Printer Friendly

FDI experience location and subsidiary mortality differences in national culture and the expansion of Korean MNEs.

Abstract:

Purpose: This study examines the relationship between an MNE's experience with foreign direct investments (FDI) and subsidiary mortality when considering learning limitations that an MNE are susceptible to under the influence of national cultural differences.

Methods: We apply event history analysis on a sample of subsidiaries of Korean MNEs during 1990-2006.

Results: We find that subsidiary mortality has a negative relationship with FDI experience from countries culturally similar to both the MNE's home country and the focal host country: an inverse U-shaped relationship with both experience from countries culturally similar to the home country but different from the focal host country, and experience from countries culturally different from the home country but similar to the local host country. Experience from countries culturally dissimilar to both the local host country and the home country has a negative effect when a subsidiary is located in countries dissimilar to the home country, and an insignificant effect when a subsidiary is located in countries similar to the home country.

Conclusions: Our findings show that prior experience has different effects depending on where it is accumulated and that MNEs are susceptible to limitations both when drawing inferences from their FDI experiences and when applying these experiences toward subsequent FDIs.

Keywords: FDI * Cultural differences. Experience * Subsidiary mortality * Erroneous inferences * Experience misapplication

Introduction

Prior international experience has long been considered a critical factor in an multinational enterprise's (MNE) success as it embarks into new foreign direct investments (FDI) (Johanson and Vahlne 1977). Yet the effect of prior FDI experience on subsequent FDI performance remains unclear (Barkema and Schijven 2008). Decades of research have yielded inconsistent results: One group of studies has found a beneficial effect of an MNE's prior FDI experience on the performance of subsequent FDIs (e.g., Delios and Beamish 2001: Fang et al. 2007), while another group has found a detrimental effect (e.g., Barkema et al. 1996: Gaur and Lu 2007: Makino and Delios 1996). A third group has failed to confirm a consistent or significant relationship between FDI experience and subsequent FDI performance (e.g., Barkema et al. 1997; Li 1995).

We intend to improve our understanding of the experience-performance relationship in FDI by systematically investigating the conditions where a beneficial or a detrimental effect of FDI experience may occur. According to the experiential learning perspective, prior experience enables a firm to learn and perform better when conducting similar activities in the future (Cyert and March 1963: Levitt and March 1988). Evidence from manufacturing settings has confirmed the presence of a learning curve and a beneficial effect of experience (Yelle 1979). However, organizational learning scholars have also long acknowledged that experience may not always be beneficial because firms may draw erroneous inferences from experience and misapply experience to dissimilar contexts (Feldman 1986: Huber 1991: Levinthal and March 1993; Levitt and March 1988). The prevalence and detrimental consequences of these two types of learning limitations (i.e., drawing erroneous inferences from and misapplying experience) have been evidenced in strategic settings such as acquisitions and strategic alliances (Finkelstein and Haleblian 2002: Haleblian and Finkelstein 1999: Heimeriks 2010: Zollo 2009).

Effectively learning from and applying experience is inherently difficult in FDI because the MNE needs to learn within a foreign cultural context, and experience gained from one culture is often subsequently applied in another. National cultural differences may compromise both an MNE's ability to learn from its prior FDI operations (Bhagat et al. 2002: Bjorkman et al. 2007: Uhlenbruck 2004) and its ability to appropriately apply the lessons learned toward subsequent investments (Barkema et al. 1996: Nadolska and Barkema 2007). Given these potentially hindering effects of cultural differences, it is possible that an MNE may draw erroneous inferences from experience as well as misapply experience to subsequent FDIs. Recently, scholars have begun paying attention to leaning limitations in FDI and the consequently detrimental effect of FDI experience (Kim et al. 2010: Nadolska and Barkema 2007). However, these studies have mainly focused on an MNE's misapplication of experience, leaving the possibility that an MNE may draw erroneous inferences from prior experience unexamined.

In order to provide a systematic investigation of the effect of experience by considering the possibilities of both aforementioned learning limitations, when examining the mortality of a subsequent FDI (hereafter "focal subsequent FDI"), we group the MNE's prior FDI experience into subgroups based on two types of national cultural differences that may increase the likelihood of those learning limitations. The first is the difference between the MNE's home country and the host countries of its previous FDIs; this cultural difference erodes the MNE's ability to effectively draw inferences from previous FDIs (Barkema and Drogendijk 2007; Johanson and Vahlne 1977; Uhlenbruck 2004) and may therefore increase the likelihood of drawing erroneous inferences. The second is the difference between the host countries of the MNE's previous FDIs and the host country of the focal subsequent FDI; this cultural difference causes dissimilarities between the previous and focal subsequent FDI, increasing the likelihood of experience misapplication (Nadolska and Barkema 2007; Thomas et al. 2007).

Grouping MNEs' previous FDls based on these two types of cultural differences results in four distinctive experience types that may have different effects on the focal subsequent FDI: (1) Double-proximate experience, referring to previous FDIs in countries that are culturally similar to both the MNE's home country and the host country of the focal subsequent FDI; (2) home-proximate experience, referring to previous FDIs in countries that are culturally similar to the MNE's home country but different from the host country of the focal subsequent FDI; (3) host-proximate experience, referring to previous FDIs in countries that are culturally different from the MNE's home country but similar to the host country of the focal subsequent FDI; and (4) double-distant experience, referring to previous FDls in countries that are culturally different from both the MNE's home country and the host country of the focal subsequent FDI. We examine the effect of these four types of experience on the mortality of MNEs' subsequent FDIs using a longitudinal dataset of FDIs conducted by South Korean MNEs between 1990 and 2006.

This study enriches the literature on FDI experience by addressing two research gaps that may have contributed to the inconsistent empirical findings discussed earlier. First, prior research has by-and-large assumed a linear and positive relationship between FDI experience and subsequent FDI performance; and the causes of possibly detrimental effects of experience have therefore not been well explored. We address this gap by considering learning limitations such as drawing erroneous inferences and misapplying experience that may result in a detrimental experience effect. Second, previous studies tended to either assume that FDI experiences are homogenous (e.g., Fang et al. 2007) or focus on experience specific to a focal host environment (e.g., Gaur and Lu 2007; Delios and Beamish 2001). We advance this literature by distinguishing four types of FDI experience defined based on two types of cultural differences that may increase the likelihood of the aforementioned learning limitations.

Theory Development

Limitations in Learning from Experience: Background

The commonly accepted positive relationship between experience and performance is based on two premises: First, that firms will draw accurate inferences and gain valuable knowledge from prior experience, and second, that firms will discern the relevance of this experience and apply it appropriately toward subsequent activities. However, the organizational learning literature has indicated that these premises are often violated because firms are susceptible to learning limitations (Feldman 1986; Fiol and Lyles 1985; Huber 1991; Levinthal and March 1993; Levitt and March 1988).

The first type of limitation is related to a firm's drawing of inferences from prior experience. Levitt and March (1988, p. 325) noted that learning is "superstitious" when, "the subjective experience of learning is compelling, but the connections between actions and outcomes are mis-specified." Underlying this misspecification of causalities is the ambiguous nature of experience--what has happened and the causalities involved are not clear (Levinthal and March 1993: Levitt and March 1988). Zollo (2009) suggested that outcome ambiguity, defined as the uncertainty surrounding the assessment of a decision or task outcome, could also result in the mis-contribution of outcomes and lead to erroneous inferences concerning causalities. Alongside the problems caused by the ambiguous nature of experience, there are also individual- and organizational-level constraints that hinder effective learning. At the individual level, the processing of information and interpretation of experience are constrained by individual cognitive and rational limitations (Feldman 1986: Levitt and March 1988). At the organizational level, the frames that a firm uses to interpret experience and draw inferences are susceptible to constraints such as organizational history and politics (Levinthal and March 1993; Levitt and March 1988). These inherent difficulties in learning from experience may cause prior experience to generate erroneous references and mislead the firm in its subsequent similar activities (Huber 1991 : Levitt and March 1988).

