Equation Modeling: A Multidisciplinary Journal, 14(4), 535–569. https://doi.org/10.1080/10705510701575396

34.

I. S.

Guay

R. P.

Kim

Harold

C. M.

Lee

J. H.

Heo

C. G.

Shin

K. H.

(2014). Fit happens globally: A meta‐analytic comparison of the relationships of person–environment fit dimensions with work attitudes and performance across east Asia, Europe, and north America. Personnel Psychology, 67(1), 99–152. https://doi.org/10.1111/peps.12026

35.

Oyserman

Coon

H. M.

Kemmelmeier

(2002). Rethinking individualism and collectivism: Evaluation of theoretical assumptions and meta-analyses. Psychological Bulletin, 128(1), 3–72. https://doi.org/10.1037/0033-2909.128.1.3

36.

Ramesh

Gelfand

M. J.

(2010). Will they stay or will they go? The role of job embeddedness in predicting turnover in individualistic and collectivistic cultures. The Journal of Applied Psychology, 95(5), 807–823. https://doi.org/10.1037/a0019464

37.

Robinson

S. L.

Morrison

E. W.

(1995). Psychological contracts and OCB: The effect of unfulfilled obligations on civic virtue behavior. Journal of Organizational Behavior, 16(3), 289–298. https://doi.org/10.1002/job.4030160309

38.

Sajjadiani

Sojourner

A. J.

Kammeyer-Mueller

J. D.

Mykerezi

(2019). Using machine learning to translate applicant work history into predictors of performance and turnover. The Journal of Applied Psychology, 104(10), 1207–1225. https://doi.org/10.1037/apl0000405

39.

Schaufeli

W. B.

Shimazu

Hakanen

Salanova

De Witte

(2019). An ultra-short measure for work engagement: The UWES-3 validation across five countries. European Journal of Psychological Assessment, 35(4), 577–591. https://doi.org/10.1027/1015-5759/a000430

40.

Shipp

A. J.

Furst‐Holloway

Harris

T. B.

Rosen

(2014). Gone today but here tomorrow: Extending the unfolding model of turnover to consider boomerang employees. Personnel Psychology, 67(2), 421–462. https://doi.org/10.1111/peps.12039

41.

Spector

P. E.

Fox

(2002). An emotion-centered model of voluntary work behavior: Some parallels between counterproductive work behavior and organizational citizenship behavior. Human Resource Management Review, 12(2), 269–292. https://doi.org/10.1016/s1053-4822(02)00049-9

42.

Spector

P. E.

Fox

(2010). Counterproductive work behavior and organisational citizenship behavior: Are they opposite forms of active behavior? Applied Psychology, 59(1), 21–39. https://doi.org/10.1111/j.1464-0597.2009.00414.x

43.

Spurk

Hirschi

Wang

Valero

Kauffeld

, (2020). Latent profile analysis: A review and “how to” guide of its application within vocational behavior research. Journal of Vocational Behavior, 120, 103445–103521. https://doi.org/10.1016/j.jvb.2020.103445

44.

Vroom

V. H.

(1964). Work and motivation. Wiley.

45.

Wheeler

A. R.

Halbesleben

J. R. B.

Shanine

Donovan

D. J.

(2012). Snowball sampling in management research: A review, analysis, and guidelines for future research. InAnnual Meeting of the Southern Management Association.

46.

Woo

S. E.

(2019). Big data opportunities for advancing turnover theory: A case for inductive and abductive research. Industrial and Organizational Psychology, 12(3), 330–333. https://doi.org/10.1017/iop.2019.55

47.

Woo

S. E.

Allen

D. G.

(2014). Toward an inductive theory of stayers and seekers in the organization. Journal of Business and Psychology, 29(4), 683–703. https://doi.org/10.1007/s10869-013-9303-z

48.

Yam

K. C.

Klotz

A. C.

Reynolds

S. J.

(2017). From good soldiers to psychologically entitled: Examining when and why citizenship behavior leads to deviance. Academy of Management Journal, 60(1), 373–396. https://doi.org/10.5465/amj.2014.0234