Dual-earner couples’ willingness to relocate abroad: the reciprocal influence of both partners’ career role salience and partner role salience

ABSTRACT

Successful international assignments are important for international organizations. Research has shown that employee willingness to relocate internationally strongly depends on spouse’s willingness to follow. However, the mechanisms driving these effects are not thoroughly investigated. This study gives more insight into the processes that explain both partners’ willingness to (co-)relocate internationally. We examine the influence of both partners’ career role and partner role salience on each other’s (co-)relocation willingness. On the basis of Identity theory, Interdependency theory, and Attachment theory, we hypothesize combined interaction effects of career and partner role salience. Data were collected from 226 couples (professional employees and their spouses) working in a multinational Anglo-Dutch company. Results show that, in particular, spouses’ willingness to follow their partners abroad is determined by both career importance and partner role salience. We conclude that, for theory and for organizational practice, it is indeed crucial to involve both the employees and their spouses in the decision-making for an international relocation.

KEYWORDS:

• International relocation
• dual-earner couples
• partner and career salience

International organizations experience increasing difficulties in finding able and willing candidates for their international positions (Brookfield Global Relocation Services [GRS], 2012; cf., Mol, Born, Willemsen, van der Molen, & Derous, 2009). Because successful international assignments are crucial for organizations engaged in global business ventures (Takeuchi, Yun, & Tesluk, 2002), researchers have attempted to determine the factors influencing international assignment acceptance rates. Factors, such as employee’s language ability (Kim & Froese, 2012; Mol et al., 2009) cultural flexibility (Mol et al., 2009), personality traits like openness, conscientiousness and emotional stability (Konopaske, Robie, & Ivancevich, 2005; Mol et al., 2009; Wan, Hui, & Tiang, 2003), foreign experience (Mol et al., 2009; van der Velde, Bossink, & Jansen, 2005), and family factors (Konopaske et al., 2005, 2009; Wan et al., 2003) are related to individuals’ willingness to relocate.

However, these studies also show that focusing solely on the employee him- or herself is not sufficient to predict willingness to accept an international transfer (Bhaskar-Shrinivas, Harrison, Shaffer, & Luk, 2005; Black, Gregersen, Mendenhall, & Stroh, 1999). One cause is that employees are increasingly part of a dual-earner couple (i.e., a couple in which both partners are highly educated and have (almost) full-time jobs, Brookfield GRS, 2012; see, e.g., Harvey & Buckley, 1998; Kupka & Cathro, 2007). The spouse’s¹ attitude towards the employee’s international assignment is often crucial for the success and continuation of the international assignment (e.g., Bhaskar-Shrinivas et al., 2005; Black & Stephens, 1989; Harvey, 1998; Konopaske et al., 2009; Shaffer & Harrison, 1998; Stroh, 1999; Tharenou, 2008). Likewise, the spouse has an important role in the decision to accept an international assignment (Aryee, Chay, & Chew, 1996; Brett, Stroh, & Reilly, 1993; Challiol & Mignonac, 2005; Eby & Russell, 2000; Konopaske et al., 2005; Tharenou, 2008). The more positive a (potential) trailing spouse’s attitude towards an international relocation, the more willing the employee is to accept. In other words, the reasons for employees to either accept or reject an international assignment are related not only to their own attitudes and preferences but to their spouses’ attitudes and preferences as well (e.g., Adler, 1986; Brookfield GRS, 2012).

Employees will take into account issues concerning their job and the organization they work for. That is, they will consider the offer from the perspective of their role as an employee. Simultaneously, they will consider the offer from the perspective of being a partner, for example, by envisioning the advantages and disadvantages an assignment may have on their spouses’ career, social life, and general well-being. Similarly, in their decision to follow the employee abroad, spouses will consider the opportunity from the perspective of being a (supportive or nonsupportive) partner, taking into account the consequences of a relocation for the employee.

To advance the knowledge about international relocation decisions, the present study takes into account (1) the factors influencing employees’ willingness to relocate, (2) the factors influencing the spouses’ co-relocation willingness (i.e., willingness to follow), and (3) how partners influence each other’s (co-)relocation willingness. Thus, to fully explore relocation decisions as a reciprocal process, we adapt a dyadic perspective exploring how employees and spouses mutually influence each other. On the basis of three theories, we present our hypothesized research model (Figure 1).

Figure 1. Research model.

The left part of the model represents the spouse variables, while the right part of the model represents the employee variables. Hypotheses 1 and 2 represent the within-person interaction effects, Hypotheses 3 and 4 represent the crossover interaction effects from employee and spouse predictors to spouse willingness and vice versa, and Hypotheses 5 and 6 represent the spillover interaction effects from employee predictors to spouse willingness and vice versa.

First, individuals fulfil different roles in life, which will influence their attitudes and decisions. Identity theory (Burke & Reitzes, 1991; Rothbard & Edwards, 2003) states that the more salient a role, the higher is the probability that an individual will act in accordance with this role. Role salience is the level of importance or value attached to performing in a given role area (Amatea, Cross, Clarke, & Bobby, 1986). We argue that the decisions of dual-earner couples to accept an international relocation depend on the extent to which they (both) value two major life roles: the career and the partner role.

Second, on the basis of the Interdependency theory (Kelley & Thibaut, 1978), we explore the effects of the interaction of the career and partner roles. Interdependency is an essential factor in relationships (e.g., Kelley & Thibaut, 1978). In general, partners in intimate relationships influence each other’s behaviours. An international assignment may even increase the level of interdependency, because expatriate couples leave friends and family behind, and partners on assignments have to rely more on each other for emotional support. Furthermore, expatriate spouses often have to give up their current jobs, and try to find a new job in the host country, increasing their feelings of dependence (see Kupka & Cathro, 2007; Shaffer & Harrison, 2001). Overall, this (increased) interdependency will influence the decision-making process of employees and spouses, such that they will not only take into account their own preferences but will let their decision be influenced by the attitudes and preferences of their partners as well.