The second type of limitation is related to a firm's application of its experience in subsequent activities. Correct inferences drawn from the firm's prior experience are only valuable for subsequent activities to the extent that the previous and subsequent activities are highly similar. However. firms are susceptible to misapplying experience toward dissimilar subsequent activities, resulting in the so-called "negative transfer effect" that harms subsequent activities (Finkelstein and Haleblian 2002). There are two major mechanisms underlying this experience misapplication phenomenon. The first is that managers' cognitive and rational limitations constrain their abilities to detect fundamental differences between prior and subsequent activities. Haleblian and Finkelstein (1999) noted that similar activity types can be highly heterogeneous (e.g., related and unrelated acquisitions): managers can be confused by those surface similarities and consequently inappropriately generalize prior experience toward subsequent activities that are substantially different. The second mechanism is that managers may generalize experience without deliberately assessing its applicability with regard to the focal circumstances. This may occur when experience enhances managers' confidence in their competence at a focal activity more than it enhances their actual competence (Zollo 2009), or when experience enhances managers" familiarity with certain routines and practices that are not directly applicable to subsequent endeavors (Levinthal and March 1993; Levitt and March 1988).

Figure 1 summarizes the possible effects of experience on a subsequent activity when considering the aforementioned two types of learning limitations. As shown in the top section of the figure, if experience generates correct inferences then the effect of experience on a subsequent activity will be beneficial when applied to similar activities (path from Box 1 to 3), detrimental if applied to a dissimilar activity or a different circumstance (path from Box I to 5), or neutral if not applied (path from Box 1 to 8). However, as indicated in the bottom section of the figure, if experience primarily generates erroneous inferences then the effect of experience can only be detrimental if applied (path from Box 6 to 10) or neutral if not applied (path from Box 6 to 8).

[FIGURE 1 OMITTED]

National Cultural Differences and Limitations in Learning from FDI Experience

National cultural differences have been identified as the major source of barriers deterring learning and knowledge transfer in FDI (Barkema et al. 1996; Bhagat et al. 2002; Bjorkman et al. 2007; Nadolska and Barkema 2007; Uhlenbruck 2004). By shaping individuals' beliefs, perceptions, and behaviors, national culture defines the knowledge relevant to conducting business within a country (Kirkman et al. 2006). Culture also permeates companies' structures, business practices, and management styles (Lubatkin et al. 1998), as well as underlies a country's rules, regulations, and institutions (Barkema and Schijven 2008). As a result, national cultural differences challenge both the MNE's ability to learn from its prior FDI experience and its ability to appropriately apply the knowledge gained from that experience (Barkema and Drogendijk 2007; Uhlenbruck 2004).

Specifically, two types of national cultural differences are pertinent. The first is the difference between the MNE's home country and the host countries it had previously invested in (hereafter "previous-FDI-home-country cultural difference"), and the second is the difference between the host countries of previous investments and the host country of a focal subsequent FDI (hereafter "previous-FDI-subsequent-FDI cultural difference").

Previous-FDI-Home-Country Cultural Difference and Drawing Erroneous Interferences

FDI is a type of strategic activity that involves significant causal and outcome ambiguity, increasing the difficulty of learning and the likelihood of drawing erroneous inferences (Nadolska and Barkema 2007: O'Grady and Lane 1996: Petersen et al. 2008; Thomas et al. 2007). This is particularly true when an MNE's previous FDIs are located in host cultures that are significantly dissimilar to the MNE's home country culture.

As Cohen and Levinthal (1990, p. 128) posited, it is a firm's prior related knowledge that confers its absorptive capacity, defined as the "ability to recognize the value of new information, assimilate it, and apply it to commercial ends." Because the knowledge an MNE may learn from its previous FDIs is embedded within the host cultures while the MNE's prior knowledge base is largely imprinted by its home country culture, previous-FDI-home-country cultural differences limit the MNE's absorptive capacity and erode its learning ability (Uhlenbruck 2004). When previous-FDI-home-country cultural differences are significant the MNE may be unable to effectively detect and comprehend the information it receives from its previous FDIs. This lack of absorptive capacity will constrain the MNE from accurately deciphering the causalities involved in previous FDIs. Consequently, prior FDI experience may not always generate useful knowledge; in contrast, it may result in erroneous inferences that can mislead the MNE in its future investments. Furthermore, significant previous-FDI-home-country cultural differences accentuate individual cognitive and rational limitations. For example, managers who have developed managerial knowledge and mindsets under the MNE's home culture may find it difficult to understand how things are done in a new culture, making them susceptible to FDI decision errors.

The low learning ability of MNEs in culturally dissimilar countries and subsequent adverse consequences have been documented in previous studies. For example, Uhlenbruck (2004) found that MNEs' capabilities to learn from their experiences in Central and Eastern Europe (CEE) decreased as the cultural distance between the MNE's home country and host country increased. Barkema and Drogendijk (2007) found that the subsequent FDIs of Dutch MNEs, particularly those without sufficient international experience in the CEE region (which is culturally dissimilar to the Netherlands), did not benefit from the MNEs' previous operations in CEE countries.

Previous-FDI-Subsequent-FDI Cultural Difference and Misapplying Experience

Generalizing prior FDI experience is challenging because there may be significant heterogeneity across FDI operations located in different countries. Although each FDI may teach an MNE general lessons about operating in foreign cultures (Barkema and Schijven 2008), the unique context features of different host environments may make some of the knowledge an MNE gained from its previous FDIs (e.g., knowledge about how to interact with local suppliers) a location-bounded asset that can only be applied to the same or a similar host environment (Johanson and Vahlne 1977, 1990; Rugman and Verbeke 1992). Consequently, experience may only benefit a subsequent FDI when managers can correctly assess the similarities and differences between the prior and subsequent FDIs, and then appropriately generalize or discriminate (Haleblian and Finkelstein 1999).

However, this assessment could be challenging because differences between FDIs may be subtle and two FDIs can have high surface similarities (e.g., similar functional areas or the same type of customers), yet substantial underlying differences (e.g., how human resources are managed). These underlying differences make it difficult for mangers to apply previous FDI experiences via analogical thinking (Gavetti et al. 2005). Managers may therefore not be able to recognize these underlying differences and take the applicability of prior experience for granted based on surface similarities (Barkema and Drogendijk 2007; O'Grady and Lane 1996). Additionally, cultural differences also make it difficult to accurately evaluate the outcomes of FDIs. This outcome ambiguity may make managers become overconfident in their abilities to conduct FDIs and apply their experiences without diligently assessing the differing circumstances in subsequent FDIs (Petersen et al. 2008; Zollo 2009).

The underlying differences between previous FDIs and the subsequent focal FDI may be salient when there are significant previous-FDI-subsequent-FDl cultural differences. These differences decrease the relevance of the knowledge gained from previous FDIs. At the same time, the level of uncertainties and outcome ambiguity faced by an MNE also increases when a subsequent FDI's host culture is significantly different from those the MNE had previously conducted business within. In this context, misapplying experience is very likely to occur (Barkema 2006; Haleblian and Finkelstein 1999). The literature has provided evidence for this experience misapplication in FDI. For example, Barkema (2006) noted that Dutch retailer Ahold applied its United States acquisition experience toward acquisitions in Latin America and Asia, yielding disappointing results. Nadolska and Barkema's (2007) study of 25 Dutch MNEs showed that MNEs tend to generalize prior acquisition experience toward subsequent acquisitions in different cultures, also without successful results.

Grouping MNE Prior FDI Experience

Since the two types of cultural differences discussed above influence the likelihood that the aforementioned two types of learning limitations may occur, it is necessary to consider them simultaneously when examining the effect of FDI experience. One way to achieve this is grouping an MNE's previous FDIs based on the levels of these two types of cultural differences.' As illustrated in Fig.2, if we divide the two types of cultural differences into a low and high level respectively there will be four possible combinations represented by the four quadrants in the figure. However, the combinations that are practically possible when grouping an MNE's previous FDIs depend on the location of a focal subsequent FDI relative to the MNE's home country. For example, assume an MNE is originally from South Korea. For a subsequent FDI initiated in a country culturally similar to South Korea (e.g., China), the MNE's previous FDIs can be categorized either in Quadrant l (e.g., previous FDIs in Vietnam, which is similar to both China and South Korea) or Quadrant 3 (e.g., previous FDIs in the U.S., which is dissimilar to both China and South Korea). For a subsequent FDI initiated in a country culturally dissimilar to South Korea (e.g., the U.S.) the MNE's previous FDIs can be categorized in Quadrant 2 (e.g., previous FDIs in China, which is similar to South Korea but dissimilar to the U.S.), Quadrant 3 (e.g., previous FDIs in Germany, which is dissimilar to both South Korea and the U.S.), or Quadrant 4 (e.g., previous FDIs in Canada, which is similar to the U.S. but dissimilar to South Korea).

