Introduction

In international business (IB) research, boundary-spanning refers to a set of inter- and intraorganizational communication and coordination activities aimed at enhancing integration amid a wide range of cultural, institutional, and organizational differences in multinational enterprises (MNEs) (Schotter et al., 2017). MNEs often rely on expatriates to execute boundary-spanning (Au & Fukuda, 2002; Yang et al., 2022) due to their cultural and language skills (Barner-Rasmussen et al., 2014) and their ability to transfer explicit and tacit knowledge between clusters (Harzing et al., 2016). When expatriates are assigned to foreign subsidiaries, their boundary-spanning helps MNEs address the challenges of reconciling variations in language, culture, and behavioral norms worldwide (Au & Fukuda, 2002) and of establishing a consistent corporate strategy (Meyer et al., 2020).

As such, IB research often emphasizes a positive view of expatriates’ boundary-spanning, with a prominent focus on utilizing social capital theory to explain this advantage. This theory proposes that the social relationships expatriates create through boundary-spanning activities are considered “social capital” from which individuals and organizational units can gain economic and relational benefits (Kostova & Roth, 2003). Studies employing this theoretical lens have linked the boundary-spanning of globally mobile employees (e.g., expatriates) to positive outcomes, such as improved adjustment and performance (Liu & Shaffer, 2005), knowledge-sharing (Mäkelä, 2007), and inter- and intraunit connections (Reiche et al., 2009). In contrast, research outside the IB field demonstrates that boundary-spanning may also lead to negative outcomes, such as role overload (Wang et al., 2019) and role stress (Stamper & Johlke, 2003). Role theory, which states that social roles shape and influence individuals’ behavior (Anglin et al., 2022), is commonly employed to explain these adverse effects.

In this study, we integrate both social capital theory and role theory to explore the outcomes of expatriates’ boundary-spanning. Given that successful boundary-spanning relies on collaboration with host-country employees (Hsu et al., 2022), we mainly examine its impact on expatriate–host-country employee relationships – an aspect underexplored in the current literature. Thus, we address the following research question: What are the potential positive and negative consequences of expatriates’ boundary-spanning for the relationships between expatriates and host-country employees in MNEs? Our study contributes to the boundary-spanning literature in the IB field by offering a novel, empirically validated theoretical framework of expatriates’ boundary-spanning, which provides a more realistic, balanced, and accurate perspective. It also explains three distinct dimensions of expatriates’ boundary-spanning – functional, linguistic, and cultural – in producing both positive and negative relational outcomes. Our framework also offers practical implications for MNEs and globally mobile employees.

Theory and hypotheses

Expatriates’ boundary-spanning

In our review of the literature, we identified three dimensions of expatriates’ boundary-spanning. Functional boundary-spanning utilizes expatriates’ domain-specific knowledge and expertise to transfer core competencies between units, narrow knowledge gaps, and standardize management practices (Harzing et al., 2016). Linguistic boundary-spanning aims to address language-related issues beyond functional realms (Harzing & Pudelko, 2014). It involves more than literal translation as language serves as an important means of transmitting tacit and explicit knowledge (Welch & Welch, 2008). Cultural boundary-spanning refers to activities that exploit expatriates’ cross-cultural knowledge and understanding to navigate cultural differences and establish connections (Backmann et al., 2020). In what follows, we theorize on how these dimensions of expatriates’ boundary-spanning collectively influence the relational outcomes between expatriates and host-country employees and how each dimension contributes to such mechanisms. Our theoretical framework is depicted in Fig. 1.

Fig. 1
figure 1

Theoretical model for the current research

Full size image

The double-edged effects of expatriates’ boundary-spanning

Social capital theory proposes that individuals can derive mutual benefits from their relationships with others (Kostova & Roth, 2003). Social capital is represented by social interaction, trust, and shared visions and goals between individuals (Reiche et al., 2009; Tsai & Ghoshal, 1998). Through boundary-spanning, expatriates and host-country employees collaborate closely, forming interpersonal ties that are the foundation of social capital (Shimoda, 2013). Such collaboration and connection allow each party to interact and gain what they need from the respective working relationship, thus fostering reliance on each other. Reciprocity also plays a significant role in shaping these exchange relationships (Molm et al., 2007). In turn, this task-related and informal interdependence and reciprocity contribute to the development of mutual trust. Mutual trust is characterized by two parties displaying complementary levels of vulnerability in a shared context based on positive expectations of each other’s intentions and behaviors (Korsgaard et al., 2015). This kind of trust is a crucial aspect in the interpersonal relationships between expatriates and host-country employees (Shimoda, 2013).