Third, on the basis of Attachment theory, we argue that, when one’s partner values the partner role, this will provide a secure base of attachment, and as such interacts with other roles (i.e., career role) to predict (co-)relocation willingness. In sum, we expect that the decision to accept an international assignment will be influenced (1) by employees’ own attitudes regarding the roles as employee and as partner—that is, their career role salience and partner role salience—and (2) by career role salience and partner role salience of their spouses.

This study has various contributions to theory. First, on an intra-individual level, we explore the combined impact of the value an individual attaches to one’s career role (“being a careerist”) and one’s partner role (“being a partner”) on his/her willingness to accept an international relocation or to follow. Simultaneously, we take into account an inter-individual perspective, as we investigate how the characteristics of one partner influence the other partner’s willingness reciprocally. As such, we gain insights into how partners influence each other’s willingness to (co-)relocate. Such a dyadic perspective is relatively rare in expatriation research (e.g., Bhaskar-Shrinivas et al., 2005; van Erp, Giebels, van der Zee, & van Duijn, 2011).

Furthermore, we contribute to the expatriation literature by providing further insights into the decision-making of potential expatriate couples, on the basis of the integration of Identity theory, Interdependence theory, and Attachment theory. We explain the decision-making process beyond the mere “organizational perspective”, and link the theoretical base of these theories to the relational dynamics that influence couples’ willingness to move abroad.

Career and partner role salience

In daily life, individuals hold multiple identities (e.g., employee, partner, daughter, or sibling). The more salient an identity or role, Identity theory states, the more committed the individual is to a specific role, and the higher is the probability that an individual will make those behavioural choices that are “in agreement with the expectations attached to that identity” (Stryker & Burke, 2000, p. 286). As such, the identification with or importance attached to a role, affects the process and outcome of the decision-making (Markham & Pleck, 1986; O’Neil, Fishman, & Kinsella-Shaw, 1987).

The prospect of moving abroad for one’s (partner’s) job affects an individual, not only in the work domain but in the private domain as well. In this study, we therefore focus on two relevant roles: being in a career and being in a relationship (as partner). Career role salience refers to the value one attaches to one’s career and the extent to which one identifies with it. That is, individuals high in career role salience attach more value to excel as, for example, an engineer, a manager, or a scientist. Partner role salience refers to the value one attaches to one’s role as a partner and the extent to which one identifies with it. That is, being high in partner role salience implies that individuals will aspire to be a good partner, for instance, by satisfying the other’s needs. A relocation request will evoke the individuals’ partner identity and careerist identity (Powell & Greenhaus, 2012); both the employee and the spouse will consider to what extent their partners’ needs are fulfilled or hampered by an international relocation and both the employee and the spouse will consider the consequences of an international relocation on their (current) careers. These considerations will influence individuals’ willingness to (co-)relocate: depending on the importance both partners attach to the career and/or the partner role areas, they will be more or less willing to relocate or to follow.

Employees’ willingness to relocate

In accordance with Identity theory, we argue that employees who identify strongly with their careers, will see the opportunity of an international relocation as part of their job and career. Furthermore, because they value their identity as a “careerist”, they will be more willing to make an effort or even sacrifice, in order to perform well in their job. Therefore, they will be more willing to accept an international assignment than employees scoring low on career role salience. In the same vein, Identity theory predicts that individuals high in partner role salience will go to great lengths to live up to the expectations of their identity as a partner. In their decision to (co-)relocate, people high in partner role salience will take their spouses’ interests more strongly into account than individuals low in partner role salience. An international relocation is usually less attractive for the spouse than for the employee. The spouse has to give up a job and/or interrupt a career, will not be financially independent, and has to face the difficulties of finding a new job abroad (e.g., experiencing language barriers, work permit difficulties) or becomes a full-time housekeeper. Following Identity theory, employees with higher partner role salience are expected to give more weight to their spouses’ (personal and career) interests when deciding on their willingness to relocate. They strongly identify with their role as a partner and act accordingly. That is, employees who are high in partner role salience will more strongly consider the impact of an international assignment for his/her spouses’ life.

We, therefore, expect that the combined importance one attaches to being a careerist and being a partner will influence the decision to accept an international assignment. In other words, employees’ partner role salience and career roles salience will interact. We expect that high career role salience is associated with more willingness to accept an international assignment for employees and that this effect will be less strong when employee’s partner role salience is high.

Hypothesis 1:

Employee’s partner role salience moderates the positive relation between employee career salience and willingness to accept an international assignment, with weaker positive relations when partner role salience is high.

Spouses’ willingness to follow

Spouses who identify strongly with their careers, will regard a relocation as less favourable because it poses a threat to their own careers (Challiol & Mignonac, 2005). Following the employee (i.e., their partner) abroad generally involves leaving one’s job and interrupting one’s career. At the same time, spouses with high partner role salience will attach great value to satisfying the needs of their partners. They will give more weight to their partners’ (personal and career) interests than spouses low in partner role salience. We therefore expect that spouses’ willingness to follow their partners abroad will be jointly influenced by their career role salience and partner role salience. We expect that spouses’ career role salience will be negatively related to spouses’ willingness to follow, and this association will be attenuated by spouses’ partner role salience.

Hypothesis 2:

Spouse’s partner role salience moderates the negative relation between spouse career salience and willingness to follow, with weaker relations when partner role salience is high.

Interdependence theory: partners mutually influence each other’s willingness

As stated, important decisions, such as accepting an international assignment, are the result of a reciprocal process between the employee and his/her spouse. The Interdependence theory thus provides a theoretical basis to explain such decisions. Both couple members are highly dependent on each other for the attainment of joint as well as individual outcomes (Rusbult & Van Lange, 2003; 2008; see also Emerson, 1976; Harvey, 1998). To better understand relocation willingness, the predictions following from Identity theory should be integrated with those from Interdependence theory. Identity theory explains which outcomes are considered important (i.e., outcomes related to the partner role or to the career role), whereas Interdependence theory explains why individuals are influenced by other’s outcomes (i.e., for the attainment of certain outcomes). We will explain how these two theories can be integrated and, as such, further elucidate the mechanism behind the decision to accept an international relocation.

