Yes, they can do it! Exploring female expatriates’ effectiveness

Abstract

Research has highlighted the under-representation of women in international assignments. While individual and organisational causes for this minor presence have been deeply explored, there is a lack of empirical studies on the effectiveness of female expatriates. Moreover, these studies have focused on a particular facet of effectiveness, and findings are usually based on the perceptions of women of their own work. Additionally, only a few of them include male expatriates, which undermine the possibility of making comparisons.

This paper explores gender contingencies on expatriate effectiveness using different groups of measures: adjustment, premature return, performance, commitment and job satisfaction. The author carried out four logit models and, overall, and she did not found significant differences on effectiveness between female expatriates and men ones. Moreover, not-significant differences favour women. Theoretical and practical implications are contemplated.

1

Introduction

There is general consensus that the development of a global orientation is imperative for today's organisations in order to be competitive (Tung, 2016). Given that top managers’ international experience in multinational companies seems to be a relevant factor for the success of the company's internationalisation strategy (Carpenter, Sanders, & Gregersen, 2001; Hutzschenreuter & Horstkotte, 2013), leaders with effective intercultural skills are essential for companies competing in a borderless world. As a result, the possession of a global mind-set seems to be a prerequisite for advancement to higher management positions in globalising firms (Altman & Baruch, 2013; Bird & Mendenhall, 2016). Previous researchers have argued that because overseas work experience has been proven to be the fastest way to acquire global capabilities (Caligiuri & Tarique, 2009; Insch, McIntyre, & Napier, 2008), it is often used as a surrogate measure of global mind-set (Tung, 2004). Thus, international mobility can be a competitive differentiator for organisations in a global context (Altman & Shortland, 2008; Caligiuri & Tarique, 2009; Insch et al., 2008).

Research has highlighted the underrepresentation of women in expatriation (Shortland, 2014; Altman & Shortland, 2008; Insch et al., 2008; Tung, 2004), and recent surveys have shown that women account for only 25% of IAs (Brookfield Inc., 2016). Taken overall, this suggests that there are still obstacles to the presence women in the international workforce.

This shortcoming poses two concerns. Firstly, because IAs have become a key requirement for career advancement (Engle, Dowling, & Festing, 2008), women's lack of international experience may be an obstacle to their career progress. In this regard, 37% of the respondents to Brookfield's (2016) survey reported that the low acceptance rates of IAs among female candidates are having adverse impacts on the creation of balanced senior management teams in their companies. Secondly, from a point of view of maximising talent, companies cannot afford to limit their talented human resources pools by excluding women (Paik & Vance, 2002).

Research to date has explored in depth the causes of the absence of women managers in IAs (see Altman & Shortland, 2008; Salamin & Hanappi, 2014; Shortland, 2014, for recent reviews). Since Adler's (1984c) influential paper regarding international female underrepresentation, most research has focused on three widespread myths: the supposed lower willingness of women to accept IAs (e.g. Adler, 1984a, 1984b, 2002); the existence of host-country prejudices against female expatriates and the rejection of women in the foreign culture (e.g. Adler, 2002; Caligiuri, Joshi, & Lazarova, 1999; Hutchings, Michailova, & Harrison, 2013); and the cultural distance between host and home countries, which includes different roles being distributed to women and leads organisations to send men rather than women abroad (Hutchings et al., 2013; Mathur-Helm, 2002; Tung, 2008). Paradoxically, the possibility that a woman may be as effective as a man in an international location has been underexplored, even though from an organisational approach women's supposed ineffectiveness should be the only valid criterion for excluding them from IA selection processes.

Our study aims to address this gap in research. More specifically, our main objective is to explore whether women are less successful on IAs than men are, and, as a consequence, are not sent abroad. In order to achieve this objective, first, relevant literature and the empirical evidence concerning the effectiveness of women in IAs is discussed. Then, the theoretical explanations that have been offered for the evidence found are reviewed. Conclusions from this review lead us to suggest that there are no differences between male and female expatriate effectiveness. We empirically test this claim on a sample that includes both men and women. Finally, we report the main theoretical and practical implications of our findings.

