The Interplay of Subsidiary Performance and Cultural Distance in International Downsizing Decisions

Existing literature has primarily identified financial and/or strategic considerations as drivers of international downsizing decisions. A subsidiary’s performance seems to have a superior meaning among those considerations. Cultural distance is also an important factor, although its influence on downsizing is ambiguous: It increases coordination costs and, hence, lowers performance; however, cultural distance might also entail benefits of diversity and diversification. We challenge prior perspectives on the role of cultural distance in international downsizing decisions by showing that cultural distance also bears influences beyond efficiency considerations. Using insights from social identity and self-categorization theory, we argue that cultural distance influences how decision-makers interpret the efficiency of the subsidiary in terms of its performance. We expect that the lower the cultural distance, the more positively performance outcomes will be interpreted, leading to less workforce downsizing in a culturally proximate subsidiary as compared to one that is culturally more distant. Results of our large-scale empirical analysis of more than 14,000 downsizing decisions of German MNCs throughout 60 industries and across 54 host countries support our predictions.


Introduction
Workforce downsizing, understood as an intentional, permanent, and systematic reduction of an organization's workforce (e.g., Freeman and Cameron 1993), represents a widespread management practice (Datta et al. 2010). Especially in times of macroeconomic turbulence, downsizing is a very common tool for firms to fight financial distress. A multinational corporation (MNC) that employs personnel in a multiplicity of countries must tackle the difficult question which foreign subsidiary will be downsized and is obliged to balance layoff decisions across diverse regulatory and cultural environments (Pajunen 2008;Pull 2008). Such de-internationalizing decisions represent an important aspect of the cycles and waves of internationalization that MNCs undergo (e.g., Berry 2013). It is also not unusual that corporate restructuring strategies of MNCs, which involve layoffs at larger scale in a host country, produce significant resistance from stakeholders inside and outside the firm (Blazejewski 2009;Carroll 1984).
In studying these difficult decision processes, previous studies on international downsizing have largely attributed the decision to downsize foreign workforce to efficiency-based considerations (e.g., Arte and Larimo 2019;Berry 2013). In this regard, prior literature has primarily identified the discrepancy between expected and actual performance outcomes (Jagersma and van Gorp 2003) as well as strategic considerations (Benito and Welch 1997;Boddewyn 1979)-such as shifting resources to locations where they are expected to work more efficiently-as reasons for international downsizing decisions.
However, the view proposing strategic decisions to only be governed by efficiency-based considerations can be challenged by approaches that advocate a 'behavioral' perspective in organization and strategy research (e.g., Aharoni 1966;Cyert and March 1963;Powell et al. 2011). This alternative view suggests that decisionmakers are individuals with different backgrounds and motivations and it aims "to strengthen the empirical integrity and practical usefulness of strategy theory by grounding strategic management in realistic assumptions about human cognition, emotion, and social interaction" (Powell et al. 2011(Powell et al. , p. 1369. In this view, decisions are not solely driven by efficiency-based considerations but also by personal attitudes and preferences such as, for example, a tendency to allocate resources evenly across different business lines, regardless of their individual efficiency (Bardolet et al. 2011).