First, Identity theory explains how individuals’ role saliences influence their relocation willingness. If Mary and John are asked to relocate for Mary’s organization, some specific career goals of Mary are satisfied when she accepts the assignment, whereas John’s career goals are likely hampered. However, when John is high in partner role salience, his “partner-role needs” may be met by accepting Mary’s international job opportunity. That is, by accepting Mary’s relocation offer, John communicates his concern for Mary’s needs and confirms his partner identity.

Second, in line with the Interdependence theory, we argue that individuals will be influenced not only by their own role saliences but also by those of their partners. Indeed, previous research has indicated that an international assignment represents not just an individual work transition, but a stressful life event and a “family issue” as well (Baldridge, Eddleston, & Veiga, 2006; Bhaskar-Shrinivas et al., 2005; Caligiuri, Hyland, Joshi, & Bross, 1998; van Erp et al., 2011). In a relationship, individuals will be aware of the importance their partners attach to their career and to their relationship. In their decision whether or not to relocate, they will, therefore, take into account not only their own interests but also the interests of the partner (Interdependence theory), especially when they are high in partner role salience (Identity theory). More specifically, when John is high in partner role salience, he will be strongly influenced by Mary’s needs or attitudes. The more importance Mary attaches to her career, the more willing John will be to help attain her career goals, and thus to follow Mary abroad. Similarly, when Mary is high in partner role salience, she will be strongly influenced by John’s needs. The more importance John attaches to his career, the more willing Mary will be to help attain the career goals of John, and the less willing she will be to accept an international assignment.

In sum, Identity theory explains the influence of individual’s own attitudes on willingness to relocate. In combination with the Interdependency theory, it explains the influence of one’s partner’s attitudes on willingness to relocate. Based on both theories, we expect that employee’s willingness to accept an international assignment will be in particular negatively associated with spouse’s career role salience, when employee’s partner role salience is high. Furthermore, we expect that spouse’s willingness to follow will in particular be positively related to employee’s career role salience when spouse’s partner role salience is high. We thus expect the combined importance of career and partner roles of both partners to influence employees’ willingness.

Hypothesis 3:

Employee partner role salience moderates the negative relation between spouse career salience and employee’s willingness to accept an international assignment, with stronger relations when employee’s partner role salience is high.

Hypothesis 4:

Spouse partner role salience moderates the positive relation between employee career salience and spouses’ willingness to follow, with stronger relations when spouse’s partner role salience is high.

A secure base for life-changing decisions: attachment theory

Attachment theory (Feeney, 2007) suggests that individuals function best, when they have a secure base from which they can grow as an individual. A secure base gives individuals the confidence and the courage they need to make independent excursions. When couples decide to move abroad they leave behind important bases of security (e.g., social networks of family and friends). Because many other bases of security are left behind, there will be a need for one’s partner as an important and secure base of attachment. That is, in order to cope with the challenging prospect of an international relocation, partners may reciprocally support each other. An employee whose spouse is high on partner role salience, (i.e., the latter shows high concern for the employee’s interests and needs) is provided with a secure base for attachment by the spouse. In other words, in line with Attachment theory, we expect that the partner role salience of the spouse positively influences the employee’s willingness to (co-)relocate.

Again, we argue that, in addition to the partners’ partner role salience, the partners’ career role salience is important as well. Following Interdependence theory, employees may be less willing to accept an international assignment when their spouses attach great value to their own careers. However, and following Attachment theory, when the spouse is also high in partner role salience, the employee will feel more freedom to “go out” and accept an international assignment (irrespective of their partners’ career role salience). After all, spouse’s high partner role salience provides the employee with a secure base of attachment. Therefore, we hypothesize:

Hypothesis 5:

Spouse’s partner role salience moderates the negative relationship between spouse’s career role salience and employee’s willingness to accept an international assignment, such that this relation is less negative when spouse’s partner role salience is high.

Similarly, from an interdependence perspective, spouses are likely influenced by the importance the employee attaches to his/her career, in that a higher career salience of the employee increases spouses’ willingness to move abroad. Following Feeney (2007), this positive effect of career role salience will be stronger when the employee is high in partner role salience. Therefore, we hypothesize:

Hypothesis 6:

Employee’s partner role salience moderates the positive relationship between employee’s career role salience and spouse’s willingness to follow, such that this relation is more positive when employee’s partner role salience is high.

Method

Sample and procedure

Our sample consists of professional employees and their spouses from a multinational Anglo-Dutch company in the Netherlands. The employees work in two specific departments, from which employees are frequently sent to work abroad. Only workers with professional jobs were included in the study, since clerical and non-professional workers are not considered for international assignments in this company. Thus, an international assignment is “part of the job” for the employees in our sample, although the exact timing of such a relocation can be negotiated. The average time of a relocation is 4 years; and the destinations are in all continents, all over the world.

The employees and their spouses received a separate questionnaire. Except for some background questions, the questionnaires for employees and spouses were identical. They were asked to fill out the questionnaire honestly and in private, and send them back separately. The low correlations between the scores of the employee (E) and the spouse (S) indicate that indeed, there was no confounding between partners, and that employees and spouses have independently filled out the questionnaires.

To involve as many (under-represented) female employees as possible, the employee and spouse questionnaires were sent to all the professional women in the two departments (N = 196). In addition, we drew an age-based stratified sample (N = 487) from the total population of men working in these two departments. To obtain respondents distributed representatively across age, we used three age strata: younger than 30, between 30 and 45, 45 years and older.

In total, 395 heterosexual couples (135 female employees (plus spouses) and 260 male employees (plus spouses) returned completed questionnaires. For the purpose of the present study, only employees who were part of a dual-earner couple were selected, which reduced the number of eligible responses to 226 couples. A dual-earner employee was defined as an employee belonging to a couple in which both spouses were employed for at least 32 h per week (Harvey, 1998).