Expatriates’ boundary-spanning and the resulting formation of mutual trust help disseminate the shared values and collective goals of organizations (Tsai & Ghoshal, 1998), particularly those with which these individuals may have had prior limited familiarity (e.g., subsidiaries for expatriates and MNEs for host-country employees). In turn, both expatriates and host-country employees gain a deeper understanding of and stronger affective connection with their counterpart organizations (Schaubroeck et al., 2013), thereby cultivating a stronger sense of organizational identification (Tsai & Ghoshal, 1998). As a result, expatriates can feel a sense of belonging to subsidiaries, and host-country employees perceive a closer connection with MNEs. Altogether, through the development of mutual trust, expatriates’ boundary-spanning enhances the mutual identification pattern for both expatriates and host-country employees, strengthening their sense of belonging to or identification with their counterpart organizations (Rousseau, 1998).

Hypothesis 1

Expatriates’ boundary-spanning facilitates the development of mutual trust between expatriates and host-country employees, which in turn promotes (a) expatriates’ subsidiary identification and (b) host-country employees’ MNE identification.

Role theory suggests that the varied expectations embedded within their roles can create stress for individuals (Kahn et al., 1964). Therefore, when working under implicit pressure and juggling multiple roles, including their official jobs and on-demand boundary-spanning activities, expatriates often experience work-related stressors, such as role ambiguity, role conflict, and role overload (Au & Fukuda, 2002; Stamper & Johlke, 2003). These role stressors exacerbate adverse effects for expatriates and can potentially lead to emotional exhaustion (Zhao & Jiang, 2022), which is characterized by the feeling of being overextended and depleted of psychological and physical resources (Maslach et al., 2001). Specifically, the challenge of managing multiple unclearly defined roles frequently exposes expatriates to overwhelming stressors, making them susceptible to emotional exhaustion (Bhanugopan & Fish, 2006).

Furthermore, high levels of role stressors may leave expatriates with limited physical, emotional, and cognitive resources to handle their workload (Bakker & Demerouti, 2007). In response, expatriates may activate self-regulatory mechanisms to alleviate stress (Repetti, 1992). These mechanisms involve surface-level social integration (Fan et al., 2023) whereby expatriates limit their interactions with host-country employees to only necessary boundary-spanning activities while refraining from using their already-scarce resources to conduct extra activities or adjusting their attitudes and behaviors to socialize with host-country employees. This diminished adaption and socialization may lead host-country employees to view expatriates as dissimilar to themselves in terms of values, attitudes, and behaviors. Consequently, host-country employees may perceive stressed expatriates as not belonging to their ingroup, thereby categorizing them as outgroup members (Toh & DeNisi, 2007).

Hypothesis 2

Expatriates’ boundary-spanning increases their role stressors, which in turn (a) exacerbates their emotional exhaustion and (b) increases their chances of being categorized as outgroup members by host-country employees.

The effects of the subdimensions of expatriates’ boundary-spanning

Each dimension of expatriates’ boundary-spanning has a distinct primary knowledge base (Barner-Rasmussen et al., 2014) and composition of explicit and tacit elements (Li & Scullion, 2010). Accordingly, we refine the double-edged framework by examining the role of each dimension separately. In theorizing their effects on mutual trust, we differentiate two forms of trust discussed in the literature: cognition-based trust, which is based on the rational evaluation of a person’s reliability, dependability, and competency, and affect-based trust, which builds on emotional bonds (McAllister, 1995).

First, functional boundary-spanning mainly deals with the domain-specific capabilities and firm-specific behavioral standards that expatriates bring to foreign subsidiaries (Harzing et al., 2016; Takeuchi et al., 2019). Dealing with these types of explicit knowledge demonstrates expatriates’ abilities and behavioral integrity, which likely increases host-country employees’ cognition-based trust in them (Tomlinson et al., 2020). Additionally, host-country employees likely reciprocate expatriates’ boundary-spanning efforts by providing information to assist them in functioning and problem-solving in the host country (Farh et al., 2010) and in delivering expected work outcomes (Ang & Tan, 2016). Such reciprocal actions demonstrate host-country employees’ abilities and behavioral integrity, likely fostering expatriates’ cognition-based trust in them.