We include this broader perspective on strategic decision-making in our study when analyzing the interplay of cultural distance and subsidiary performance on foreign downsizing decisions. Within the predominating efficiency-based logic, cultural distance between home and host country has mostly been seen as a factor that creates additional costs of coordination for the MNC (e.g., Johanson and Vahlne K 1977;Hutzschenreuter and Voll 2008). Therefore, it is assumed to lower the performance outcomes of foreign subsidiaries leading to higher downsizing (or divestment) propensities (Barkema et al. 1996;Berry 2013;Li 1995). We ask, however, whether cultural distance has effects on international layoff processes that go beyond the efficiency-based considerations demonstrated by prior literature. Specifically, we base our arguments on insights from social identity and self-categorization theory (Hogg and Terry 2000;Tajfel 1982;Tajfel and Turner 1986;Turner 1987), which suggest that individuals define and differentiate themselves from others on the basis of salient and observable characteristics. Furthermore, individuals use this assessment to construct in-groups of similar others, whom they feel more attracted to and outgroups of dissimilar others, vis-à-vis whom they feel more alienated and greater personal distance. We apply these insights to firms' international downsizing decisions and argue that decision-makers are more willing to dismiss employees whom they perceive to be culturally different to themselves. We argue that this behavioral tendency influences how MNC decision-makers assess subsidiary performance: The more culturally proximate a foreign subsidiary is, the more positive or lenient the interpretation of subsidiary performance as a predictor of the amount of dismissals will be. That means, on the other hand, when cultural distance is high, decision makers will interpret subsidiary performance more severely and will be more likely to dismiss employees. Overall, we expect that downsizing decisions in foreign subsidiaries will be based on different evaluations of performance, depending on the cultural distance of that subsidiary's host country.
To analyze the interaction of subsidiary performance and cultural distance in international downsizing decisions, we employ an extensive dataset comprising more than 14,000 downsizing decisions of German MNCs throughout 60 industries and across 54 host countries, including the explicit financial figures of the foreign subsidiaries. With this study, we contribute to extant literature on the process of MNC decisionmaking regarding foreign downsizing. We show that in addition to efficiency-related or strategic reasons, non-economic factors, specifically the perception of cultural similarity, influence firms' decisions to dismiss employees in foreign locations. Due to our study's deep level of data availability on the subsidiary level, we are able to discern economic and non-economic effects of cultural distance in international downsizing decisions more than previous studies have been able to.
We proceed as follows: In Sect. 2, we develop our theoretical framework and formulate our baseline and main hypotheses. In Sect. 3, we describe our empirical research design before we present the results in Sect. 4. Section 5 discusses the findings and implications for management practice and future research. Section 6 concludes the study with a short summary.

Theory and Hypotheses
Prior research has shown that reducing commitment in a foreign venture can have a number of different motivations, mostly related to the efficiency of the firm (e.g., Berry 2013). These studies have suggested that the major reason why firms withdraw from foreign subsidiaries is a discrepancy between expected and actual outcomes.
MNCs often lack information at the moment of entering the foreign market and are not able to correctly predict future developments of a foreign venture. Only after investments have been made, the MNC is able to observe if affiliates deliver the intended results. If the foreign ventures perform below expectations, a likely consequence has been shown to be the termination of these activities (Boddewyn 1979;Coudounaris et al. 2020;Jagersma and van Gorp 2003). Prior research has viewed these financial considerations as one of the strongest drivers of divestment (Benito and Welsh 1997;Berry 2013;Fisch and Zschoche 2012), an effect that has been confirmed in a recent meta-analysis by Schmid and Morschett (2020). 1 Based on these effects that have been established in prior literature, we begin our theorizing with the following baseline hypothesis.
Baseline: Subsidiary performance has a negative effect on the number of dismissals in a firm's foreign subsidiary.
In addition to a subsidiary's performance, prior literature has also been concerned with the relationship between its cultural distance to the home country and dismissals in that foreign subsidiary. Cultural distance is one of the essential ingredients in the analysis of MNC decision-making (Beugelsdijk et al. 2018b;Håkanson and Ambos 2010;Stahl and Tung 2015). Hence, it has played a central role in previous studies on international divestment (Barkema et al. 1996;Benito 1997;Pattnaik and Lee 2014;Mohr et al. 2018;Schmid and Morschett 2020). Within this literature, the majority of scholars have suggested differences between home and host country, for example in attitude, values, or behavior, to create additional costs of coordination for the MNC (Beugelsdijk et al. 2018b;Johanson and Vahlne 1977;Kostova et al. 2016;Li and Guisinger 1992). Cultural distance, in this regard, has been understood as the liability of foreignness a firm faces in a culturally distant host country (Lou and Mezias 2002). According to prior studies, this liability will result in challenges related to knowledge transfer, information processing, as well as the general adaption to the foreign market (e.g., Javidan et al. 2005;Schmid and Morschett 2020;Sousa and Tan 2015).