We term previous FDIs in Quadrant I double-proximate experience since they are located in countries culturally proximate to both the MNE's home country and the host country of a focal subsequent FDI, previous FDIs in Quadrant 2 home-proximate experience since they are located in countries culturally proximate to the MNE's home country, previous FDIs in Quadrant 3 host-proximate experience since they are located in countries culturally proximate to the host country of a focal subsequent FD1, and previous FDIs in Quadrant 4 double-distant experience since they are located in countries approximate to neither the MNE's home country nor the host country of a focal subsequent FDI.

In the next section we develop hypotheses regarding the effect of an MNE's level of each of these four types of experience on the mortality of subsequent FDIs.

Hypotheses

The benefits that could result from the accumulation of FDIs have been investigated at both the MNE and subsidiary levels. A major benefit is associated with an MNE's learning from its prior FDIs by generating experiential knowledge that could potentially enhance both the MNE's overall performance and subsequent FDls' performance (e.g., Barkema and Drogendijk 2007; Deliois and Beamish 2001; Lu and Beamish 2004). In contrast, the possible downsides of the accumulation of previous FDIs, in particular at the subsidiary level, have received limited attention. Those identified in the literature include the transaction and coordination costs associated with international diversification and possible diseconomies of agglomeration, both of which are particularly salient at a high level of FDI accumulation and more pertinent when examining parent-level (i.e., an MNE) performance and behavior (Arregle et al. 2009; Lu and Beamish 2001, 2004). Below, we investigate the effects of the accumulation of the previously defined four types of experience on subsidiary level performance as gauged by subsidiary mortality by considering an MNE's likely susceptibility to learning limitations under the influence of cultural differences.

Double-Proximate Experience

Double-proximate experience refers to an MNE's previous FDIs in countries that are culturally similar to both the MNE's home country and the host country of a focal subsequent FDI (i.e., the previous-FDI-home-country cultural difference and previous-FDI-subsequent-FDl cultural difference are both small). As shown in Fig. 2, this type of experience is only possible when a focal FDI is located in a country that is culturally similar to the MNE's home country (e.g., for a Korean MNE initiating an FDI in China this experience refers to its previous FDIs in Vietnam, Singapore, etc.).

An MNE's ability to learn from this experience should be high since the previous-FDI-home-country cultural difference is small. Consequently, the likelihood that the MNE may draw erroneous inferences from this experience is low. Furthermore, this experience should be highly relevant to a focal subsequent FDI given the small previous-FD1subsequent-FDI cultural difference, reducing the likelihood of experience misapplication. Since the likelihood of both of the two learning limitations is low, this experience should generate valuable experiential knowledge (e.g., knowledge about certain types of customers) that is highly applicable to a focal subsequent FDI (Barkema et al. 1996; Johanson and Vahlne 1977). Following the path from Box 1 to 3 in Fig. 1, which indicates a beneficial effect of experience when correct inferences are drawn from experience and appropriately applied to similar subsequent activities, we would therefore expect a beneficial effect of double-proximate experience. Specifically, the experiential knowledge generated by double-proximate experience could benefit a focal subsequent FDI in ways such as reducing uncertainties the MNE may face, decreasing the likelihood of making mistakes, and allowing the MNE to better perceive and evaluate business opportunities in the focal host country (Benito and Gripsrud 1992; Erikkson et al. 1997; Johanson and Vahlne 1977). We therefore hypothesize that:

Hypothesis 1: The mortality of a subsequent FDI in a culturally similar country will have a negative relationship with the MNE's experience in countries culturally similar to both the MNE's home country and the focal host country.

Home-Proximate Experience

Home-proximate experience refers to an MNE's previous FDIs in countries that are culturally similar to the MNE's home country but dissimilar to the host country of a focal subsequent FDI (i.e., previous-FDI-home-country cultural difference is small and previous-FDI-subsequent-FDI cultural difference is large). This type of experience is only possible when a focal FDI is located in a country that is culturally dissimilar to the MNE's home country (e.g., for a Korean MNE initiating an FDI in the U.S. this refers to the MNE's prior experience in China, Vietnam, etc.).

The benefits of experience accumulated in countries proximate to the MNE's home country on the MNE's subsequent foreign expansions has been well documented (e.g., Barkema et al. 1997: Johanson and Vahlne 1977). The essence of the argument is that the small cultural difference between the host and home countries allows MNEs to learn how to operate in foreign countries without simultaneously dealing with significant cultural differences. However, while this high learning ability lowers the likelihood that the MNE will draw erroneous inferences from this type of experience, the large previous-FDI-subsequent-FDI cultural difference implies that the MNE is susceptible to inappropriately applying this experience toward a focal subsequent FDI. The underlying differences between previous FDIs and a focal subsequent FDI may be significant due to the different national cultural contexts: however, they may share surface similarities such as similar products, technologies, or entry modes that could make the MNE inappropriately perceive them as similar (Barkema and Drogendijk 2007). The effect of home-proximate experience may follow the path from Box 1 to 5 in Fig. 1 such that correct inferences are drawn from experience, but the firm misconceives dissimilar events as similar and inappropriately applies that experience: this results in a detrimental experience effect.

Ceteris paribus, inappropriately applying home-proximate experience toward a subsequent focal FDI may be very common when an MNE has a low level of this experience type. On the one hand, the initial accumulation of home-proximate experience results in an increasing amount of knowledge that may be irrelevant to a focal subsequent FDI (e.g., knowledge concerning how to manage a subsidiary in a similar foreign culture) but is available for the MNE to draw on. This requires the MNE to deliberately assess the applicability of knowledge interred from its home-proximate experience toward a focal subsequent FDI. However, the MNE may have a strong tendency to generalize its home-proximate experience without sufficient caution and assessment. Since the previous-FDIs-home-country cultural difference is small, the MNE is likely to exploit a significant amount of its home-country-developed knowledge base and encounter few cultural challenges during the operation of FDIs proximate to the MNE's home country. The initial accumulation of home-proximate experience may therefore lead the MNE to develop overconfidence in the generalizability of home country resources and its ability to operate in foreign countries. This overconfidence may increase the MNE's tendency to apply home-proximate experience even when the focal subsequent FDI is in a significantly different culture.

However, once the MNE has accumulated a certain level of home-proximate experience the likelihood of misapplying this experience should decrease. Haleblian and Finkelstein (1999) noted that compared to novices, experts (those with more experiences) are more likely to perceive the underlying differences between events. Undertaking additional FDIs in countries culturally proximate to the MNE's home country will confront the MNE with additional novel obstacles and issues related to cultural differences, providing it with opportunities to learn how to detect underlying differences between operations in foreign cultures. Facing a greater number of novel issues may also challenge the generalizability of the MNE's home-country-developed knowledge, improving managers' abilities to effectively evaluate the gaps between their prior knowledge and the knowledge required to succeed in a foreign culture (Petersen et al. 2008). The MNE may consequentially become increasingly capable of detecting knowledge generated by home-proximate experience that is not applicable to the focal FDI, and as indicated in Box 7 in Fig. 1, intentionally choose not to apply irrelevant knowledge. At the same time, a high level of home-proximate experience may allow the MNE to accumulate additional knowledge that is applicable toward a focal subsequent FDI (for example, general knowledge regarding dealing with foreign currencies and complying with different regulations) and lead to a beneficial effect of experience.

The above discussion implies an inverse U-shaped relationship between the level of home-proximate experience and the mortality of a focal subsequent FDI. A low level of home-proximate experience may harm the focal subsequent FDI because of the high likelihood of experience misapplication, but a high level of the same experience may benefit the focal FDI because of the diminished likelihood of experience misapplication as well as the accumulation of general international knowledge beneficial to a focal subsequent FDI. We expect that the mortality rate of the focal subsequent FDI is highest when the MNE has a level of home-proximate experience that would provide irrelevant knowledge for the MNE to draw on, yet is insufficient for the MNE to learn to effectively assess the applicability of the knowledge gained toward a focal subsequent FDI. We hypothesize that:

Hypothesis 2: The mortality of a subsequent FDI in a culturally dissimilar country will have an inverse U-shaped relationship with the MNE's experience in countries culturally similar to the MNE's home country but dissimilar to the focal host country.