Next, cultural boundary-spanning requires individuals to accommodate different values, beliefs, and norms (Johnson et al., 2006), thereby transferring more tacit knowledge, which is difficult to formalize and is practice and context-specific (Ambrosini & Bowman, 2001). In this process, the use of cultural knowledge and metacognition influences the development of affect-based trust (Chua et al., 2012). Namely, when expatriates use their knowledge to interpret situations and interactions unique to different cultural contexts, their adept behavior helps prevent misunderstanding and potential tensions that may otherwise hinder the formation of affect-based interpersonal relationships. In this process, host-country employees also tend to extend empathetic emotional support to expatriates experiencing adjustment difficulties (Farh et al., 2010). Such reciprocal behavior fosters host-country employees’ affect-based trust in expatriates and vice versa.

Compared with functional and cultural boundary-spanning, the relationships between linguistic boundary-spanning and affect- and cognition-based mutual trust are both likely moderate. Research suggests that limited shared language proficiency in multinational teams may reduce members’ perceived task competence and dependability, which may diminish their perceived ability- and integrity-based trust. This situation can also evoke language-based anxiety among team members with lower language proficiency that reduces their emotion-based intentions to trust by not showing their vulnerabilities to people with higher language proficiency (Tenzer et al., 2014). Thus, expatriates who perform linguistic boundary-spanning likely facilitate shared language proficiency with host-country employees and, accordingly, develop cognition-based and affect-based mutual trust. However, even with a shared language, language barriers may persist in the form of pragmatic and prosodic differences (Tenzer et al., 2021). Thus, although linguistic boundary-spanning may resolve some visible language barriers and enhance mutual trust, hidden language barriers may still trigger negative cognitive and affective responses in others.

Hypothesis 3

(a) Functional boundary-spanning is more strongly related to cognition-based mutual trust and (b) cultural boundary-spanning is more strongly related to affect-based mutual trust than the other dimensions of boundary-spanning.

Performing functional boundary-spanning tends to be seen as within the realm of expatriates’ formal responsibilities, which reinforces high demands and expectations from host-country employees. These heightened demands and expectations may lead to additional workload and role stressors for expatriates. The complex work demands from different parties can trigger expatriates’ role ambiguity and role conflict, especially when they receive insufficient support from host-country employees, causing a significant burden to fall on expatriates’ shoulders. On the other hand, despite the mental challenges of working with foreign languages (Volk et al., 2014) and the cognitive effort required to interact with diverse cultures (Rozkwitalska, 2017), expatriates can often mitigate role stressors stemming from linguistic and cultural boundary-spanning for two key reasons. First, both expatriates and host-country employees play a role in enhancing mutual linguistic and cultural understanding. While expatriates often make efforts to learn local languages and cultures (Zhang & Harzing, 2016), host-country employees also contribute to fostering mutual understanding by acting as cultural interpreters and communication facilitators (Vance et al., 2009). Second, given that expatriates’ positions connect headquarters and subsidiaries, they may anticipate and internalize expectations related to bridging linguistic and cultural gaps (Wiener, 1982). As a result, expatriates often demonstrate relatively high efficiency in coping with role stressors associated with linguistic and cultural boundary-spanning.

Hypothesis 4

Functional boundary-spanning is more strongly related to role stressors than the other dimensions of boundary-spanning.

Methods

Data and participants

We collected three datasets in 2022, the first two of which were used to develop a scale to measure expatriates’ boundary-spanning. Dataset 1 was collected through a Japanese online research panel for exploratory factor analysis (EFA). It included 355 participants (57% male, Mage = 39.04) who were Japanese expatriates within the last 3 years. Dataset 2 was collected through a Chinese online research panel for confirmatory factor analysis (CFA) and included 660 employees (330 Chinese expatriates and 330 host-country coworkers, 50.94% female, Mage = 36.92) who worked in a Chinese MNE within the last 3 years. Dataset 3 was the main dataset and was collected through author connections in the energy engineering industry of Chinese MNEs based in Laos, Vietnam, the Philippines, and Malaysia. We chose this industry due to the significant deployment of expatriates resulting from the “Belt and Road Initiative,” providing us access to paired data for expatriates and host-country employees. We collected dyadic data (64% overall response rate) from a total of 177 expatriate–host-country coworker pairs in two waves with a 2-week interval using paper-based questionnaires or online surveys (expatriates: Mage = 33.91, 50.28% female, Msubsidiary tenure = 2.63; host-country coworkers: Mage = 31.84, 51.98% female, Msubsidiary tenure = 3.37).