2

Literature review

The low representation of women in IAs has been addressed from a variety of perspectives. These different approaches can be grouped into three main categories:

• (a)

  Individual: women's lack of interest in expatriation (Adler, 1984a, 1984b, 2002; Stroh, Varma, & Valy-Durbin, 2000); different emphasis on personal vs. professional life (Fischlmayr & Kollinger, 2010; Linehan & Walsh, 2000; Selmer & Lauring, 2011; Shen & Jiang, 2015; Tharenou, 2008).

• (b)

  Organisational: biases in the selection process (Harris, 2002; Paik & Vance, 2002; Varma, Toh, & Budhwar, 2006); higher selection requirements for women than for men (Linehan & Walsh, 2001); concerns about the physical safety of women because of the inherent risks to travelling to underdeveloped countries (Napier & Taylor, 2002); lack of mentors’ counselling (Linehan & Walsh, 1999; Selmer & Leung, 2003b); lack of effective models of female leadership (Linehan, 2002).

• (c)

  Cultural: host-country prejudices or rejection to women (Hutchings et al., 2013; Stroh et al., 2000; Tung, 2004); different roles attributed to women in the host country (Hutchings et al., 2013; Mathur-Helm, 2002; Tung, 2008).

Research has been directed mainly at examining individual and organisational causes.

With regard to the former, several studies have pointed out that women are interested in IAs, and would be likely to accept them (Adler, 1984a, 1984b, 1984c; Linehan & Scullion, 2004; Stroh et al., 2000; Tharenou, 2010). In addition, recent studies on self-initiated expatriates – individuals who voluntarily resort to the international sphere to seek employment – have found that the percentage of women exceeds that of men in this type of international experience, and that the percentage of women in self-initiated expatriation is larger than in company-assigned expatriation (Andresen, Biemann, & Pattie, 2015; Cerdin & Pargneux, 2010; Vance & McNulty, 2014). Scholars interpret these data as evidence of the interest and engagement of women in international careers. In sum, despite some contrasting evidence (van der Velde, Bossink, & Jansen, 2005), there is some consensus on the willingness of women to undertake IAs.

With regard to organisational causes, female expatriates report that they are less likely to be sent abroad by their organisations (Stroh et al., 2000). Moreover, some studies have found that women are required to be more highly qualified than men, including having a higher level of education, greater technical competence and more managerial expertise (Linehan & Walsh, 2000). It has also been suggested that that if there are several equally qualified male and female candidates for an IA, a man is usually preferred (Forster, 1999). Although these results should be treated with caution because they have been inferred from women's perceptions, the conclusion from this research must be that organisations have a rather reluctant attitude to sending women on IAs. According to different authors, this attitude can be related to assumptions concerning the masculine nature of international locations, so there is sufficient evidence to confirm that there may be some implicit discrimination against women (Hutchings, Lirio, & Metcalfe, 2012; Insch et al., 2008; Paik & Vance, 2002; Tharenou, 2010).

As has been previously noted, studies that explore whether cultural distance might be a problem for the effectiveness of expatriate women in comparison with men are scarce. This may be related to the lack of an accurate definition of expatriate effectiveness (Cerdin & Pargneux, 2009). Bearing this limitation in mind, scholars have mainly focused on three indicators for measuring effectiveness put forward by Caligiuri and Tung (1999): premature termination; performance; and adjustment.

With regard to premature termination, which is considered an assignment failure rate, researchers have noted that the rate of premature return of men is similar to that of women (Caligiuri & Tung, 1999; Forster, 1999; Tharenou, 2010).

With respect to performance, Adler found that 97% of a sample of US female expatriates thought they were successful in their overseas assignment. Although this percentage seems unrealistic (Caligiuri et al., 1999), other researchers have found that female expatriates’ performance is satisfactory regardless of the country they come from and the one they are assigned to (e.g. Adler, 1984b; Caligiuri & Tung, 1999; Linehan and Scullion, 2004).