Host-Proximate Experience

Host-proximate experience refers to an MNE's previous FDIs in countries that are culturally similar to the host country of a focal subsequent FDI, but dissimilar to the MNE's home country (i.e., previous-FDI-home-country cultural difference is large and previous-FDI-subsequent-FDI cultural difference is small). This type of experience also is only possible when a focal FDI is located in a country that is culturally dissimilar to the MNE's home country (e.g., for a Korean MNE initiating an FDI in the U.S. this experience refers to the MNE's prior experience in the United Kingdom, Canada, etc.). Host-proximate experience can generate knowledge that is highly relevant to a focal subsequent FDI due to the small previous-FDI-subsequent-FDI cultural difference (Barkema et al. 1996; Uhlenbruck 2004). However, the large previous-FDI-home-country cultural difference may erode the MNE's ability to effectively learn from this experience and increase the likelihood of drawing erroneous inferences.

Learning effectively from host-proximate experience is particularly difficult when the MNE only has a low level of this type of experience. In order to correctly interpret its experiences with previous FDIs in countries culturally dissimilar to the home country's culture the MNE must learn and understand the business practices, management styles, and consumer preferences specific to the host culture. This process may involve unlearning some home-country-developed knowledge and practices, as well as searching for and building relationships with local learning sources (Li and Scullion 2006: Tsang and Zahra 2008). These learning tasks take time and require significant trial and error. Home country mindsets and corporate practices can have a long-lasting imprinting effect on the management of overseas subsidiaries (Lubatkin et al. 1998; Nadolska and Barkema 2007). Identifying knowledge-holders within the local culture and building relationships can be both time-consuming and challenging, it is therefore likely that during this initial learning period causal relationships will be mis-specified and erroneous inferences will be drawn.

As the MNE accumulates greater host-proximate experience its likelihood of mis-specifying causalities and drawing erroneous inferences from this type of experience may decline. The MNE learns to operate within those host countries; accumulated host-proximate experience also provides the MNE with a knowledge base that will facilitate new learning (Cohen and Levinthal 1990: Levitt and March 1988). Individual managers will unlearn some of their home country-developed mindsets, acquiring novel ideas and solutions emanating from local sources. At the organizational level new structural adaptations and interpretational frames may be developed based on accumulated host-proximate experience (for example, regional headquarters may be formed and interpretations or decisions may be made at the regional level) (Arregle et al. 2009; Goerzen and Makino 2007). The MNE's ability to untangle causal relationships becomes augmented following this improvement in its learning process.

Erroneous inferences drawn from prior experience will harm a focal subsequent FDI if applied, as indicated in the path from Box 6 to 10 in Fig. 1. Since the initial accumulation of host-proximate experience is likely to generate erroneous inferences, the MNE must be cautious when applying this experience toward the focal subsequent FDI. However, due to the small prior-FDI-subsequent-FDI cultural difference, the MNE may take the applicability of host-proximate experience toward a subsequent FDI for granted and generalize it without deliberately evaluating the accuracy of the drawn inferences (O'Grady and Lane 1996). Given the high likelihood of drawing erroneous inferences when home-proximate experience is low combined with the MNE's strong tendency to apply this experience, we would expect a detrimental effect of home-proximate experience on the subsequent FDI at this stage.

As the level of host-proximate experience continues to increase, the MNE's likelihood of drawing erroneous inferences decreases and the erroneous inferences drawn early on are likely to be either corrected or abandoned. A beneficial effect of host-proximate experience is therefore likely to be observed once the MNE accumulates a certain level of host-proximate experience. When combining a decreasing detrimental effect at a lower level and an increasing beneficial effect at a higher level of host-proximate experience, we expect that:

Hypothesis 3: The mortality of a subsequent FDI in a culturally dissimilar country will have an inverse U-shaped relationship with the MNE's experience in countries culturally similar to the focal host country but dissimilar to the MNE's home country.

Double-Distant Experience

Double-distant experience refers to an MNE's previous FDIs in countries that are culturally similar to neither the host country of a focal subsequent FDI nor to the MNE's home country (i.e, previous-FDI-home-country cultural difference and previous-FDI-subsequent-FDI cultural difference are both large). This experience is possible regardless of a focal subsequent FDI's location. For example, for a Korean MNE initiating an FDI in the U.S. this experience refers to the MNE's prior experience in Germany, Denmark, etc.; if a subsequent FDI is in China this experience refers to the MNE's prior experience in the U.S., Germany, etc.

Similar to the logic underlying Hypothesis 3, the large previous-FDI-home-country cultural difference erodes the MNE's ability to effectively learn from this experience and consequently the likelihood for the MNE to draw erroneous inferences is high. However, the erroneous inferences possibly drawn from double-distant experience may not harm a focal subsequent FDI because the MNE's tendency to apply this experience to the focal FDI may be low. This is because even an MNE with a low level of double-distant experience may be able to decipher the differences and be cautious when applying this experience toward a focal subsequent FDI given the significant previous-FDI-subsequent-FDI cultural difference. Additionally, since the previous-FDI-home-country cultural difference is large, the MNE may face great challenges when exploiting its home-developed resources and exploring new resources in the host countries of double-distant experience. As a result, unlike in the case of home-proximate experience where the MNE may become overconfident with its ability to generalize its home-developed resources and operate in foreign countries, the accumulation of double-distant experience may not lead to such overconfidence and the misapplication of the experience.

The above discussion implies that the effect of double-distant experience may follow the path from Box 6 to 8 in Fig. 1--this experience may have a neutral effect on a focal subsequent FDI because it may not be applied. However, the literature indicates that the accumulation of double-distant experience may benefit a focal subsequent FDI by improving the MNE's ability to learn in foreign cultures in general, it has been acknowledged that rich experience in foreign cultures will provide an MNE with a heterogeneous knowledge base concerning foreign cultures (Barkema and Drogendijk 2007). This knowledge base will enhance the MNE's ability to tackle and solve novel issues, developing skills that will permit learning in foreign cultures that are likely to be generalized across cultures (Barkema and Drogendijk 2007; Levinthal and March 1993). We may therefore observe a negative relationship between the level of double-distant experience and the mortality of a focal subsequent FDI. Given these two possible effects of double-distant experience, we propose two competing hypotheses and leave the empirical analysis to test which is true in our sample:

Hypothesis 4a: The mortality of a subsequent FDI will have a neutral relationship with the MNE's experience in countries culturally dissimilar to both the MNE's home country and the focal host country.

Hypothesis 4b: The mortality of a subsequent FDI will have a negative relationship with the MNE's experience in countries culturally dissimilar to both the MNE's home country and the focal host country.

Methodology

Sample

We tested our hypotheses using a panel data of FDIs conducted by South Korean manufacturing MNEs listed on the Korea Stock Exchange (KSE) between 1973 and 2006. We dropped those that were initiated after 2004 to exclude pending projects or projects that were still under construction.

As previously discussed, the groupings of experience are different when a focal sub-sequent FDI is located in a country culturally similar to the MNE's home country versus when it is located in a dissimilar country. Therefore, we divided our data into two sub-samples, with one sample including FDIs in countries culturally similar to South Korea, and the other including those in countries that are dissimilar to South Korea. We used a widely used classification of national culture, cultural clusters as outlined by Ronen and Shenkar (1985), to distinguish countries that are culturally similar to South Korea from those that are dissimilar. Specifically, countries belonging to the same cultural cluster as South Korea, i.e. Far Eastern Asia, were defined as countries culturally similar to South Korea, while other countries were defined as culturally dissimilar to South Korea. Using cultural clusters is appropriate here because arbitrary judgment would be needed in order to define a high and low level of cultural differences if a continuous measure of national cultural differences was used. The Ronen and Shenkar classification is based on a synthesis of cross-cultural studies of work-related attitudes and values, and has been widely used in the literature (Barkema et al. 1996: Barkema and Drogendijk 2007).

Following these sampling criteria, we obtained 1432 FDI projects initiated by 251 MNEs. Among these FDI projects, 736 were located in the Far Eastern Asia cultural cluster and the rest were in other cultural clusters. We tracked the life history of these FDI projects annually from 1990 to 2006. Although some of the FDI projects in the data were launched prior to 1990, we were only able to begin observing the life history of these FDI projects from 1990 due to the unavailability of valid data prior to that time. However, this issue of left censoring does not present a serious threat to the validity of our analysis because 90% of the FDI projects in the data were begun after 1990. Additionally, we per formed a robustness test after excluding FDI projects founded before 1990 and obtained a similar results pattern.