Of the 226 couples, 139 (62%) were Dutch couples, the remaining 38% were English/American couples on assignment in the Netherlands. All employees had completed at least a higher vocational education (i.e., they had all obtained at least a Bachelor’s degree). Equal percentages (79%) of the male and female employees worked full time (i.e., 40 h per week). Both male and female employees mentioned a wide range of job titles including petroleum engineer, civil engineer, research or production technologist, geologist, advisor, and planning coordinator. The responses did not suggest any gender differences in types and levels of job.

A larger variation was found regarding the educational level of the spouses of the employees. An academic educational level (Masters’ degree or higher) was claimed by 65% of the male and 59% of the female spouses. The type of jobs in which spouses were employed also showed considerable variation, such as HR manager, HR consultant, manager back office, manager, economist, assistant professor, sales officer, and engineer. About 49% had children living at home.

Measurements

All variables were measured using established items and scales.

Career role salience and partner role salience were assessed by means of the five-item career role salience and the four-item partner role salience subscales of the life-role salience scales (LRSS, Amatea et al., 1986). The LRSS have been validated and positively reviewed (Campbell & Campbell, 1995; Eby, Douglas Johnson, & Russell, 1998). Evidence for both reliability and discriminant validities of the LRSS has been reported (Amatea et al., 1986). The items were rephrased where necessary in order to be applicable to unmarried, cohabitating spouses. A sample item of partner role salience is: “Having a successful relationship with my partner is the most important thing in my life”. A sample item of career role salience is “I enjoy thinking about and making plans for my future career”. Responses were given on a 5-point scale (1 = fully disagree, 5 = fully agree). The alpha coefficients of career salience were .73 for employees and .82 for spouses, and reliabilities of partner role salience was .81 for employees, and .84 for spouses. Although the mean scores for partner role salience were somewhat high, possibly pointing at social desirability, the standard deviations were large enough to find meaningful variance in the measures.

Willingness to accept an international assignment

This variable was measured by asking employees (not spouses) to indicate how willing they were to accept an international assignment (based on Noe, Steffy, & Barber, 1988). Respondents answered this question on a Likert scale varying from 1 = “very unwilling” to 5 = “very willing”. In this organization, employees have frequent opportunities to relocate overseas, and many of the employees actually have worked abroad, including some of the participants at the time of the study. The question whether they are willing to relocate would concern a “new” relocation to another country. Therefore, respondents are likely to have an accurate perception of what a relocation entails. Moreover, previous research has shown that relocation willingness is strongly related to actual decisions to relocate (Turban, Campion, & Eyring, 1992). While the organization offers many opportunities to work abroad, and expects the employees to work abroad at some time, the exact timing of the relocation can be negotiated. Hence, the decision for employees whether or not to relocate is influenced by other factors, such as how their spouses would think of a relocation, and how important each partner considers their career and partner role.

Willingness to follow the employee on an international assignment

This variable was measured by asking spouses to indicate how willing they were to follow their spouse on an international assignment. Responses on these questions were given on a Likert scale, varying from 1 = “very unwilling” to 5 = “very willing”.

Control variables

We controlled for a range of variables, which previous research had indicated as related to relocation willingness and willingness to follow (van der Velde et al., 2005). Income was measured by asking respondents to indicate their income by choosing one of nine income categories, ranging from earning 12,500 USD or less a year to earning 112,500 USD or more a year (in intervals of 12,500 USD). Job satisfaction was measured using seven items from Hackman and Oldham’s (1975) Job Diagnostic Survey. Examples of items are “Generally, I am very satisfied with my job” and “I consider my job rather unpleasant” (reverse coded). The respondents used a 5-point answering scale ranging from 1 (“fully disagree”) to 5 (“fully agree”). The alpha coefficients were .86 for employees and .90 for spouses. Moreover, we controlled for job tenure (in years), children living at home (0 = no; 1 = yes), and age of the spouse (in years; r = .87 with employee age). Information on children living at home and spouse age provides an indication of the phase a family is in (i.e., whether it is a “young” family vs. an “older” family). Finally, we controlled whether employees were currently relocated on an international assignment (0 = no; 1 = yes). As respondents’ educational level was generally high (i.e., university degree), we did not control for this variable.

Analyses

We first conducted the Harman’s single-factor test and confirmatory factor analysis on the multi-item scales: partner role salience, career role salience, and job satisfaction (all scales rated both by the employee and the spouse) to test the factor structure underlying the data (CFA with Lisrel 8.80; Jöreskog & Sörbom, 2008). The hypothesized model was tested with the proposed six multi-item factors under study, and was compared with alternative models. The proposed six-factor model obtained acceptable fit (χ² = 1623.70, df = 439, p < .05, RMSEA = .09, NNFI = .94, CFI = .98; Hu & Bentler, 1999), and all items loaded significantly on the latent variables. The proposed model fitted significantly better than a four-factor model in which for both employee and spouse career role salience and partner role salience loaded on one factor each (Δχ² = 734.55, Δ df = 9, p < .001), a three-factor model, where the employee and spouse items loaded on the three variables job satisfaction, career role salience, and partner role salience (Δχ² = 1533.44, Δdf = 12, p < .001), as well as a two-factor model in which all employee-items loaded on one factor and spouse-items loaded on one factor (Δχ² = 3289.28, Δdf = 14, p < .001), and finally, a one-factor model with all items loading on one factor (Δχ² = 3479.43, Δdf = 15, p < .001). We also tested a model including paths from all of the items to an unmeasured latent factor to control for common method variance (Podsakoff, MacKenzie, & Podsakoff, 2012). This model obtained a significant better fit than the proposed model (Δχ² = 147.17, Δdf = 38, p < .001). However, the fit statistics appeared to be only slightly better than the proposed model, and accounted for 21% of method variance, which is below the standard threshold of 25% (Andrews, Kacmar, Blakely, & Bucklew, 2008). Hence, even though common method variance may exist, this has little effect on the factor structure and is not a pervasive problem. Moreover, this common method factor may also represent variance among the constructs that are due to the relationships among the variables (Podsakoff et al., 2012). Therefore, we concluded that the measures were adequate to use in the current study: The variables represent different constructs, thus, it was deemed appropriate to proceed with our analyses.