Finally, another group of studies focuses on cross-cultural adjustment (hereafter, CCA). While it is true that this concept is not a performance indicator per se, it has been assumed that if expatriates are unable to adjust to the host location, they are likely to perform poorly. Some authors (e.g. Caligiuri & Tung, 1999; Culpan & Wright, 2002) have suggested that women may face difficulties in adjusting to cultures where there is both low female participation in the workforce and a lower percentage of women managers. However, comparative studies of adjustment have concluded that, overall, women adjust at least as well as men (Cole & McNulty, 2011; Haslberger, 2010; Salamin & Davoine, 2015; Selmer & Leung, 2003a).

Additionally, both Selmer and Leung's (2003a) study and its replications found that female expatriates have significantly higher interaction and work adjustment levels than their male counterparts. This is an important finding, because it has been suggested that interaction adjustment – namely, the perceived comfort related to the interpersonal relations and the interaction with other people from the host country - may be the fundamental dimension for CCA, since both general and work adjustment are based on interpersonal relationships (Black & Stephens, 1989). In connection to this, some researchers have suggested that female expatriates may be better positioned than men to handle an expatriate assignment (Shortland & Altman, 2011), due basically to their better social, interpersonal and communication skills and to their capabilities such as team- and relationship-building (Vance & McNulty, 2014). It has also been noted that female expatriates benefit from visibility and a “halo effect”, because host-country nationals assume that those women would not have been sent abroad if they were not competent (Napier & Taylor, 2002).

Overall, it seems that the success of women on global assignments has been generally confirmed. Nevertheless, most studies have utilised a single criterion of expatriate success, which offers only a partial view of performance. Also, the vast majority of studies have involved only women participants, which prevents comparisons with male expatriates (e.g. Linehan & Scullion, 2004). Moreover, it seems that female expatriates’ success has been defined by the outcomes of the IA (adjustment, permanence or performance) but without taking into consideration personal outcomes for the individual expatriate. Regarding this point, Cerdin and Pargneux (2009:7) suggested that “the individual's definition of IA success can be different from that of the organizational”, and proposed that career success and job success should be considered when measuring expatriate success.

The purpose of this study was therefore to examine gender differences in expatriate effectiveness from a broader perspective. In doing so, I explored whether women were less likely than men to perform, not in terms of a particular indicator that has been traditionally used as a measure of performance, but in terms of wide range of measures.

I additionally included personal outcomes for the individual in order to better understand the intricacies of female performance in international contexts. Thus, they were explored gender differences in satisfaction and commitment. In the context of expatriation, the manifestation of job satisfaction is satisfaction with the assignment (Firth, Chen, Kirkman, & Kim, 2014). Dissatisfaction is linked to the desire to leave the organisation, so some authors assume that if the rate of expatriate abandonment of the assignment is higher than that of domestic employees, this is precisely because of the lack of satisfaction. Organisational commitment refers to commitment to the goals and values of the organisation, which is reflected in the willingness to achieve those objectives and in the desire to remain in the organisation. The relationship between expatriation and organisational commitment both to the multinational company and to the subsidiary has been widely studied, and these studies suggest a strong relationship between satisfaction in the workplace, and commitment (see Hechanova, Beehr, and Christiansen (2003) for a meta-analytic review).

3

Method

3.1

Sample characteristics

139 Spanish medium and high skilled workers (37 women and 102 men) took part in this study. All of them were Assigned Expatriates, since they were assigned by their companies to their international post. At the time the survey they had on average been in their current location for almost two years (21.65 months). 63.71% of the respondents were working in countries in the European Union, 16.93% in America, 11.29% in the Middle East and the remaining 8.06% in Asia and Oceania. They were slightly above 34 years old on average (34.23), and almost all of them (93.55%) had attended university for at least three years.

3.2

Instruments and measures

The data were collected using an electronic survey conducted from May to August 2016. Several companies were asked to participate. They were advised of the rationale for the study and were asked to send a questionnaire to their staff abroad. Ten organisations agreed to participate.

The on-line questionnaire contained the study measures and also several demographic questions (gender, age, education) and questions related to the IA (time-on-assignment, host country). This instrument has been considered well-suited for sending to expatriates in several countries (Dillman, 2000). In order to minimise any potential problem of method bias, all items were assigned in random order, and some of them had reverse-polarity, to reduce the likelihood of obtaining uniform answers (Podsakoff, MacKenzie, & Podsakoff, 2012). The instrument was pilot tested on 4 former expatriates (2 men and 2 women). It was included a cover letter informing the potential respondents that participation was voluntary and that the results were confidential. Both the cover letter and the questionnaire were provided in Spanish.