Although based on these theoretical considerations, there is a clear mandate to predict a positive relationship between cultural distance and the number of dismissals in foreign subsidiaries, recent results have been less conclusive, often showing insignificant effects (e.g., Coudounaris et al. 2020;Schmid and Morschett 2020). In explaining these ambivalent findings, Schmid and Morschett (2020), for example, have argued that firms may realize that adaption to culturally distant countries simply takes time and be less of a problem after an initial phase of adjustment (e.g., Sousa 1 Our approach to international downsizing using the number of dismissals in a firm's foreign subsidiary is more nuanced compared to previous research that has very often measured divestment as a binary variable only (e.g., Dai et al. 2013;Pattnaik and Lee 2014). We chose this approach for two reasons: (1) Our study's level of data availability allows us to use a more fine-grained measurement than that commonly used in prior research. (2) From a theoretical perspective, our approach allows for a broader understanding of the organizational decline trajectory (e.g., Cameron et al. 1987;Whetten 1980) that may end in divestment. At the same time, we recognize that different strategic motivations could lie behind a firm's decision to dismiss employees versus the decision to divest. We address this issue at length in the discussion section of our study. and Tan 2015; Wilkinson et al. 2008). Hence, firms may not react to performance shortcomings of their foreign subsidiaries immediately. Furthermore, other recent studies have argued for positive performance effects of cultural distance by highlighting the economic benefits that the diversity and cross-market learning inherent to cultural distance may bring (Gomez-Mejia and Palich 1997;Lücke et al. 2014;Park and Ungson 1997). Particularly Stahl and his colleagues (e.g., Stahl and Tung 2015;Stahl et al. 2016;Tung and Stahl 2018) have argued for a more balanced treatment of cultural distance in international business studies, which, instead of overemphasizing a negative view on foreignness, distance and related differences of all kinds, should-in line with a positive organizational scholarship perspective-focus more on the dynamics, processes, and conditions that enable organizations to benefit from diversity.
In this study, we add to this discussion by employing a behavioral interpretation of cultural distance to explain why cultural distance may influence a firms' assessment of a foreign subsidiary's performance and hence its economic contribution to the parent firm. We introduce behavioral aspects of cultural distance as a novel determinant in international downsizing decisions for two reasons.
First, with regard to general decision-making, a growing research stream has evolved during the last decades which demonstrates that decision-makers do not behave strictly according to efficiency considerations but that there are also cognitive biases and socio-psychological mechanisms that direct human decisions in a seemingly 'less rational' manner (e.g., Kahneman and Tversky 1979;Powell et al. 2011). The analysis of these aspects of decision-making has delivered various insights into the nature of human beings when it comes to economically relevant decisions. Those research developments, which build a bridge to social psychology (and recently neuroscience), have led to new and strong pillars in the economics and finance disciplines as well as strategic and international management (e.g., Chittoor et al. 2019;Du et al. 2019;Fitzsimmons et al. 2017).
Second, with regard to the specific role of cultural distance, scholars have long acknowledged that the difficulties and challenges of cultural distance arise not only from the lack of knowledge of how a culturally distant host country functions. They also arise from the perceived foreignness that alienates individuals in home and host country from one another (e.g., Beugelsdijk et al. 2018b;Tihanyi et al. 2005). Especially the latter aspect may exercise socio-psychological influences on individual decision-making that occur independently of economic considerations.