Measures

All measures (see the online appendix for the full scales) other than the control variables were assessed using a seven-point Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree). To ensure scale validity and verb accuracy, we conducted a back-translation process (Brislin, 1980).

Independent variable

We developed a new scale for expatriates’ boundary-spanning following Hinkin’s (1998) scale-development procedure. From 70 initial items generated through interviews and the literature, we retained 29 items after assessing clarity and content validity. EFA and CFA on two separate datasets validated the measure. The EFA with Oblimin rotation on Dataset 1 (n = 355) supported a three-factor solution. After eliminating three items with cross-loadings, our final 26-item scale includes nine items for functional, seven items for linguistic, and ten items for cultural boundary-spanning (see the appendix). Next, the CFA results from Dataset 2 (n = 660) supported the proposed three-dimensional structure of this scale, χ2[296] = 1077.60, p < 0.01; Tucker–Lewis index (TLI) = 0.94; comparative fit index (CFI) = 0.94; standardized root mean-square residual (SRMR) = 0.03; root mean squared error of approximation (RMSEA) = 0.06. Similar results were obtained from the 177 dyads from Dataset 3 (reported by host-country coworkers; χ2[296] = 409.17, p < 0.01; TLI = 0.96; CFI = 0.97; SRMR = 0.04; RMSEA = 0.05). The EFA and CFA demonstrated consistent results among the three reflective indicators, allowing us to compute expatriates’ boundary-spanning as a multidimensional construct comprising functional, linguistic, and cultural dimensions. In line with previous research [e.g., Connelly et al. (2019)], we utilized expatriates’ boundary-spanning as the entire measure when we tested the hypotheses in which the overall boundary-spanning construct was of interest, whereas to test the hypotheses dealing with the specific types of boundary-spanning separately, we employed each indicator as a separate measure.

Mediating variables

Mutual trust, cognition-based mutual trust, and affect-based mutual trust were reported by both expatriates and host-country coworkers (11 items; McAllister, 1995). We found a high level of agreement in trust between host-country employees and expatriates: interrater agreement (rwg[j]) (James et al., 1984) showed rwg[j] values of 0.86 for mutual trust, 0.80 for cognition-based mutual trust, and 0.67 for affect-based mutual trust. Thus, we computed the measures for mutual trust, cognition-based mutual trust, and affect-based mutual trust based on the indicator of the square root of the product (SRP) (Smith & Barclay, 1997) because the SRP includes both level and agreement facets:

$${\text{SRP}}\;\left( {{\text{mutual trust}}} \right)\; = \;\left[ {\left( {{\text{Expatriates}}\hbox{'} {\text{ trust}}} \right)\; \times \;\left( {{\text{Host}} - {\text{country coworkers}}\hbox{'} {\text{ trust}}} \right)} \right].^{{{\text{1}}/{\text{2}}}}$$

Role stressors were indicated by expatriates (17 items; Rizzo et al., 1970; Beehr et al., 1976).

Dependent variables

Subsidiary identification (reported by expatriates) and MNE identification (reported by host-country coworkers) were both measured using an organizational identification scale (six items; Mael & Ashforth, 1992). Emotional exhaustion was reported by expatriates (nine items; Maslach & Jackson, 1981), and outgroup categorization was reported by host-country coworkers (seven items; Leonardelli & Toh, 2011).

Control variables

We included expatriates’ and host-country coworkers’ age, gender, industry, tenure in their subsidiaries, and foreign language proficiency (Liu & Jackson, 2008).

Results

Preliminary results

We first examined the variance inflation factors (VIFs) to test for multicollinearity issues. The VIFs ranged from 1.06 to 2.38, falling far below the accepted cutoff value of 10 (Myers, 1990) and thereby indicating no multicollinearity. To mitigate potential common method variance (CMV), we followed the suggestions of Podsakoff et al. (2003) and collected measures from different sources and at different time points in the research-design stage. Furthermore, we conducted a Harman’s single-factor test and found that the variance explained by the first factor was 25.58%, falling short of accounting for half of the total variance. We also performed unmeasured latent common factor analysis. Despite a significant chi-square difference (Δχ2 = 271.23, Δdf = 39, p < 0.05) before and after including the method factor, the hypothesized factors explained 56.92% of the total variance, whereas the method factor explained 17.98%. These findings led us to conclude that CMV did not play a major role in our analysis.