3.3

Measures

I adopted the scales in order to assess widely accepted constructs of CCA (Black, 1998; Black & Stephens, 1989); job performance (Black & Porter, 1991); perceived effectiveness; premature return (Cammann, Fichman, Jenkins, & Kiesh, 1979); organisational commitment (Meyer and Allen, 1991) and job satisfaction (Macdonald & Maclntyre, 1997). The reliabilities for all these constructs were above 0.70 and therefore acceptable (Nunnally, 1978).

Gender. Gender was assessed using a single item, asking whether the respondent was male (1) or female (0).

Time in Assignment. Time in location was assessed by asking participants to state (in months) the amount of time they had been in their current assignment.

Expatriate Effectiveness. Effectiveness in the IA was assessed through two measures. The first, job performance, was the evaluation provided by the company; while the other (self-perceived effectiveness) indicated the perception of the expatriate of his/her own performance.

Job performance. The original scale created by Black and Porter (1991) was adapted. Expatriates were asked about their latest performance evaluation and requested to indicate where that rating would place them in relation to their peers on a percentage basis along 5 dimensions: (1) overall performance, (2) ability to interact with others, (3) completing tasks on time, (4) quality of performance, and (5) achievement of work goals. The resulting scale consisted of 8 items. Cronbach's alpha for this scale was α = 0.875 in this sample.

Self-Perceived effectiveness. To measure the self-perceptions of expatriate effectiveness I adapted a scale developed by Chen et al. In it, expatriates were asked about their perceptions of their own performance. It consisted of 8 items, which are responded on a 5-point scale, where 1 = strongly disagree and 5 = strongly agree. The reliability in this sample was α = 0.79.

Adjustment. To measure this factor the classical adjustment scale developed by Black (1998) and Black and Stephens (1989) was used. Although this scale has recently been subject to criticism (e.g. Haslberger, Brewster, & Hippler, 2013; Hippler, Caligiuri, Johnson, & Baytalskaya, 2014), it remains a widely used instrument to examine expatriate CCA and to make comparisons among different countries and cultures (Cole & McNulty, 2011; Selmer and Leung, 2003a, 2003b). This scale includes 3 dimensions of adjustment (general, relational and work adjustment) through 14 items. Each item was rated using a 7-point Likert-type scale with anchors of 1 = strongly disagree to 7 = strongly agree. The sub-scale of cultural adjustment assesses the expatriate's adjustment to living conditions. The sub-scale of interactional adjustment assesses the expatriate's adjustment to socialising with other people. The sub-scale of work adjustment assesses the expatriate's adjustment to the requirements and conditions of the new job. In the current sample the reliabilities of the scales were 0.73, 0.83, and 0.74 for general, interaction, and work adjustment respectively. For the total scale (CCA) the reliability was 0.89.

Premature return. The turnover intention scale developed by Camman and his colleagues (Cammann et al., 1979) was adapted to measure this variable. This is a 3-item scale with a 7-point Likert format ranging (from 1 = strongly disagree to 7 = strongly agree). The reliability of this scale in the current sample was 0.71.

Organisational Commitment. To measure the affective commitment it was applied the sub-scale constructed and validated by Meyer and Allen (1991). This sub-scale entailed eight items that the participants responded to along a 7-point Likert scale, from 1 = strongly disagree to 7 = strongly agree. Reliability for this scale was α = 0.87 in this sample.

Job Satisfaction. Job satisfaction was assessed using the 10-item scale developed by Macdonald and Maclntyre (1997). This scale assesses the individual's overall satisfaction with several job aspects on a 5-point scale ranging from 1 = strongly disagree to 7 = strongly agree. This scale has shown to be internally consistent with reliability α = 0.77.