Building on these insights from prior research, we argue that increasing cultural distance may aggravate problems associated with social identification, which will lead MNC decision-makers to feel a lesser social connection and personal attraction to employees in culturally distant foreign subsidiaries. We base this argument on insights from social identity theory (Tajfel 1982;Tajfel and Turner 1986) and selfcategorization theory (Turner 1987) which build on each other and together suggest that individuals define and differentiate themselves from others on the basis of observable-and hence often demographic-characteristics. Based on this assessment, individuals then construct in-groups of similar others, whom they feel more attracted to and out-groups of dissimilar others, vis-à-vis whom they feel more alienated and greater personal distance. Social identity theory extends this categorization and pro-poses that via the construction of in-groups, individuals aim to maintain a positive self-image by comparing their own in-group favorably to the out-group. Although previous studies on decision-making in foreign downsizing have not explicitly made a connection to social identity theory, the idea of forming in-and out-groups according to foreignness relates to Perlmutter's (1969) concept of ethnocentrism, a mindset that elevates home nationals of the MNC and makes the MNC home country the reference point against which managers should make all crucial decisions (Bohas et al. 2021).
We argue that this social identity-related effect of cultural distance influences how MNC decision-makers interpret a foreign subsidiary's performance results: In culturally more proximate subsidiaries, decision-makers will have a more favorable attitude toward performance outcomes. In culturally more distant countries, on the other hand, a lack of social identification with the local workforce will mean that performance outcomes may be judged more harshly, leading to a higher amount of dismissals. In investigating this interplay of subsidiary performance and cultural distance as determinants of dismissal intensity, we draw on the tenets of social identity theory (Tajfel 1982;Tajfel and Turner 1986) and self-categorization theory (Turner 1987) which both suggest that the process of perceiving and categorizing oneself and others into distinct social groups must be triggered by the recognition and salience of a particular dimension of similarity and therefore the likelihood that individuals will use this dimension as a basis for forming in-and out-groups.
With regard to international downsizing decisions, we argue that when subsidiary performance and cultural distance interact, such a trigger occurs and hence, decision-makers will use cultural distance to form in-and out-groups. This interpretation builds on findings of Maddox and Chase (2004), who suggest that situational factors, such as an immediate context, which supports a particular type of categorization, increase the salience of a similarity dimension. For example, particularly a foreign subsidiary's poor performance may accentuate a categorization process when the performance shortcomings are interpreted by decision-makers at firm headquarters as a threat to the prosperity of their individual in-group (i.e., the performance and survival of the MNC). Following social identity theory, individuals will react to such a perceived threat with various perceptual, affective, or behavioral responses in an effort to protect and sustain their in-group vis-à-vis the threatening out-group (Sammarra et al. 2021). We build this argument on Goldberg et al. (2010) who show that the two main motives of social identification-uncertainty reduction and status enhancement-moderate the relationship between perceived demographic dissimilarity and group outcomes. In other words, the more a particular social category allows individuals to reduce uncertainty in their identity or to increase the status of their in-group, the more likely individuals are to use that demographic dimension to form social categories. Therefore, we argue that when cultural distance is high and therefore perceived social identification is low, decision-makers will judge performance outcomes more harshly and be more likely to downsize a foreign subsidiary than they would in a culturally more proximate foreign subsidiary. With this behavior, decision-makers protect their in-group at the expense of the out-group.
On the other hand, when a foreign subsidiary is culturally more proximate and MNC decision-makers therefore feel more socially identified with the subsidiary's K employees, in-group favoritism may lead decision-makers to judge subsidiary performance more favorably. This interpretation reflects previous studies which use social identity theory to explain why transgressions, such as, for example, unethical behavior, are punished less severely when they are committed by members of an in-group as compared to when they are committed by out-group individuals (e.g., Ellemers et al. 1997;Kundro and Nurmohamed 2021;Van Vugt and Hart 2004). By punishing in-group transgressors less severely, in-group decision-makers can protect the positive image of their own in-group by signaling that the transgressors' behavior was not highly problematic. This perspective is in line with other studies showing that in-group decision-makers are more likely to assume less harmful motives for in-group transgressors as compared to out-group transgressors (Brewer 1999;Weidman et al. 2020). With regard to foreign subsidiaries, such an effect may translate into assuming that good performance of close subsidiaries is valued higher than the same performance level in more distant subsidiaries.