We also ran a CFA using equality constraints for the paths from the latent variables to their indicators (for career role and partner role salience), to assess whether the items have equal loadings on their factors for employees and their spouses. A model in which these were freely estimated proved to have a better fit than a model using equality constraints for these measures (Δχ2 = 52.27, Δdf = 9, p < .001). However, it should be noted that this test is very conservative and aims to test whether the loadings of the factors are equal, which may statistically be challenging (Taris, Bok, & Meijer, 1998). The relative minor difference in fit statistics between the two models (χ2 = 1676.26, df = 448, p < .001 vs. χ2 = 1623.70, df = 439, p < .05) indicates that these differences are actually minor rather than major. Hence, we proceeded with our analyses (Taris et al., 1998).

We conducted moderated structural equation modelling (MSEM) with LISREL 8.80 (Jöreskog & Sörbom, 2008) to test the hypotheses, and simultaneously added employee and spouse responses in the model. We preferred covariance analyses over hierarchical regression analyses because the first allow for correction of measurement error. To test the hypotheses in line with previous research (e.g., Cortina, Chen, & Dunlap, 2001), single indicators were used for all latent variables, because the number of items, relative to the sample size, was large. Partner and career role salience were standardized prior to calculating interaction effects.

We followed the procedure of Mathieu, Tannenbaum, and Salas (1992; see also Cortina et al., 2001) to conduct the moderator analyses. We built a crossover model including both endogenous variables employee willingness to accept and spouse willingness to follow (cf. Figure 1). Each exogenous variable needed for the interactions had one indicator, which was the standardized scale score (Cortina et al., 2001). For the interaction terms, we multiplied the standardized score of partner role salience and the standardized score of career role salience. The paths from these latent exogenous factors to their indicators were fixed with the square roots of the scale reliabilities, whereas the error variances of each indicator were set equal to the product of their variances and one minus their reliabilities. Moreover, the correlations between partner role salience and the interaction term and between career role salience and the interaction term were set to zero. Partner role salience and career role salience were allowed to correlate (for more details, see Cortina et al., 2001). In sum, we had two sets of interactions: within-person interactions with interaction of career and partner role salience within employee and spouse, and crossover interactions with interactions of career salience of the employee with partner role salience of the spouse and vice versa. For significant interactions, we plotted the interaction patterns using simple slope analysis with slopes one standard deviation below and above the mean of the moderator (Aiken & West, 1991).

Results

Table 1 shows the mean scores, standard deviations, and correlation coefficients for all study variables. Employee job satisfaction and job tenure were negatively related to employee willingness, while being currently on assignment was positively related to employee willingness. Employee willingness was positively related to spouse willingness to follow. Spouse income and being on assignment were positively related to spouse willingness to follow.

Table 1. Means, standard deviations, and correlations between employee (E) and spouse (S) variables (N = 226).

		Mean	SD	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
1	E Income	5.28	2.31	–
2	E Job Satisfaction	3.78	.71	.03	–
3	E Job Tenure	.99	.30	.08	−.14*	–
4	S Income	2.37	2.21	−.56**	−.09	.14*	–
5	S Job Satisfaction	3.74	.79	−.01	−.13	.36**	−.01	–
6	S Job Tenure	1.42	1.40	−.10	.13	−.79**	−.01	−.53**	–
7	Spouse Age	36.26	7.52	.40**	−.04	.10	−.26**	−.01	−.09	–
8	Children	.49	.50	.28**	−.16*	.01	−.22**	−.03	−.09	.68**	–
9	On Assignment	.64	.48	.04	−.16*	−.03	−.04	.02	−.10	.19**	.04	–
10	E Career Role Salience	3.36	.85	−.36**	−.06	.06	.38**	.06	−.01	−.10	−.09	.05	–
11	E Partner Role Salience	4.03	.60	.10	.01	−.07	−.04	−.10	.08	.18**	.11	.00	−.13*	–
12	E Willingness to Accept	3.47	1.44	.10	−.22**	−.23**	−.00	.07	.13	.03	.02	.13*	.12	.10	–
13	S Career Role Salience	3.12	1.04	−.40**	.19**	.04	.12	.05	−.13*	−.26**	−.21**	.00	.23**	−.12	−.15**	–
14	S Partner Role Salience	4.07	.63	−.06	.06	−.04	.05	−.07	.05	.06	.17**	.13*	−.07	.27**	−.12	.04	–
15	S Willingness to Follow	3.35	1.35	.04	−.02	−.07	.14*	−.06	−.01	.00	−.06	.16*	.21**	.14*	.40**	−.08	.01		–

*p < .05; **p < .01; E = employee; S = spouse.

Hypothesis testing

We tested a single structural equation modelling (SEM)-model including all the proposed interaction effects. To evaluate model fit, established goodness-of-fit measures were used. The root mean square error of approximation with values of below .08 is acceptable and below .05 is good (Bentler, 1990; Hu & Bentler, 1999). Furthermore, comparative fit index (CFI) and incremental fit index (IFI) should be above .90. The postulated SEM-model with the interactions obtained an acceptable fit (χ² = 143.37, df = 55, p < .001, RMSEA = .08, CFI = .90, IFI = .90). Standardized coefficients of the models are shown in Table 2.

Table 2. Standardized coefficients of the relationships in the model.