4

Results

The aim of this part of our study is to analyse whether expatriate women and men differ in their effectiveness in IAs. Specifically, I explore whether gender introduces significant differences in adjustment, premature return, job performance, self-perceived effectiveness, satisfaction and commitment. As these include are categorical and qualitative variables, research has utilised linear models, such as ANOVA. These models involve linear regressions, which describe outcomes as linear combinations of various independent variables (also called predictors) plus some random error (Chatterjee & Hadi, 2009). Nevertheless, their limitations (Jaeger, 2008) make Logit models well-suited for the analysis of binomially distributed categorical outcomes. Thus, our estimation method to test the effect of gender on expatriate effectiveness (regardless of the chosen indicator) is to apply a non-linear regression model.

Since Logit regression provides a more accurate result at the threshold values (Cabrer, Sancho, & Serrano, 2001) I have chosen this technique. Rodgers and Ghosh (2001) noted that Logit regression models are appropriate when the dependent variable is a simple decision, the underlying random elements of the distribution are assumed to follow a binomial distribution and the error terms of the regression follow a logistic distribution. As I collapsed each dependent variable as success (i.e. adjusted, high self-perceived effectiveness) or failure (i.e. unsatisfied or intention to return) and I also considered the belief that the underlying distribution of the error term is common in this kind of study, our regression meets the above conditions.

Bearing in mind that there is some variation across equations, the logistic distribution function can be expressed as:

This seminal equation is not linear, and 0 < P_i < 1 at −∞ < Z_i < ∞. However, its logarithmic transformation (Logit) permits its use as a linear function both in X_i and β_i

This model serves to predict the probability an expatriate has to have a higher or lower perception of their effectiveness and also to identify whether gender increases or decreases this probability. In this type of model, the probability interval ranges from 0 to 1 but Z_i ranges from −∞ to ∞. Moreover, although the Logit is linear with respect to X_i, the probabilities do not have to follow this behaviour. Thus, the probability will be lower when X_i also decreases; while the probability will be higher when X_i increases.

After defining the method, I pose some questions about expatriate effectiveness over the sample period. First, I focus on expatriate adjustment. I explore whether there is any difference in the likelihood of adjustment between women and men. I also analyse whether this likelihood is different depending on the kind of adjustment. To do this, I made two estimations. Firstly, I included gender as a control variable and three independent variables, namely general, cultural and interaction adjustment, as a whole (CCA). Secondly, I distinguished between these kinds of adjustment in order to better understand gender contingencies. Table 1 shows the results of these estimations:

Table 1.

Logit model for adjustment.^a

Facets of effectiveness	Estimation 1		Estimation 2		C	McFadden	LR test
	Gender	Z-statistic	Gender	Z-statistic
GA			0.793	2.94^*	−0.271	0.022	4.19
WA			0.683	1.81^**	0.054	0.016	3.04
IA			0.660	1.88^**	−0.054	0.015	2.89
CCA	0.601	1.55			−0.162	0.012	2.45

N = 139.

GA: general adjustment; WA: work adjustment; IA: interaction adjustment; CCA: cross-cultural adjustment.

a

Convergence achieved after three iterations.

*

Denotes significance at the 95 per cent confidence level.

**

Denotes significance at the 90 per cent confidence level. LR denotes log-likelihood ratio.

As can be seen in Table 1, there were no significant gender differences in expatriate adjustment as assessed by the general factor (CCA). Nevertheless, when I analysed the three kinds of adjustment, I found that there were perceptible differences between men and women, and that these were significant at a 10% error level (estimation 2) for work adjustment and interaction adjustment. I also found that differences in the level of general adjustment are significant at a 5% error level. Specifically, I found in our sample that the likelihood of work adjustment, interaction adjustment and general adjustment was higher for females.

Next, I explored whether there were significant differences between men and women in their intention to quit the assignment. Table 2 shows the Logit model for this factor. I found no gender differences in the likelihood of premature return, as can be seen in Table 2:

Table 2.

Logit model for premature return.^a

Facets of effectiveness	Gender	Z-statistic	C	McFadden	LR test
PR	0.379	0.94	0.496	0.005	0.349

N = 139.

PR: premature return.

a

Convergence achieved after four iterations.

LR denotes log-likelihood ratio.