We collected the data from multiple sources, including the Korea Listed Companies Association (KLCA) foreign affiliate database, company websites and news, LEXIS/ NEXIS, and the websites of the Munwha Daily Newspaper and Korea Information Service.

Model Specification

We used event history analysis (EHA) to examine the hazard of the mortality of a focal subsequent FDI. Compared to traditional models such as logistic regressions, EHA has the advantage of providing a dynamic analysis of the factors that may influence the occurrence of an event over time by tracking the life history of a firm. Furthermore, EHA uses information provided by right-censored cases and this approach avoids biases that can be caused by not considering whether or not an FDI exited after the observation period (Allison 1982; Vermeulen and Barkema 2001; Yamaguchi 1991). We applied a Cox's proportional hazard model with time-varying variables (Cox and Oakes 1984). Cox model estimates the influence of explanatory variables on the hazard rate of the event (the exit of an FDI project in our study) without specifying patterns of time dependence (Cox and Oakes 1984). This model is appropriate to our study because our interest is in the effect of the explanatory variables rather than the effect of time per se. The model can be expressed as follows:

[h.sub.i](t) = [h.sub.0](t)exp([[beta].sub.1][x.sub.i1] + [[beta].sub.2][x.sub.i2] + ... + [[beta].sub.k][x.sub.ik])

where [h.sub.i](t) is the dependent variable, denoting the hazard rate for FDI i to exit from the host country at time t. [h.sub.0](t) is the baseline hazard function that is not estimated as it is common across the population and can take any form as a function of time, [x.sub.il] to [x.sub.ik] are the independent variables, and [[beta].sub.l] to [[beta].sub.k] are coefficients that will be estimated. The life history of each FDI over the study period was divided into annual spells. During each spell an FDI was at risk of exit. Each spell was treated as right censored unless the FDI exited that year. The values of the status of the subsidiaries and the values of the time-varying variables (e.g., subsidiary ownership, MNE size, age, and profitability) were lagged by 1 year and updated annually.

Because there may be dependence among FDIs imitated by the same MNE, which could lead to biased results if not controlled for (Arregle et al. 2006), we used the "Robust" and "Cluster" STATA procedures to estimate a robust coefficient (Cameron and Trivedi 2010). We also used the STATA "Stphtest" command to check if our data violates a key assumption of the Cox model, that is, the ratio of hazards for any two observations is the same across time periods. No violation of the assumption was found.

Dependent Variable

Our dependent variable is the hazard rate for an FDI to exit. The data for the EHA is a panel data where each FDI project has one observation for each year before it exited, in each year a dummy variable was used to indicate the survival status of an FDI project. The variable equaled 1 if an FDI project exited from the database and 0 otherwise.

Independent Variables

Our four independent variables are: (1) Double-proximate experience, referring to previous FDIs in countries that are culturally similar to both the MNE's home country and the host country of a focal subsequent FDI: (2) home-proximate experience, referring to previous FDIs in countries that are culturally similar to the MNE's home country but different from the host country of a focal subsequent FDI: (3) host-proximate experience, referring to previous FDIs in countries that are culturally different from the MNE's home country but similar to the host country of a focal subsequent FDI; (4) double-distant experience, referring to previous FDIs in countries that are culturally different from both the MNE's home country and the host country of a focal subsequent FDI.

Consistent with the way in which we defined our sample, we used cultural clusters to define these four types of experience in order to avoid arbitrarily defining similar and dissimilar cultures using a continuous measure of cultural differences. Specifically, for the subsample including FDIs outside of the Far Eastern Asia cluster, we divided the host countries of an MNE's FDIs prior to a focal subsequent FDI into three groups using the Ronen and Shenkar clusters: (1) Countries belonging to the Far Eastern Asia cluster; (2) countries belonging to the same cultural cluster as the host country of the focal FDI; and (3) countries that are neither in the Far Eastern Asia cluster nor in the focal host country's cluster. We then calculated the number of FDIs that the MNE has had in each of these groups of countries prior to the focal FDI to obtain home-proximate experience, host-proximate experience, and double-distant experience.

For the subsample including FDIs in the Far Eastern Asia cluster, we divided the host countries of the MNE's FDIs prior to a focal subsequent FDI into two groups: The first includes countries in the Far Eastern Asia cluster and the other includes countries outside of the Far Eastern Asia cluster. We then calculated the number of FDIs that the MNE has had in these two groups of countries prior to the focal FDI to obtain the two types of experience that are possible for this subsample (i.e., double-proximate experience and double-distant experience).

Using the number of FDIs to measure the level of FDI experience is consistent with previous studies (e.g., Barkema et al. 1996, 1997; Barkema and Drogendijk 2007; Nadolska and Barkema 2007). Theoretically, our arguments with respect to the effect of host-proximate experience are applicable to both experience in the focal host country and experience in other countries within the host country's cultural cluster. However, since host country experience has been treated as a distinct type of experience in the literature (e.g., Barkema and Drogendijk 2007: Li 1995), we tested its effect separately. Host country experience was measured by the number of FDIs the MNE had initiated in the host country prior to the focal FDI.

As a robustness check, we also grouped FDI experience using cultural clusters developed from a more recent study on national culture, the Global Leadership and Organizational Behavior Effectiveness (GLOBE) project (House et al. 2004). This project included 62 countries clustered into 10 groups (Gupta et al. 2002). One major difference between the GLOBE and the Ronen and Shenkar clusters is that eastern Asian countries were clustered into two subgroups in the former (Southern Asia and Confucian Asia), and one group (Far Eastern Asia with Japan as an exception) in the latter. Following Hutzschenreuter and Voll (2008), countries that were not included in these classifications were assigned into a cultural cluster based on their religion, language, and ethnicity using the CIA World Factbook. (2)

Control Variables

At the FDI project level we controlled for the ownership type of the project, i.e. joint ventures (JVs) or wholly owned subsidiaries (WOSs). Previous studies indicate that JVs have higher failure rates than WOSs (e.g., Gaur and Lu 2007). A subsidiary was defined as a JV when the equity ownership of the Korean MNE was less than 95 % and as a WOS otherwise.

At the MNE level we controlled for firm age, size, annual profitability, and the overall FDI experience. Older, larger, and more profitable MNEs have more resources and a greater opportunity to succeed in FDI (Barkema et al. 1996). An MNE's age was measured as the number of years since its founding, its size was measured as the logarithm of its total assets, and its profitability was measured as the return on sales (ROS). Overall FDI experience was measured as the total number of FDI projects an MNE had initiated prior to a focal FDI. Studies have shown that an incremental path of internationalization (i.e., the MNE begins international expansions in countries similar to its home country) may result in more effective learning because the MNE is able to accumulate international knowledge gradually (Hutzschenreuter and Vol 12008; Johanson and Vahlne 1977). In order to control for the differences in MNEs' internationalization paths, we added a dummy variable that equaled 1 if an MNE began its FDI expansion in South Korea's cultural cluster, and 0 if the MNE began in other cultural clusters. We controlled for industry effects (Li 1995) by adding dummy variables representing the four-digit South Korean Standardized Industry Code of an MNE's primary industry. Additionally, we controlled for industry research and development (R&D) intensity and marketing intensity, measured with the percentage of R&D and advertising expenditures over sales. MNE and industry information was obtained from the Korea Stock Exchange (KSE).

We controlled for the gross domestic product (GDP) growth rates of the host countries (Thomas et al. 2007); this data was obtained from the International Monetary Fund. Although we do not focus on the direct effect of cultural differences on subsidiary mortality, there has been evidence that cultural distance between an MNE's home country and the host country of a focal subsidiary FDI affects subsidiary performance (e.g., Barkema et al. 1997). Following Kogut and Singh (1988), we calculated the cultural distance between the host country and South Korea using the national cultural index developed by Hofstede (Hofstede 1980, 2001). Since it is possible that a Korean MNE may learn from the experience of other Korean MNEs in the host country (particularly within the same industry) (Chang 1995; Shaver et al. 1997), we controlled for the number of FDIs invested in by South Korean MNEs from the same industry as the focal MNE within the host country. The mortality of overseas subsidiaries may also be impacted by the overall legitimacy of FDI from the home country (Kostova and Zaheer 1999). We therefore controlled for the exit rate of Korean MNEs in the focal host country, measured as the ratio of Korean FDI projects that exited the country as a proportion of the total number of Korean FDI projects in the country.