We summarize our arguments in our main Hypothesis: Hypothesis: Cultural distance moderates the effect of subsidiary performance on the number of dismissals in a firm's foreign subsidiary: The lower the cultural distance, the stronger the negative effect of subsidiary performance on the number of dismissals in a firm's foreign subsidiary.

Data and Sample
The empirical analysis draws on subsidiary-level data of German multinational firms maintained by the Central Bank of Germany. Firms and individuals that are based in Germany are required to report their foreign direct investment above a balance sheet total greater than C 3 million to the Central Bank.

Dependent Variable
The dependent variable amount of dismissal measures how many employees are dismissed by subtracting the number of employees in a foreign subsidiary in the previous year from the number in the current year. This measure is more nuanced compared to previous research that very often measures divestment as a binary variable only (e.g., Dai et al. 2013;Pattnaik and Lee 2014).

Independent and Control Variables
There continues to be a lively debate on which is the most appropriate measure for cultural distance (Beugelsdijk et al. 2018a) and research constantly develops more advanced measures of cultural distance (e.g., Shenkar 2012). However, the aim of this study is not to deliver a refined measurement of the construct. Therefore, we chose to use a broadly accepted and conservative measure of cultural distance in order to make our results comparable to previous empirical findings. We first decided on the concept used for cultural distance. Notwithstanding justified criticism, the cultural measure developed by Hofstede (1980) is among the most accepted and used in international business research (Beugelsdijk et al. 2018a). Therefore, we favor this concept vis-á-vis alternative concepts such as that of Schwartz (1994) or GLOBE (House et al. 2004). Second, we selected a specific measurement. The index developed by Kogut and Singh (1988) is the most established measure (Kirkman et al. 2006;Maseland et al. 2018) and the "must have" variable in international business and management research (Shenkar et al. 2008, p. 908). It represents a Euclidean distance measure with variance correction. In order to ease the interpretation of our results, we invert cultural distance, operationalized with the Kogut-Singh-Index, to cultural proximity by subtracting each subsidiary' value of cultural distance from the maximum in the sample. In addition, we perform robustness checks (see below) using the index GLOBE (House et al. 2004) as an alternative measure of cultural distance.
To measure subsidiary performance we draw on the return on equity (roe) generated in a subsidiary and year (e.g., Delios and Beamish 2001;Gupta et al. 2020;Uhlenbruck et al. 2017). This operationalization captures the actual profitability of a subsidiary rather than forward-looking expectations of the overall corporation, which are commonly in the focus of market-based performance measures. To make sure that this measure does not cause bias, we performed robustness checks using return on assets instead of return on equity (see below). Another figure that captures the subsidiary's success is the development of sales (Nobeoka and Cusumano 1997). Hence, we include the change in subsidiary sales calculated as the difference between sales in the previous year and the current year divided by the sales in the previous year.
Further, we include a variety of variables that were identified as determinants of subsidiary downsizing in previous studies. We include the subsidiary equity share because joint ventures seem to induce more coordination problems and show a higher propensity to be divested than wholly-owned subsidiaries (Li 1995). With rising age, subsidiaries are less likely to be divested (Belderbos and Zou 2009); hence, we include subsidiary age (number of years since establishment). As unrelated subsidiaries are more difficult to manage for the parent (Duhaime and Grant 1984), we include the dummy variable subsidiary unrelatedness that indicates whether the subsidiary's activities represent a diversification in relation to the parent firm ("1" if yes, "0" otherwise). Finally, we control for subsidiary size as the subsidiary's log annual turnover and the subsidiary's function as manufacturing subsidiary ("1" if yes, "0" otherwise) and service subsidiary ("1" if yes, "0" otherwise).
Foreign affiliates might face downsizing or termination also due to strategic considerations of the parent company (Benito and Welch 1997;Boddewyn 1979). We include the variable network reconfiguration, which captures whether the parent firm is increasing or decreasing its staff in the overall international network (number of foreign employees in current year minus number of foreign employees in previous year relative to the number of foreign employees in previous year). On the parent level, we further capture firm size (Belderbos and Zou 2009) measured as (log of) total sales and foreign network performance, measured as return on sales across a parent firm's foreign subsidiaries.