Next, I analysed gender contingencies for performance. I used two indicators for performance, namely job performance and self-perceived efficacy. The results for this Logit model can be seen in Table 3:

Table 3.

Logit model for performance.^a

Facets of effectiveness	Gender	Z-statistic	C	McFadden	LR test
JP	0.357	0.81	0.993	0.004	0.631
SPE	0.427	−1.06^**	0.496	0.006	1.099

N = 139.

JP: Job Performance; SPE: self-perceived efficacy.

a

Convergence achieved after four iterations.

**

Denotes significance at the 90 per cent confidence level LR denotes log-likelihood ratio.

Our data did not show significant differences in the rates of job performance. Nevertheless, when analysing self-perceived efficacy, I found differences that were significant at a 10% error level. More specifically, I found that men's self-perceived efficacy was more likely to be high than that of women.

Finally, I considered interaction between gender and two personal outcomes, namely commitment and satisfaction. As can be seen in Table 4, I did not find significant gender differences in these indicators.

Table 4.

Logit model for individual outcomes.^a

Facets of effectiveness	Gender	Z-statistic	C	McFadden	LR test
CM	0.39	1.008	−0.272	0.005	0.631
JSAT	0.17	0.42	0.613	0.001	1.099

N = 139.

a

Convergence achieved after four iterations.

CM: Organisational Commitment; JSAT: Job Satisfaction.

LR denotes log-likelihood ratio.

5

Discussion

Most research on female underrepresentation in IAs focuses on women's willingness to accept expatriation and organisational processes that lead to sending preferably men to posts abroad. However, only a few researchers have focused on the effectiveness of female expatriates. This is a crucial issue, because from an organisational approach, women should be excluded from IAs only if they were ineffectiveness.

Some past efforts had sought to test whether female expatriates might be less effective than male in their IAs, but had done so by focusing on a particular measure (CCA, early return and performance). However, to the best of our knowledge, no prior study has examined a full range of variables. In addition, research seems have not taken into consideration personal outcomes when measuring effectiveness. Moreover, most research studied only female expatriates’ effectiveness, making it impossible to make comparisons to male expatriates. To shed light on this field, I explored potential differences in effectiveness by using four groups of indicators that have been widely used in relevant literature: adjustment, premature return, performance, commitment and job satisfaction. I ran four Logit models to test gender contingencies for these indicators, and I did not find significant differences.

In relation to adjustment, I did two estimations. Looking at both of them, we noted that all the coefficients on the dummy variables were positive, indicating that the likelihood of adjustment is higher in female expatriates. Further, the estimate for general adjustment is significant at a 95% confidence level; and for interaction and work adjustment, 90%. This finding is consistent with previous studies that did not found gender differences in the level of CCA (Haslberger, 2010; Salamin & Davoine, 2015; Selmer & Leung, 2003a). In addition, and in line with these studies, I found that female expatriates have significantly higher interaction and work adjustment levels, but also a higher general adjustment level. This finding gives some support to previous research that has suggested that female expatriates may be better positioned than men to handle an expatriate assignment (e.g. Altman & Shortland, 2008; Vance & McNulty, 2014). These facets of adjustment and interaction adjustment in particular, seem relevant to the effective handling cross-cultural differences and to building the interpersonal relationships imperative to be efficient in an IA.

The second Logit model tested the effect of gender on premature return. Although the gender coefficient is also positive, the estimate was not significant. As in previous findings regarding intention to quit the assignment (Caligiuri & Tung, 1999; Forster, 1999; Tharenou, 2010), I found no significant gender differences in this rate.

As for performance, I ran Logit models for job performance and self-perceived efficacy. Regarding the former, I found no significant differences in the performance ratings of women and men, supporting previous research in this topic (Sinangil & Ones, 2003). From the perspective of the expatriate herself, the model showed, however, that women achieved slightly lower ratings than their male counterparts. It should be highlighted that this is a significant difference. It can be seen that the influence of perceived performance on the success is higher in men than in women. This is consistent with the contribution of Fischlmayr (2002), who claims that women have a lower self-perception of efficacy. The current study results support this argument, and it is important to remark that for the first time in such studies the participants were expatriates.