Analysis and Results

Tables 1 and 2 summarize the descriptive statistics and the correlation matrix of the independent and control variables for the two subsamples respectively. In order to reduce the threat of multicollinearity, all experience variables were mean-centered. Furthermore, for both of the two subsamples each type of experience was added separately in the regression and then added together in the final model. We calculated the variance inflation factor (VIF) and found that the VIFs for our theoretical variables were below the rule-of-thumb threshold value of 10 (Neter et al. 1996). Finally, because the correlations between an MNE's overall FDI experience and the for subsets of experience were high, we conducted the analysis both with and without controlling for the overall experience. The results were similar, so we reported the results including the overall experience variable. For all MNE's first FDI, all experience variables equaled zero (i.e., no prior FDI experience). In order to ensure that results were not driven by too many zeros, we performed the analysis both with and without the first FDI of each MNE. The results were not qualitatively different. We report the results without the first FDI.

Table 3 reports the results for the subsample including FDIs located outside of the cultural cluster of South Korea. Model 1 includes only the control variables. The results for the industry dummies are omitted in all models in order to make the table more readable. Models 2 and 3 test the effects of home-proximate experience. In Model 3 the coefficient of the variable is positive and significant, and the coefficient of the squared term of the variable is negative and statistically significant. This indicates an inverse U-shaped relationship between the mortality of a focal subsequent FDI and the MNE's prior home-proximate experience, supporting Hypothesis 2.

Models 4 and 5 test the effects of host-proximate experience (excluding host country specific experience). In Model 5 the coefficients of the first and quadratic terms of the host-proximate experience are consistent with Hypothesis 3, indicating an inverse U-shaped relationship between the mortality of a focal FDI project and the MNE's host-proximate experience. Models 6 and 7 examine the effect of host country specific experience. In Model 6 the linear effect is negative but insignificant. In Model 7 both the first and quadratic terms of the variable are significant and consistent with the directions predicted in Hypothesis 3.

Models 8 and 9 examine the effect of double-distant experience. In Model 8 the linear effect is negative and significant, showing that the morality rate of a focal subsequent FDI decreases with the increase of the MNE's double-distant experience. In Model 9 neither the linear effect nor the quadratic term are significant. Therefore, taken together, Models 8 and 9 support Hypothesis 4b, which predicts a negative relationship between the mortality, of a focal subsequent FDI and the MNE's double-distant experience. In Model 10 all experience variables were added into the model, and the effects remain unchanged except that the coefficient of double-distant experience was no longer statistically significant.

[FIGURE 3 OMITTED]

Figure 3 graphs the effects of home-proximate, host-proximate, and double-distant experience on the mortality of FDIs located outside of South Korea's cultural cluster. It can be clearly seen that the three types of experience affect the mortality of a subsequent FDI in different ways. For home-proximate experience, approximately eleven experiences are needed before subsequent FDIs can benefit. For host-proximate experience, the threshold is four previous FDIs. With respect to the magnitude of the effect of the three types of experience, host-proximate experience has the strongest effect, followed by home-proximate experience and double-distant experience.

Among the control variables the MNE's age, the exit rate of other South Korean MNEs from the host country, and industry R&D intensity have significant coefficients. The results show that FDI projects initiated by older firms and by firms in industries with a high level of R&D intensity are less likely to exit. Also, the exit of FDIs by peer MNEs decrease the mortality of a focal subsequent FDI, which is contradictory to the notion that the legitimacy of a firm's FDIs in a foreign country will be negatively influenced by the exit of other projects from the same home country. This finding may suggest that the density of South Korean FDI projects in the host country already exceeds the level needed to obtain overall legitimacy for South Korean FDI, therefore, the exit of other South Korean MNEs reduces competition in the market without eroding the legitimacy of South Korean FDI in the host country.

Table4 reports the results for the subsample including FDIs located within South Korea's cultural cluster. For this sample, only double-proximate and double-distant experiences are possible when examining the mortality of a subsequent FDI. Model 1 includes only the control variables. In Model 2, we added double-proximate experience (host country specific experience was tested separately), i.e. previous FDIs in countries belonging to South Korea's cultural cluster. The variable has a significantly negative coefficient, supporting Hypothesis 1. In Model 3, we tested host country specific experience. The variable has a positive effect on subsidiary mortality, contrasting the prediction of Hypothesis 1. A possible explanation is that because the cultural difference between the home country and host country of the focal subsequent FDI is small, the MNE may choose to exploit its home country developed knowledge and practices to a great extent in the host country. The accumulation of host country specific experience therefore may lead to rigidity and a lack of response to the unique circumstances of a subsequent subsidiary (Levinthal and March 1993: Vermeulen and Barkema 2001). Our additional analysis (reported in Appendix A) seems to support this explanation by showing that the detrimental effect of host country specific experience is mainly from the MNE's experience with wholly owned subsidiaries, an ownership type that facilitates resource exploitation (Hennart 1991; Kogut and Chang 1991).

[FIGURE 4 OMITTED]

Model 4 tested the effect of double-distant experience. The coefficient of the variable is not statistically significant, supporting Hypothesis 4a which predicts a neutral relationship between this experience and the mortality of a subsequent FDI. Model 5 is the full model that includes all variables. Although the effects of double-proximate experience and double-distant experience remained unchanged, the effect of host country specific experience lost significance. Among control variables, WOSs have a lower mortality than JVs. Additionally, an MNE's age and profitability have a negative effect while an MNE's size has a positive effect on subsidiary mortality.

Figure 4 graphs the effect of double-proximate on the mortality of FDIs located within South Korea's cultural cluster. It can be clearly seen that FDI mortality decreases as double-proximate experience increases. We do not graph the effect of double-distant experience since this effect is not statistically significant as shown in Table4.

We performed two sets of analysis using the GLOBE clusters. The first used Confucian Asia as South Korea's home culture cluster: the second used Confucian Asia and Southern Asia together as South Korea's home culture cluster. For the subsample that contains FDIs located within South Korea's cultural cluster, the results were almost identical to those reported in Table4. For the subsample containing FDIs located outside South Korea's cultural cluster, the results of the two sets of analysis were similar to the results reported in Table 3, except that in the first set of analysis the coefficient of host-proximate experience (i.e., Models 4 and 5 in Table3) was not statistically significant. A possible explanation is that countries in Southern Asia share high cultural similarities with South Korea, and classifying these countries in a different cultural cluster blurs the boundary between home-proximate and host-proximate experiences.

We discussed other additional analyses in Appendix A.

Discussion and Conclusion

This study was intended to provide a comprehensive understanding of the FDI experience-performance relationship when considering the possible learning limitations an MNE may be susceptible to when operating in dissimilar cultures. We departed from the traditional assumption of a positive linear experience-performance relationship and examined a possible detrimental effect of experience by considering the learning limitations that may constrain an MNE when it draws inferences from its previous FDIs and applies its experience toward subsequent FDIs. We found that the mechanisms underlying the effect of experience on subsequent FDIs vary with the location of the MNE's prior experience as measured using two types of pertinent cultural differences: The difference between the MNE's home country and the host countries where it had previously invested and the difference between host countries of those previous FDIs and the host country of a focal subsequent FDI.

First, our results showed that a subsequent FDI located in a country culturally similar to an MNE's home country benefits from the MNE's prior double-proximate experience, a type of experience where both of the two pertinent cultural differences are small. This situation seems to meet the premise of effective learning where a firm is capable of drawing accurate inferences from its experience and applying its experience appropriately toward subsequent activities, leading to the presence of a learning curve and beneficial effect of experience (Levinthal and March 1993: Levitt and March 1988; Yelle 1979).

Second, consistent with Johanson and Vahlne's (1977) emphasis on the value of FD1 experience in countries adjacent to an MNE's home country on its international expansions, our findings show that home-proximate experience is potentially valuable to a focal subsequent FDI located in a culture that is dissimilar to the home country culture. However, a considerable number of previous FDIs (eleven in our sample) is needed in order for the MNE to learn to discern the transferability of knowledge gained from this type of experience in culturally dissimilar countries.