	Dependent variables
	Employee willingness to accept	Spouse willingness to follow
	Standardized coefficient	Standardized coefficient
Employee
Income	.08	−.01
Job satisfaction	−.13	.13
Job tenure	.04	−.17*
Children	.36*	−.35*
On assignment	.06	.16*
Career role salience	.04	.28***
Partner role salience	.09	.29**
Spouse
Spouse age	−.38*	.37*
Income	−.01	.09
Job satisfaction	.11	−.15*
Job tenure	.00	.09
Career role salience	−.14	.07
Partner role salience	−.13	−.07
Willingness to follow	.41***
Within person interactions
E career salience * E partner role salience	.03	−.39***
S career salience * S partner role salience	−.05	.36***
Crossover interactions
E partner role salience * S career salience	−.03
E career salience * S partner role salience		.18***
R^2	.27	.34

*p < .05, **p < .01, ***p < .001.

Hypothesis 1 predicted that employee partner role salience moderated the relation between employee career role salience and employee willingness. The interaction between employee partner role and career salience was not related to employee willingness (β = .03, ns). Thus, Hypothesis 1 was rejected. Hypothesis 2 predicted that spouse partner role salience moderated the relation between spouse career role salience and spouse willingness to follow. The interaction effect between spouses’ career and role salience was indeed significantly related to spouse willingness (β = .36, p < .001). Figure 2 shows the interaction effect. The relationship between spouses’ career role salience and their willingness to follow was negative for spouses with low partner role salience (β = −.29, p < .01), and it was positive for high partner role salience spouses (β = .43, p < .001). The more value spouses attach to their own career, the less willing they are to follow, unless they attach great value to their relationship (i.e., high partner role salience). This supports Hypothesis 2.

Figure 2. The interaction between spouse partner role salience and career role salience of the spouse in relation to spouse willingness to follow.

Hypothesis 3 predicted a crossover effect of spouse career role salience and employee partner role salience in relation to employee willingness to accept. The interaction effect was not significant (β = −.03, ns). Therefore, Hypothesis 3 was rejected. Hypothesis 4 predicted a crossover effect of the interaction between employee career salience and spouse partner role salience in relation to spouse willingness to follow. The interaction between employee career salience and spouse partner role salience was significant (β = .18, p < .001). Figure 3 shows the interaction pattern. When spouse partner role salience is low, the relationship is not significant (β = .10, ns), while the relationship is positive for spouses with high partner role salience (β = .46, p < .001). This supports Hypothesis 4.

Figure 3. The interaction between spouse partner role salience and employee career role salience in relation to spouse willingness to follow.

Hypothesis 5 predicted that an interaction effect of spouse’s partner role salience and spouse’s career role salience on employees willingness to accept an assignment. This relationship was not significant (β = −.05, ns), Hypothesis 5 is not confirmed. As formulated in Hypothesis 6, we found a significant interaction between employee’s career salience and employee’s partner role salience in relation to spouse willingness to follow (β = −.39, p < .001). Figure 4 shows the interaction pattern. The relationship between employee career role salience and spouses willingness to follow was positive for employees with low partner role salience (β = .67, p < .001), while this relationship was non-significant for employees with high partner role salience (β = −.11, ns). The more importance employees attach to their careers, the more willing spouses are to follow when employees’ partner role salience is low; but, when employees are high in partner role salience, their spouses’ willingness to follow is high irrespective of employees’ career role salience, which is fully in line with Hypothesis 6.

Figure 4. The interaction between employee partner role salience and career role salience of the employee in relation to spouse willingness to follow.

Discussion

In this study, we investigated factors influencing employees’ and spouses’ willingness to (co-)relocate. Based on Identity theory, Interdependence theory, and Attachment theory, we explored the role of the importance dual-earner partners attach to their career and to being a partner. Overall, the results emphasize the importance to take into account career and partner role salience of both partners simultaneously to predict (co-)relocation willingness.

Theoretical contributions of this study

Research has shown that employee willingness to relocate internationally is strongly associated with spouse willingness to follow (e.g., Aryee et al., 1996; Brett et al., 1993; Tharenou, 2008). However, studies explaining why and how (i.e., by which processes or mechanisms) employees and spouses influence each other’s (co-)relocation willingness are scant. This study gives further insights into the processes that explain which attitudes influence (co-)relocation willingness, by (1) adapting a dyadic approach, (2) taking into account important life roles of both partners, and (3) exploring the combined impact of these life roles by integrating three theories, Identity theory, Interdependence theory, and Attachment theory. In particular, we have shown that employee willingness to accept a relocation is primarily dependent upon the willingness of their partners to follow them. Furthermore, their partners’ willingness to follow is determined by complex processes that involve a high partner role salience among spouses in combination with high employee career role salience or low spouse career salience. Hence, the study shows that the strongest predictors of employee willingness reside within the spouses and the relevance of their careers and partner roles.

This study contributes to the research literature in a number of ways. First, in general, our study contributed to both Identity theory and Interdependence theory in that it showed how the combined impact of the value spouses attaches to the career role (“being a careerist”) and the partner role (“being a partner”) influence their willingness to co-relocate. More specifically, we contributed to Identity theory by showing that career decision-making behaviour is influenced by the importance of specific life roles. Moreover, we added to Interdependence theory by showing how—in a dyadic perspective—dual-earner couples mutually influence each other, for example, with respect to career orientation. Finally, we contributed to Attachment theory: this study showed that spouses who are provided a secure base by means of high partner role salience of their partner (the employee) strengthens the positive effect of employees’ career role salience on spouse’s willingness to follow. We now turn to a more specific discussion of the theoretical implications of our findings.

Spouses’ willingness to follow

Spouses’ willingness to follow their partners abroad was strongly associated with their own roles of careerist and partner. In line with Identity theory, the more value spouses attached to their career, the less willing they were to follow their partner, but only when they had a low partner role salience (H2). This suggests that those who strongly value their careers, will not put their careers at risk or on hold for the satisfaction of their partner. Hence, they are less likely to follow the employee on an overseas assignment because this may be disadvantageous to their career.