Finally, and for the first time to our knowledge, I found that the parameter estimate for organisational commitment and job satisfaction was positive, but not significant. This means that gender does not introduce significant differences in the likelihood of being satisfied with the job abroad or in committing to the organisation. As has been previously noted, there is a strong relationship between satisfaction in the workplace, and commitment. In turn, organisational commitment is linked with performance, and both negatively predict turnover. In sum, this finding reinforces some of the previous ones.

6

Limitations and future research

This paper has some shortcomings that should be acknowledged. First, because self-report was the most feasible way of obtaining this information for this research, all the constructs used in this study were self-reported. This procedure has been used in existing studies using variables such as employees’ attitudes and the intention to leave/remain (e.g. Kehoe & Wright, 2013). Nevertheless, the common-method variance poses a potential problem.

In order to assess whether common method variance may have impacted the obtained results, Harman's one-factor test was used to analyse the presence of common method bias. All the studied variables were entered into an exploratory factor analysis, using unrotated principal components factor analysis and principal component analysis with varimax rotation. The largest factor only accounted for the 10.83% of the variance. Considering these results, common method variance is not a major problem in this study. In addition, the questionnaire used was designed following the recommendations by Podsakoff, MacKenzie, Lee, and Podsakoff (2003) and Podsakoff et al. (2012), to guarantee anonymity to the respondents and to separate the dependent variables from the independent ones. Some of these recommendations can also help to avoid self-generated validity effects. These include: (1) temporal separation; (2) increasing the physical distance between constructs that are highly diagnostic for each other, and (3) randomised questionnaire designs. Whereas temporal separation was not feasible given the constraints of time and cost, questions regarding the physical distance between measures were taken into account when in the design of the questionnaire.

There are also limitations regarding the generalisation of these findings to other expatriates, because the study is limited to Spanish respondents, and also because of the high proportion of respondents whose assignments were in European Union countries (65.33%). Moreover, all respondents were AEs. Having in account that companies currently use a wide range of business expatriates (e.g. short-term, long-term, business travellers) further research is needed to address this variability. I also acknowledge that our use of a convenience sample may cause some concern, since this sample came from companies and institutions with which the researcher had contact. Therefore, future research should analyse a larger random sample in order to assess the significance of the obtained findings, including, if possible, different expatriate home and host countries.

I also suggest some avenues for future studies. Firstly, as job satisfaction and commitment was shown to be related to other criteria for expatriate effectiveness, future research should examine gender differences on these personal outcomes from IAs in connection with other measures of effectiveness. For example, higher levels of job satisfaction might serve as a source to increase self-perceived efficacy and, as a consequence, can contribute to reduce gender differences on this measure. Secondly, the lower self-efficacy with women could be linked to perceived self-confidence, and might be the object of future research.

7

Conclusions

This study yields two major findings. While previous research has explored female expatriate effectiveness from limited points of view, I analysed four groups of measures that have been widely employed in the international human resource management literature. Our findings provide further evidence to suggest that women are as effective as men in IAs, as all the indicators used to measure this effectiveness confirm. A second contribution is related to the personal results from IAs. To the best of our knowledge, prior studies have not explored gender contingencies on organisational commitment and job satisfaction in IAs. Our study confirms that there are not significant differences in these outcomes between expatriate women and men. Moreover, because I found that female expatriates have higher interaction, work and general adjustment levels than male, our findings give some support to previous research that has suggested that women enjoyed unique advantages for the handling of IAs. In sum, this study contributes to the literature by expanding the link beyond the initial phases of IAs, such as female’ willingness to accept an international experience, and female expatriates’ effectiveness in host countries. These results may have implications for international human resource management, because to achieve their international strategies, organisations should encourage their employees to undertake and successfully complete IAs, regardless of their gender. Thus, the processes for selecting managers for international assignments should consider all their talented human resources – including women – in order to improve hiring decisions. This trend could improve female representation in international experiences, on the short term. In turn, on the long term, women could improve their opportunities to be promoted to higher management positions. As Vance and McNulty (2014:37) put it, the under representation of women in IAs “represents not only a breach of ethics and fairness, but also represents an inefficient use of a multinational firm's talent pool in the face of increasing global competition”.