Third, we found an inverse U-shaped relationship between the mortality of a focal subsequent FDI located in a culturally dissimilar country and host-proximate experience. (3) This finding is consistent with recent studies' emphasis on the value of local experience (e.g., Barkema et al. 1996; Uhlenbruck 2004). Nevertheless, our results show that multiple FDIs (four in our sample) in the early period are needed in order for the MNE to figure out how to learn in host cultures that are significantly different from its home country culture. This complements Barkema and Drogendijk's (2007) argument that a certain level of international experience is needed in order for an MNE to improve its learning capability in host countries dissimilar from its home country.

Fourth, our findings show that in our sample, double-distant experience (i.e., both of the two types of pertinent cultural differences are large) has a negative relationship with the mortality of a subsequent FDI located in countries dissimilar to the MNE's home country, but a neutral relationship with the mortality of a subsequent FDI located in countries similar to the MNE's home country. This may indicate that investing in culturally different countries allows an MNE to accumulate a rich knowledge base concerning foreign cultures, and this knowledge can be generalized to improve the survival of a subsequent FDI in a dissimilar culture (Barkema and Drogendijk 2007). However, since an MNE may face fewer cultural issues when operating in a country culturally similar to its home country and its ability to learn in a similar country is high, the survival of a subsequent FDI in a similar country may not depend on the knowledge resulting from double-distant experience.

Finally, as shown in Fig. 3, both the shape and magnitude of the three types of experience vary when a focal subsequent FDI is located in a dissimilar country. Host-proximate experience provides a faster and stronger learning effect compared to home-proximate experience: The inverse U-shaped relationship is narrower and sharper. This finding suggests that host-proximate experience provides better opportunities to develop new knowledge and generates a larger amount of knowledge relevant to a focal subsequent FDI. Furthermore, in our sample home-proximate experience seems to have the strongest counterproductive effect since subsidiaries by MNEs without this experience have a better likelihood of survival. Finally, each increase in double-distant experience has the smallest effect on changes in subsidiary mortality, supporting our argument that the MNE's learning capability is low and a large amount of the knowledge gained from this experience is irrelevant to a focal subsequent FDI.

This study contributes to the international business literature in several ways. First, our findings show that MNEs are susceptible to both drawing erroneous inferences from their experiences and misapplying experiences. Scholars must therefore consider an MNE's ability to learn from experience as well its ability to discern the relevance of experience in a focal context before assuming a beneficial effect of FDI experience. Second, our findings of different effects of the four types of experience indicate that defining experience sub-groups is necessary in order to fully understand the effects of FDI experience on subsequent FDIs. This resonates with Barkema and Schijven's (2008, p. 613) argument that studies need to focus on "specific subsets of experience rather than lumping together disparate types of experience" in order to better understand the experience-performance relationship.

Third, we advance the literature by relating the likelihood of an MNE's drawing erroneous inferences and misapplying experience to two types of pertinent national cultural differences: The difference between the MNE's home country and the host countries where it had previously invested, and the difference between the host countries of previous FDIs and the host country of a focal subsequent FDI. Although scholars have acknowledged the hindering effects of these two types of cultural differences in learning and knowledge transfer in FDI (Barkema et al. 1996; Bhagat et al. 2002; Bjorkman et al. 2007; Nadolska and Barkema 2007; Uhlenbruck 2004), the extant literature has not considered both of them simultaneously when examining the effect of FDI experience. Our study shows that grouping experience by considering both these cultural differences allows us to acquire a more comprehensive understanding of the effect of FDI experience.

This study also has implications for international business practitioners. We find that the accumulation of knowledge in culturally dissimilar countries is challenging and the knowledge gained can initially be misleading, but this knowledge is rewarding once the MNE "learns how to learn" in these countries. MNE managers must therefore cautiously assess the accuracy of inferences drawn from experience in a new culture and deliberately create learning mechanisms that will enhance their firms' learning abilities in the host culture. In contrast, the accumulation of experiential knowledge from countries proximate to the MNE's home country is less challenging, but the knowledge gained could be misapplied when the MNE invests in culturally dissimilar countries. The implication is that MNE managers with limited experience in countries proximate to their home countries may need to consciously avoid depending on their home-proximate experience when operating in a dissimilar culture. Furthermore, managers may develop processes such as formal internal auditing processes in order to assess the applicability of prior experience and reduce the likelihood of experience misapplication (Zollo 2009).

Finally, this study also contributes to the organizational learning literature by providing a comprehensive investigation of the effects of experience when considering both types of learning limitations identified in the literature. Although the theoretical arguments for these learning limitations are strong as shown in Fig. l, previous studies have typically focused on one at a time. Our study shows that considering both learning limitations simultaneously allows us to draw a more complete picture of the experience-performance relationship and better understand the mechanisms underlying the effects of experience in complex settings such as FDI.

This study is also not without limitations. First, although categorizing firms' experiences in other clusters into one group (double-distant experience) simplifies the comparison of the four types of experience, there may be differences among experiences within this group. Further studies could be conducted to specifically examine differences in the effects of experience from double-distant countries. Second, for the sake of simplicity we did not consider differences in the learning effect of prior FDI projects with different performance outcomes. However, the literature has shown that the learning effect of experience is influenced by the outcomes of prior operations (Levinthal and March 1993; Sitkin 1992). In the future it would be valuable to include the outcomes of prior FDI projects when examining the effects of different experience types. Third, prior research has shown that subsidiary performance may depend on the relatedness and location specificity of a focal subsidiary (Goerzen and Markino 2007). We were not able to control for this possibility due to data limitations. Future studies could explore whether or not the effect of experience varies with the level of relatedness between a subsidiary and the parent company, as well as with the subsidiary location specificity.

Finally, although our research design allows us to identify situations where the two learning limitations are likely to occur in FDI, we were not able to directly measure these learning limitations due to data constraints (this limitation is generally true for studies examining learning limitations using secondary data (e.g., Haleblian and Finkelstein 1999; Heimeriks 2010; Kim et al. 2009; Nadolska and Barkema 2007; Zollo 2009)). It would be meaningful to use methods such as surveys in order to directly detect the existence of these learning limitations and measure their magnitude in future studies.

DOI 10.1007/s11575-013-0169-0

Received: 14.10.2011 / Revised: 30.09.2012 / Accepted: 20.11.2012 / Published online: 02.03.2013

Appendix: "Additionally Analysis" (All Results are Available from the Authors upon Request)

We conducted several additional robustness tests on our results. First, in order to test if the effects of the four types of experience vary with the ownership of a subsequent FDI, we added interaction terms between the dummy variable identifying WOSs and the experience variables. Host-proximate experience had a less detrimental effect on WOSs. This may be due to the possibility that WOSs are less dependent on the MNE's learning about local culture. Additionally, double-distant experience had a stronger beneficial effect on WOSs than on JVs located in countries dissimilar to South Korea. Since JVs typically encounter more cultural issues than WOSs (Barkema et al. 1996), this finding supports our argument that a large portion of the knowledge gained from a third cultural cluster may lose value in a different cultural cluster.

Second, in order to examine if WOS and JV experiences have different learning effects (Barkema et al. 1996: Vermeulen and Barkema 2001), we reran all the models in Tables 1 and 2 using only the MN E's WOS experience or JV experience. For the subsample including FDIs in countries dissimilar to South Korea, the effects of the three types of JV experience and three types of WOS experience were similar to those in Table 3, except that the effect of the first order of host-proximate WOS experience (the MNE's WOS experience in other countries within the focal host country's cultural cluster) was not significant. For the subsample including FDIs in countries similar to South Korea, double-distant WOS experience and double-distant JV experience had very different effects. While double-distant WOS experience had a significantly positive relationship with subsidiary mortality, double-distant JV experience had a significantly negative relationship with subsidiary mortality. Additionally, host country specific JV experience does not have a significant effect while host country specific WOS experience has a detrimental effect.

Third, in order to capture the possible effect of time on the effectiveness of learning, we measured the tour types of experience using the total number of subsidiary-year that an MNE had operated in the host country's cluster, the home country's cluster, or other cultural clusters (Delios and Beamish 2001: Gaur and Lu 2007). A subsidiary-year represented 1 year of operations by one FDI project. The results were very similar to those reported in Tables I and 2, except that the coefficient of double-proximate experience lost significance.