In contrast, when spouses’ partner role salience was high, they valued their partners’ interests over their career. More specifically, when spouses were high in partner role salience, their career role salience was positively related to their willingness to follow (H2). Plausibly, spouses who are high in both partner role salience and career role salience, know how highly their partner may value the career opportunity (i.e., because spouses themselves attach great value to a career) and they are willing to accommodate such an opportunity (i.e., because their partner role salience is high).

Furthermore, in line with our expectations, we found that employees’ career role salience was particularly positively associated with spouses’ willingness to follow when spouse partner role salience was high (H4). This indicates that spouses are more willing to follow when the career is important for the employee, and they themselves identify strongly with being a good partner.

Also, in line with Attachment theory, spouses were influenced by the combined career and partner role saliencies of the employee (H6). Spouses were more willing to follow, the more important employees valued their own career, but this effect disappeared when employees found their partner role important. In effect, when employees’ partner role salience was high, spouse willingness to follow was also high, regardless of employees’ career role salience. This may suggest that spouses of employees high in partner role salience are confident that their interests will be taken into account on an assignment, and, therefore, are generally more willing to follow. Spouses of employees low in partner role salience require additional arguments to follow the employee abroad. Employee’s strong career orientation may be a deciding factor which explains this relationship.

Employees’ willingness to relocate

For employee willingness to accept, spouse’s willingness to follow was the strongest predictor, showing that especially the spouse plays a crucial role in the decision process. However, in contrast with our expectations (H1, H3, H5), employees’ willingness to relocate was neither to their spouses’ partner or career role saliencies nor to their own career or partner role saliencies (contradicting Identity theory), nor to combinations of these.

The clearly different findings for employees and spouses may be explained by the fact that “following” represents a larger personal risk than “accepting” and, therefore, required a more secure attachment base by a high employee partner role salience. Note that employees already have a secure attachment base in the company: they do not leave the company when expatriated (see, e.g., Shaffer & Harrison, 2001).

In sum, the current study has important implications for theory. In general, willingness to relocate and actual relocation decisions have been theoretically explained through different mechanisms, such as identity, interdependence, and attachments, which both partners in a relationship have. The current study shows that these theories do not operate in isolation, but they interact, and thus relocation decisions are not made solely on the basis of one’s salient identity (e.g., as a careerist) but also on one’s interdependent relationship with one’s partner, and the attachments people have in a relationship to fulfil their role as a partner. They should therefore be mutually taken into account when predicting the willingness of employees and their partners in relocation decisions. Beyond partner and career role salience, employee willingness to relocate was influenced by human capital factors, such as income, experience and job tenure, and a family factor: the presence of children. These factors may be seen as indicators of other important life roles: being a bread winner and a parent. To gain further insight in the motivators and barriers to relocate internationally, we suggest for future research to include these important life roles and their mutual influence as well.

Practical implications of this study

This study has a number of important practical implications for (multinational) organizations. Our study clearly shows that the employee and spouse both take into account each other’s career outcomes. Thus, organizations should involve both partners in the decision-making process for an international relocation. We advise international human resource managers to take a “couple” approach to relocation, since reasons for employees to either accept or reject an international assignment are related both to employees’ attitudes and preferences as well as to spouses’ attitudes and preferences. HR policies should focus on both the identity of a careerist, with interventions such as coaching and talent assessments, and the identity as a partner, including being a father/mother, with interventions such as commuting faculties and local schooling for children, not only for the employee but also for the spouses. Career facilities for spouses may, for example, include help in finding a job abroad (e.g., the organization may participate in a network of expatriating companies and semi-government institutions), assisting with acquiring work permits, and facilitating a study break. Since spouses with longer job tenure are more confident to find a suitable job elsewhere and are less worried about a “gap” in their résumé when following their partner, the relocation decision will have to be “tuned in” with the career rhythm of the spouse as well. So management development or talent development will have to involve, partly, (facilitating) “couple development”. In sum, international relocations should be viewed as family events, more than (merely) as individual employee experience (e.g., Bhaskar-Shrinivas et al., 2005).

Limitations and future research suggestions

There are some limitations of this study that need to be discussed. A first important issue is that we did not measure actual relocation. Since, in general, people are not able to accurately predict the emotional impact of an event, our results on (co-)relocation willingness may be either too optimistic (when individuals overestimate the positive emotional impact of relocation) or too conservative (when individuals overestimate the negative emotional impact of relocation; see, e.g., Wilson & Gilbert, 2005). However, previous studies have shown that willingness to relocate in general is a good predictor of actual behaviour, when a specific assignment is offered (Brett & Stroh, 1995). For example, Turban et al. (1992) found a correlation of .46 between willingness to relocate and an actual relocation decision made during the following year. Moreover, since the study was done in an organization with a strong cultural norm of expatriation, we feel that measuring willingness instead of actually going would show more variance in our dependent variable. Moreover, an advantage of our present sample in comparison with other studies (e.g., Aryee et al., 1996; Borstorff, Harris, Field, & Giles, 1997; Brett & Stroh, 1995; Konopaske et al., 2005) is that many of our couples had previous international experience, and the employees in our study have considered an international relocation as a logical career step, because they all work in departments from which it is common to be sent to work abroad. More than 60% of the respondents involved in the survey had already been on an international assignment, or were currently on assignment in the country where the study took place, which makes it likely that the answers they gave in response to our hypothetical “relocation offer” is a good indicator of actual future relocation decisions (Larson, 2006). In reality, it may be that couples relocate multiple times across their careers, and therefore, it remains important to assess willingness to relocate in the future, even though employees and partners have past experiences with relocations. Moreover, for organizations and managers, it may also be very important to understand the willingness and predictors of willingness of those working not only in their home country but also in international contexts, as especially these employees may be approached to conduct another international relocation in the (near) future.