Referring to host country characteristics, favorable market conditions such as economic growth reduce divestment probabilities (Benito 1997). We therefore include GDP growth (source: World Bank) and market size (log GDP, source: World Bank) to control for the attractiveness of the local market. Finally, we include industry and time dummies in all models. Table 1; the pairwise correlations are presented in Table 2. Interestingly, there is nearly no correlation (r = 0.004) between subsidiary performance and cultural proximity, which is remarkable, given the generally expected influence of cultural distance on performance. It seems that cultural distance does not lead per se to lower performance and, hence, dismissal. Obviously, it is a much more complex relationship that is affected by a number of influences (e.g., the cultural background and experience of individual managers, the nature of the product or service offered by the firm, and so on). Further, subsidiary performance (measured as roe) and change in subsidiary sales have nearly no correlation (r = -0.001). While both variables measure an aspect of firm performance, their individual meaning seems very different. Change in subsidiary sales reflects the consumers' valuation of the products and services offered by the firm. It can vary strongly from year to year, depending also on factors that are beyond the firm's influence (e.g., macroeconomic effects). Return on equity reflects the efficiency of money invested. Certainly, the firm has more flexibility to optimize the roe, for example by cutting costs or (in the short run) by capital outflow. The variance inflation factors are close to one (mean: 1.19; maximum: 1.39) indicating that no problems with multicollinearity exist.

Summary statistics and variable measurement are displayed in
Regression results are displayed in Table 3. As a Breusch-Pagan/Cook-Weisberg test reveals, the data are heteroscedastic. We therefore employ an ordinary least squares (OLS) estimation with robust standard errors. Model 1 is the base model, containing just the controls. In Model 2, we introduce the simple effects of cultural proximity and subsidiary performance. Model 3 is the full model, in which we interact cultural proximity and subsidiary performance.
In Model 2, the coefficient of subsidiary performance is negative but slightly misses significance (β = -0.003; p = 0.102). In Model 3, however, the coefficient of subsidiary performance is negative and significant (β = -0.139; p = 0.000). To draw a clear conclusion whether we find support for our baseline assumption that subsidiary performance has a negative influence on the amount of subsidiary dismissal, we have to exercise a marginal effects analysis (see below). To test our main Hypothesis, we interact the variables cultural proximity and subsidiary performance (Model 3). The interaction is negative and significant (β = -0.258; p = 0.000). With   an R squared of 6%, the explained variance is low. The most important reason behind this seems to be the heterogeneity of firms in this sample. Every firm has very specific endowments with management culture, organizational architectures, decision routines, firm values (and so on), which affect the severity of downsizing. Our empirical model does not control for such heterogeneities, hence, the overall influence of cultural distance and performance is low. Since our goal was not to maximize the explanative power of the full model but to carve out the joint effect of cultural proximity and subsidiary performance, we do not think that a small explained variance is a severe problem.
To interpret the moderating effect of cultural proximity, we follow a recent approach introduced by Busenbark and colleagues (2022) that has been implemented, for example, by Steinberg, Asad, and Lijzenga (2022). We visualize the marginal effects of subsidiary performance on amount of dismissal over different values of cultural proximity (ranging from the fifth to the 95th percentile). Figure 1 shows that the relationship between subsidiary performance on amount of dismissal is negative for all values of cultural proximity. Hence, we can conclude that our baseline hypothesis is supported. Further, we see that the relationship becomes stronger for higher values of cultural proximity, which delivers additional support for our main hypothesis. The higher the cultural distance between home and host country, the more employees are dismissed in that foreign subsidiary, compared to culturally more proximate foreign subsidiary at the same performance level.