We also conducted higher-order CFA to validate our higher-order model with ten first-order factors and two second-order factors. Expatriates’ boundary-spanning and mutual trust were second-order factors, while the remaining factors were considered first-order factors. Since the original measures consisted of many indicators with a limited sample size, we reduced the number of indicators for each latent variable using item parceling [e.g., Sekiguchi et al. (2017)]. The results indicated that the hypothesized model fit the data well (χ2[676] =1166.90, p < 0.01; TLI = 0.92; CFI = 0.92; SRMR = 0.08; RMSEA = 0.06). The descriptive statistics are presented in Table 1.

Table 1 Descriptive statistics and correlations
Full size table

Hypothesis testing

To test Hypotheses 1a, 1b, 2a, and 2b, we conducted bootstrapping-based mediation tests using the PROCESS macro (Hayes, 2018) with 5000 resamples to produce 95% confidence intervals (CIs) around the estimated indirect effects. Tables 2 and 3 report the regression results for the positive and negative paths in our model. According to the mediation results, expatriates’ boundary-spanning was associated with expatriates’ subsidiary identification (indirect effect = 0.20, SE = 0.07, 95% CI [0.07, 0.36]) and host-country coworkers’ MNE identification (indirect effect = 0.22, SE = 0.08, 95% CI [0.08, 0.40]), both of which were mediated by mutual trust. Thus, Hypotheses 1a and 1b were supported. Expatriates’ boundary-spanning was also associated with emotional exhaustion (indirect effect = 0.41, SE = 0.09, 95% CI [0.25, 0.60]) and outgroup categorization (indirect effect = 0.26, SE = 0.08, 95% CI [0.12, 0.43]), both of which were mediated by role stressors. Therefore, Hypotheses 2a and 2b were supported.

Table 2 Regression results for the positive effects of expatriates’ boundary-spanning
Full size table
Table 3 Regression results for the negative effects of expatriates’ boundary-spanning
Full size table

To test Hypotheses 3a, 3b, and 4, we performed relative weight analysis to examine the relative importance of each type of boundary-spanning on cognition-based and affect-based mutual trust and role stressors (Tonidandel & LeBreton, 2011). The results are shown in Table 4 and Fig. 2 and provide support for Hypothesis 3b and 4. Hypothesis 3a was not supported because cultural boundary-spanning had the strongest relationship with cognition-based mutual trust followed by functional and linguistic boundary-spanning. In summary, all our hypotheses except Hypothesis 3a were fully supported.

Table 4 Results of the relative weight analysis
Full size table
Fig. 2
figure 2

Relative weight analysis results

Full size image

Discussion and conclusion

Theoretical contributions

The main theoretical contribution of this study is that we develop a comprehensive framework of the double-edged phenomenon of expatriates’ boundary-spanning that incorporates two theoretical lenses – namely, social capital theory and role theory. Through a social capital theory lens, we identify several positive relational outcomes of expatriates’ boundary-spanning. Specifically, it benefits organizations by increasing mutual trust between expatriates and host-country employees, which further leads to a mutual organizational identification pattern. On the other hand, through a role theory lens, our study reveals the potential dark side of expatriates’ boundary-spanning, which can result in negative emotional outcomes and outgroup categorization due to expatriates’ conflicting roles. By combining these two theoretical lenses, our framework sheds light on both the positive and negative relational outcomes of expatriates’ boundary-spanning, providing a more accurate and realistic understanding of this phenomenon in the IB field.

Our study further contributes to the theoretical understanding of how different types of expatriates’ boundary-spanning – namely, functional, linguistic, and cultural – contribute to double-edged relational outcomes. First, our findings suggest that functional boundary-spanning primarily fosters cognition-based mutual trust mainly through the exchange of explicit knowledge (Harzing et al., 2016; Takeuchi et al., 2019), whereas cultural boundary-spanning promotes not only affect-based mutual trust but also cognition-based mutual trust. These findings might reflect how engaging in cultural boundary-spanning helps transform context-specific and hands-on knowledge (i.e., tacit knowledge) into clear and comprehensible information (i.e., explicit knowledge) during cross-cultural knowledge transfer. Thus, while cultural boundary-spanning develops affect-based mutual trust through tacit knowledge-sharing, it also promotes cognition-based mutual trust through explicit knowledge-sharing, making it the most influential in developing social capital. In contrast to functional and cultural boundary-spanning, linguistic boundary-spanning does not have a strong direct influence on mutual trust. Rather, it might be a necessary condition for other types of boundary-spanning to operate effectively (Backmann et al., 2020; Barner-Rasmussen et al., 2014). Indeed, language is deeply embedded in the IB context, and overcoming language barriers is crucial for functional and cultural boundary-spanning to promote mutual trust (Tenzer et al., 2014).