Another issue pertains to the measurement of willingness to accept and to follow, as they were both measured using single-item measures. However, there is evidence that single-item measures can be as valid as multiple-item measures (see, e.g., Nagy, 2002). Moreover, social desirability may be an issue with employees’ measures of partner role salience, as people may be more likely to portray themselves as being “good” partners. However, we have found sufficient variance in this measure in order to predict willingness to accept and follow, and thus the responses from the employees and their spouses vary enough to find meaningful patterns.

Another concern is whether our results can be generalized to other organizational types and other cultures. As we performed this study in an organization with a strong cultural norm of expatriation and a context of high-level expatriate services, providing expatriates with optimal circumstances to expatriate, the importance of the situational circumstances is minimized, creating an ideal context to explore more personal influences on expatriation willingness. Moreover, we did not explicate a (hypothetical) destination. However, it is unlikely that participants would have referred to an objective cultural distance measure had we done so. Rather, people have a (sometimes very accurate, sometimes rather shallow, sometimes pretty wrong) perception of destinations. When a real and specific relocation offer is made, only then will (and can) couples start investigating their potential location more thoroughly.

Furthermore, cultural distance as a country-level measure may oversimplify matters. For instance, many developing countries have capitals with facilities resembling one’s in Europe or the US (good private hospitals, international schools, high-quality restaurants) attracting a large international community, whereas their rural areas may lack even the most basics needs, like running water (see also van Erp, van der Zee, Giebels, & van Duijn, 2014). If we had explicated a hypothetical destination in the current study, that could have created noise rather than further insight on the results (cf. Noe et al., 1988).

Nonetheless, we emphasize the need for research to further investigate the role of cultural distance in dual-earner couples’ relocation willingness. Although there is strong evidence that dissimilarities in culture indeed impede international adjustment (Bhaskar-Shrinivas et al., 2005), the effect of cultural distance on relocation willingness has received less attention. The effect of cultural distance on willingness to relocate may be more complex. Although some studies report a negative association between cultural distance and relocation willingness (e.g. Aryee et al., 1996) others did not find significant results (Wagner & Westaby, 2009). First, larger cultural differences entail more uncertainty, and are therefore less attractive for potential expatriates and spouses. However, new cultural experiences may be viewed as challenging and exciting as well. This may be especially true for the current sample of employees who already decided to work for an international organization. In other words, they may look forward to an international adventure.

Second, employees may be influenced differently by cultural distance than their spouses. Employment opportunities for spouses may decline the larger cultural distance between home and host country, reducing willingness to follow especially for spouses with high career role salience. Future research may shed more light on the impact of cultural differences in the relationships between career and partner role saliences on the decisions to relocate internationally.

Moreover, the role of predicted length of an assignment may also be relevant. When an international assignment is relatively short (e.g., less than a year), a partner may be more likely to follow when it does not disrupt his/her career in the home country, and employment opportunities at the destination are insecure. However, for longer predicted relocation assignments, it becomes more important for the spouses to settle in the destination country, build a life and networks, and therefore, may also predict the effects of career and partner role salience on willingness to follow. Hence, for future research, it would be important to further assess these effects. A final limitation is that the cross-sectional design makes it difficult to determine the causal direction of the relationships. However, since an important characteristic of expatriate research is that participants frequently move from one country to another, as such increasing the difficulty to trace them over time, it is not surprising that longitudinal research is still rare in expatriation literature (Hechanova, Beehr, & Christiansen, 2003). This brings us to our suggestions for future research in this area.

There is a need for longitudinal designs in future studies because it would enable us to examine a couple’s decision-making process over time, and the actual assignment. Moreover, it would provide a more robust insight into how previous relocation decisions and experiences affect later relocation decisions. Such a developmental perspective for dual-earner families (Sekaran & Hall, 1989) makes it possible to study what happens when a couple’s family cycle and the two partners’ career cycles get “out of sync”.

Another suggestion for future research is to distinguish career salience from the extent to which participants assume an international relocation is a career advantage. Again, in the present study, our employees worked for an international company where international relocations were an inevitable part of employees’ jobs. The employees know that it is part of their job, but the exact timing can be discussed and negotiated. Not every internationally oriented organization expects every employee (in a specific job), to move abroad at one point in their career. In such cases, researchers may want to distinguish between having a strong career identity and whether individuals see an international assignment as a career advantage, as it may be expected that only those who value their career (i.e., high in career role salience) and see an international relocation as a career advantage are highly willing to accept an international assignment.

Finally, since the present study was conducted in one multinational company (by examining one organization, we were able to study the influence of important career and partner roles, while we keep/hold the context stable), we were not able to study the effect of various expatriate arrangements on employees and spouses. Therefore, we suggest that future research examines how the support provided by an organization to the potentially expatriate family before, during, and after an assignment affects employee and spouse willingness to relocate. In fact, few studies have directly investigated the effect of HR support on willingness to relocate (Borstorff et al., 1997). Brett and Stroh (1995) argue that when spouses perceive their organization as supportive, their willingness to accept will increase; and career support concerns have been found to be negatively related to spouse willingness to relocate (Konopaske et al., 2005). Given that an increasing number of organizations are developing and adjusting their expatriate packages to meet the needs of dual-earner couples, evaluating the actual effects of this support on the willingness to consider a move and the actual acceptance rate by such couples, as well as their satisfaction with support on previous assignments, needs further study.

Conclusion

Successful international assignments are crucial for international organizations, and in dual-earner couples, the spouse’s attitude is crucial for the couples’ willingness to relocate. Especially, spouses are less willing to relocate when the career identification of their partners was higher, however this disappeared when their partners also highly valued their partner role. The study shows that relocation decisions are not solely dependent upon employees, but that their willingness is primarily related to their partners’ willingness to follow them. Spouse willingness to follow is predicted by joint roles of employee and spouse career and partner role salience, which thus relates ultimately to whether employees are willing to relocate internationally for their companies.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. We use the term spouse to refer to the partner, either married or cohabiting, of the employee/potential expatriate. We will use the term partner(s) to refer to both the employee and the spouse.