To substantiate our findings, we perform a number of robustness checks. Results are displayed in Table 4. In Model 1, we use an alternative measure of subsidiary performance, which is return on assets (e.g., Coombs and Gilley 2005) instead of return on equity. The hypothesized effect of the interplay of subsidiary performance and cultural proximity is negative and significant (β = -0.008; p = 0.005) as in our main model. In Model 2, we use another alternative measure of subsidiary performance, which aims at capturing changes in performance difference instead of absolute values (e.g., Fredrickson et al. 1988). We measure the change in return on equity as the difference between return on equity in the current year and the previ-     ous year. The interaction effect of subsidiary performance and cultural proximity is negative, however, not significant (β = -0.010; p = 0.172). In Model 3, we combine the previous two alternative measures and operationalize subsidiary performance as change in return on assets. Here, the interaction effect is, again, negative and significant (β = -0.008; p = 0.038).
Besides varying the measurement of performance, we also test a different measure of cultural proximity. In Model 4 (Table 4), we use the cultural values proposed by the GLOBE project (House et al. 2004) instead of Hofstede (1980). With this alternative measure, the interaction effect of cultural proximity and subsidiary performance (measured with return on assets) is negative and significant (β = -0.004; p = 0.004). In sum, the robustness checks, which contain alternative measures of subsidiary performance and cultural proximity, largely support the results of our main analysis.

Discussion
The goal of this study was to investigate the interplay of cultural distance and subsidiary performance in international downsizing decisions and therefore contribute to knowledge on the determinants of de-internationalization. Results of our analyses confirm previous research suggesting that subsidiary performance is an important driver of downsizing decisions. Further, we show that cultural distance influences how MNC decision-makers interpret the efficiency of the subsidiary in terms of its performance. The lower the cultural distance between home and host country, the stronger is the dismissal-preventing effect of subsidiary performance. That means performance will be interpreted more favorably in more proximate subsidiaries and more unfavorably in subsidiaries that are more distant. These results have important implications for research and management practice.
Our study extends research on multinational divestment and de-internationalization that highlights the coordination costs associated with cultural distance. Most prior studies have presented strong theoretical arguments why cultural distance between home and host country will increase coordination costs for the MNC and, therefore, lead to lower efficiency and a higher tendency for downsizing (Beugelsdijk et al. 2018b;Johanson and Vahlne 1977;Kostova et al. 2016;Li and Guisinger 1992). However, recent results have presented less conclusive evidence of this relationship, often showing insignificant effects (e.g., Coudounaris et al. 2020;Schmid and Morschett 2020). Our findings are in line with these studies. We suggest that-instead of influencing subsidiary performance directly-the effect of cultural distance comes to bear more indirectly through behavioral considerations. We argue that greater differences in cultural terms may lead MNC decision-makers to feel a lesser personal attraction, social identification, and familiarity to employees in those foreign subsidiaries and, hence, will interpret subsidiary performance based on this behavioral effect.
The results of our analysis also provide interesting insights on the interaction of cultural distance with subsidiary performance and its consequences for de-internationalization. We find that high subsidiary performance will prevent dismissal in particular in culturally proximate countries. Theories of social identity (Tajfel 1982;Tajfel and Turner 1986) and self-categorization (Turner 1987) support our suggestion that decision-makers are triggered into using cultural distance to form social categories, alienating more culturally distant subsidiaries and consequentially increasing the number of dismissals there. We believe that these results can encour-age further research on the relationship between cultural distance and the subsidiary (or MNC) performance.
The results of our study bear important managerial implications. Decision-makers in MNCs should be aware of potential behavioral biases when making multinational downsizing decisions. Downsizing and layoff decisions are very critical and might raise the attention not only of the employees who are directly affected by them. Therefore, it is especially important to present transparent and comprehensible decision determinants. Moreover, it is in the interest of long-term MNC performance that only those subsidiaries are downsized that are performing least favorably or that have the lowest strategic value for the MNC. Managers might not be aware that they have more or less empathy for different cultural groups. One measure MNCs might use is to establish cultural diversity within the group of people that makes these strategic decisions. Different cultural backgrounds and diverse international (work) experience may diminish any potential dominance of one cultural group.