Second, our findings show that all three types of boundary-spanning cause role stressors and subsequent negative relational outcomes, which are potential costs in cultivating social capital and achieving positive outcomes. In particular, functional and cultural boundary-spanning require greater effort from expatriates to understand and exchange knowledge with host-country employees, resulting in relatively strong role stressors. In contrast, linguistic boundary-spanning requires less effort as language processing becomes more automatic as one’s foreign language proficiency increases (Volk et al., 2014). Similar to the case of developing social capital, linguistic boundary-spanning can be an essential medium in facilitating other types of boundary-spanning (Barner-Rasmussen et al., 2014). Altogether, our findings suggest that cultural boundary-spanning has the strongest double-edged effects; linguistic boundary-spanning shows the weakest effects; and functional boundary-spanning presents asymmetric effects, with negative outcomes possibly exceeding positive outcomes.

Practical implications

Our study also has practical implications for MNEs and managers. First, MNEs should implement policies that amplify the positive effects of expatriates’ boundary-spanning while mitigating the negative ones. To maximize the positive effects, global managers should assess expatriates’ functional, linguistic, and cultural knowledge to ensure they possess the necessary skills and offer training if necessary to meet the demands and expectations of host-country employees. To minimize the negative effects, global managers should clarify job responsibilities to help expatriates anticipate and cope with challenges and be attentive to expatriates’ well-being by providing more organizational support and encouraging assistance from host-country employees.

Second, our implications extend to other globally mobile employees, such as self-initiated expatriates and third-country nationals engaged in boundary-spanning within MNEs’ global operations (Barmeyer et al., 2020; Furusawa & Brewster, 2019). They should be aware of both the positive and potential detrimental impacts of boundary-spanning. As more people engage in cross-border work, our findings can guide them in effectively navigating boundary-spanning in multinational operations.

Limitations and future research directions

First, our main dataset might have been vulnerable to selection bias due to our use of paired data from only one coworker per expatriate and our adoption of the singular energy engineering industry comprising only Chinese MNEs. Additionally, considering that our main dataset was collected in 2022, the second year of the COVID-19 pandemic, host-country employees’ responses to Chinese expatriates may have been influenced. Thus, future studies could collect dyadic data from multiple coworkers per expatriate and broaden the sample by including data from other industries, time periods, and countries. Second, despite the use of multisource data, our study design has limitations related to causal inference. Future studies could adopt a longitudinal design and examine the development of expatriate–host-country employee relationships over time. Such an approach could constructively replicate and extend the results we obtained. Third, we only measured the organizational identification of expatriates with subsidiaries and of host-country employees with MNEs. Future studies could measure dual organizational identification (Ishii, 2012; Vora & Kostova, 2007) and explore the dynamic relationship between the two foci of organizational identification.

Future research could also refine or extend our theoretical framework by examining different aspects of social capital theory and role theory. For example, when applying social capital theory, one promising research avenue is exploring how culture influences the dynamics of mutual trust between expatriates and host-country employees. This research is relevant given that various cultures foster different types of trust (Chua et al., 2012), and it is significant to understand how inter-cultural trust evolves in hierarchical relationships (Bueechl et al., 2023). Such an investigation could add nuanced insights into the double-edged effects of boundary-spanning. When applying role theory, it would be beneficial to explore other potential consequences of role stressors like counterproductive behavior (Belschak & Den Hartog, 2009). Additionally, future research could explore possible positive outcomes (Au & Fukuda, 2002), such as role consensus (Anglin et al, 2022) and role novelty (Kawai & Mohr, 2015). In conclusion, the double-edged framework we proposed and empirically validated in this study simultaneously highlights both the bright and dark sides of expatriates’ boundary-spanning and opens new avenues for research on globally mobile individuals and the impact of their boundary-spanning activities in the MNE context.