Finally, we acknowledge that the study has several limitations. Most importantly, the empirical analysis is based on a completely anonymous dataset. A more granular knowledge, especially regarding the composition of executive teams and their cultural backgrounds, would allow us to gain a deeper understanding of the psychological factors in decision-making processes. This is not possible with the present data but could help to uncover why certain preferences and attitudes towards employees from different cultural regions may dominate other decision determinants. Future research might draw on experiments or surveys to deliver a more nuanced understanding of the multinational downsizing decision process.
Second, while our use of a more fine-grained variable-the number of dismissals in a firm's foreign subsidiary-extends prior research which has very often measured divestment as a binary variable only (e.g., Dai et al. 2013;Pattnaik and Lee 2014), it also introduces the question of potentially different strategic considerations behind dismissals and divestment. When a firm fully divests a foreign subsidiary, this foreign involvement is terminated. Dismissals in foreign subsidiaries, however, may strategically, in the course of organizational decline, either be intended as a stepwise process towards full divestment or-alternatively-be aimed at bringing the subsidiary back on track. Since our data does not allow for any insights into the strategic considerations behind the foreign downsizing decisions, we cannot exclude this possibility. At the same time, due to a number of reasons, we believe that our results are not biased by this fact. First, prior research on retrenchment has shown that layoffs do not contribute to turnaround success (Tangpong et al. 2015), and hence, they would not be likely to ultimately prevent the full divestment of the foreign subsidiary. Furthermore, by measuring the number of dismissals in a firm's foreign subsidiary, we include the full range of downsizing in foreign subsidiaries from only a few employees to nearly all employees of that subsidiary being dismissed. In addition, we control for several variables (e.g., change in subsidiary sales), which may influence a firm's strategic thinking on whether a foreign subsidiary could be brought back on track or not. Finally, from a more theoretical point of view, we believe that against the background of our research question-which argues that international downsizing decisions are influenced by considerations that are not based on financial and/or strategic aspects-the contribution of our study is not diminished by the possibility that strategic considerations behind firm's downsizing decisions may differ. An interesting avenue for future research could, however, be to specifically investigate the antecedents and consequences of decision-making on foreign dismissals more deeply.
Third, it would be valuable to include the home country into the considerations of multinational dismissal decisions. Presumably, a firm's home country will be less affected by dismissal due to decision makers' emotional attachment rather than efficiency consideration. As we lack performance data of the parent company, we cannot analyze this question in the paper at hand. Finally, it might be valuable to expand the investigation to different dimensions of distance. Previous literature has highlighted distance concepts ranging from geographic distance, economic distance to institutional distance and several more (e.g., Ambos and Håkanson 2014). It would be interesting to see whether there are similar effects when it comes to alternative concepts of distance.

Conclusion
In this study, we challenge the established, yet bounded, perspective that investors base their de-internationalizing downsizing decisions only on financial and/or strategic considerations. From this viewpoint, cultural distance between home and host country is generally expected to increase coordination costs and, therefore, lead to higher downsizing propensities. Instead, we ask whether cultural distance has effects on international layoff processes that go beyond the efficiency-based considerations demonstrated by prior literature. Specifically, we base our arguments on the insights from social identity (Tajfel 1982;Tajfel and Turner 1986) and self-categorization (Turner 1987) theory and argue that decision-makers are more willing to dismiss employees whom they perceive to be culturally different to themselves. Results show that this behavioral tendency influences how MNC decision-makers assess subsidiary performance: The higher the cultural proximity, the more positive decision-makers will interpret performance outcomes. Our results therefore contribute to knowledge on the determinants of de-internationalization and open up multiple avenues for future research to investigate more closely how exactly behavioral factors influence decision-making in foreign divestments.
Funding Open Access funding enabled and organized by Schmalenbach-Gesellschaft and German Academic Association for Business Research.

Conflict of interest M. Flickinger and M. Zschoche declare that they have no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4. 0/.