Fourth, due to data limitations, we were not able to determine the reasons that an FDI was terminated. It is possible, however, that an FDI project might be terminated after the parent company achieved its goals in the FDI. In this case, the exit of the FDI is an indicator of success instead of failure. In order to ensure that our results were not affected by falsely defining FDI exit, following Delios and Beamish (2001), we coded the dependent variable as I only if the FDI was less than 10 years old when it exited from the data and reran the regressions. The results remained largely unchanged except that the coefficient of double-distant experience lost significance in the subsample including FDIs in culturally dissimilar countries and the coefficient of host country specific experience lost significance in the subsample of FDIs in culturally similar countries.

Acknowledgements: We thank Editor Joachim Wolf and three anonymous reviewers for their constructive comments and suggestions. Our appreciation also goes to Dr. Dean Xu for his helpful comments on an earlier version of the paper. This work was supported by the National Research Foundation of Korea Grant funded by the Korean Government (NRF-2012-S1A3A22012SIA3A2033412). Sangcheol Song was the recipient.

References

Allison, P. D. (1982). Discrete-time methods for the analysis of event histories. In S. Leinhardt (Eds.). Sociological methodology (pp.61-98). San Francisco: Jossey-Bass.

Arregle, J.-L., Hebert, L., & Beamish, P. W. (2006). Mode of international entry: The advantages of multilevel methods. Management International Review, 46(5). 597-218.

Arregle, J.-L., Beamish, P. W., & Hebert, L. (2009). The regional dimension of MNEs' foreign subsidiary localization. Journal of International Business Studies. 40(1), 86-107.

Barkema, H. (2006). How to avoid ruining your company through international acquisitions: The Ahold case. In C. L. Cooper & S. Finkelstein (Eds.), Advances in mergers and acquisitions (pp. 105-126). Amsterdam: Elsevier.

Barkema, H. G., & Drogendijk, R. (2007). Internationalising in small, incremental or larger steps? Journal of International Business Studies, 38(7). 1132-1148.

Barkema, H. G., & Schijven, M. (2008). How do firms learn to make acquisitions'? A review of past research and an agenda for the future. Journal of Management. 34(3), 594-634.

Barkema, H. G., Bell, J. H. J.. & Pennings, J. M (1996). Foreign entry, cultural barriers, and learning. Strategic Management Journal, 17(2), 151-166.

Barkema, H. G., Shenkar, O., Vermeulen, F., & Bell, J. H. J. (1997). Working abroad, working with others: How firms learn to operate in international joint venture. Academy of Management Journal. 40(2), 426-442.

Benito, G. R. G., & Gripsrud, G. (1992). The expansion of foreign direct investments: Discrete rational location choices or a cultural learning process. Journal of International Business Studies, 23(3). 461-476.

Bhagat, R. S., Kedia, B. L.. Harveston, P. D.. & Yriandis, H. C. (2002). Cultural variations in the cross-border transfer of organizational knowledge: An integrative framework. Academy of Management Review, 27(2). 204-221.

Bjorkman, I., Stahl, G. K., & Vaara, E. (2007). Cultural differences and capability transfer in cross-border acquisitions: The mediating roles of capability complementarity, absorptive capacity, and social integration. Journal of International Business Studies, 38(4), 658-672.

Cameron, A. C., & Trivedi, P. K. (2010). Microeconometrics using Stata (Revised ed.). College Station: Stata.

Chang, S. J. (1995). International expansion strategy of Japanese firms: Capability building through sequential entry. Academy of Management Journal. 38(2). 383-407.

Cohen, W. M., & Levinthal, D. A. (1990). Absorptive capacity: A new perspective on learning and innovation. Administrative Science Quarterly, 35(1), 128-152.

Cox, D. R., & Oakes, D. (1984). Analysis of survival data, monographs on statistics and applied probability. London: Chapman & Hall.

Cyert, R. M.. & March, J. G. (1963). A behavioral theory of the firm. Cambridge: Blackwell Business.

Delios, A., & Beamish, E W. (2001). Survival and profitability: The roles of experience and intangible assets in foreign subsidiary performance. Academy, of Management Journal, 44(5), 1028-1038.

Erikkson, K., Johanson, J., Majkgard, A., & Sharna, D. D. (1997). Experiential knowledge and cost in the internationalizaton process. Journal of International Business Studies, 28(2), 337-360.

Fang, Y., Wade, M., Delios, A., & Beamish, P. W. (2007). International diversification, subsidiary performance, and the mobility of knowledge resources. Strategic Management Journal, 28(10), 1053-1064.

Feldman, J. (1986). On the difficulty of learning from experience. In H. P. Sims & D. A. Gioa (Eds.), The thinking organization: Dynamics of organizational social cognition (pp. 263-292). San Francisco: Jossey-Bass.

Finkelstein, S., & Haleblian, J. (2002). Understanding acquisition performance: The role of transfer effects. Organization Science, 13(1), 36-47.

Fiol, C. M.. & Lyles, M. A. (1985). Organizational learning. Academy of Management Review. 10(4), 803-813.

Gaur, A. S., & Lu, J. W. (2007). Ownership strategies and survival of foreign subsidiaries: Impacts of institutional distance and experience. Journal of Management, 33(1), 84-110.

Gavetti, G., Levinthal, D. A,. & Rivkin, J. W. (2005). Strategy making in novel and complex worlds: The power of analogy. Strategic Management Journal, 26(8), 691-712.

Goerzen, A.. & Makino, S. (2007). Multinational corporation internationalization in the service sector: A stud), of Japanese trading companies. Journal of International Business Studies, 38(7), 1149-1169.

Gupta, V.. Hanges, P. J., & Dorfman, P. (2002). Cultural clusters: Methodology and findings. Journal of World Business, 37(1). 11-15.

Haleblian, J., & Finkelstein, S. (19991. The influence of organizational acquisition experience on acquisition performance: A behavioral learning perspective. Administrative Science Quarterly, 44(1). 29-56.

Heimeriks, K. H. (2010). Confident or competent? How to avoid superstitious learning in alliance portfolios. Long Range Planning, 43(1). 57-84.

Hennart, J.-F. (1991). The transaction costs theory of joint ventures: An empirical study of Japanese subsidiaries in the United Stales. Management Science, 37(4). 483-497.

Hoftede, G. (1980). Culture's consequences: International differences in work-related values. London: Sage.

Hofstede, G. (2001). Culture's consequences: International differences in work-related values. New, bury Park: Sage.

House, R. J.. Hanges, P. J., Javidan, M., Dorfman. P.. & Gupta. V. (2004). Culture, leadership, and organizations: The GLOBE study of 62 societies. Thousand Oaks: Sage.

Huber. G. P. (1991). Organizational learning: The contributing processes and the literatures. Organization Science, 2(1). 88-115.

Hutzschenreuter, T., & Voll, J. (2008). Performance effects of "added cultural distance" in the path of international expansion: The case of German multinational enterprises. Journal of International Business Studies. 39(1). 53-70.

Johanson, J., & Vahlne, J.-E. (1977). The internationalization process of the firm: A model of knowledge development and increasing foreign market commitments. Journal of International Business Studies, 8(1). 25-34.

Johanson, J., & Vahlne, J.-E. (1990). The mechanism of internationalism. International Marketing Review, 7(4). 11-25.

Kim, T.-Y.. Delios, A., & Xu, D. (2010). Organizational geography, experiential learning and subsidiary exit: Japanese foreign expansions in China. 1979-2001. Journal of Economic Geograpty, 10(4). 579-597.

Kirkman, B. L.. Lowe, K. B., & Gibson, C. B. (2006). A quarter century of culture's consequences: A review of empirical research incorporating Hofstede's cultural values framework. Journal of International Business Studies, 37(3). 285-320.

Kogut, B., & Chang, S. J. (1991). Technological capabilities and Japanese foreign direct investment in the United States. Review of Economics & Statistics, 73(3). 401-413.

Kogut, B., & Singh. H. (1988). The Effect of national culture on the choice of entry mode. Journal of International Business Studies. 19(3). 411-432.

Kostova, T., & Zaheer, S. (1999). Organizational legitimacy under conditions of complexity: The case of the multinational enterprise. Academy of Management Review, 24(1). 64-81.

Levinthal, D. A., & March. J. G. (1993). The myopia of learning. Strategic Management Journal 14.95-112.

Levitt, B., & March, J. G. (1988). Organizational learning. Annual Review of Sociology, 14(1), 319-340.

Li, J. (1995). Foreign onto and survival: Effects of strategic choices on performance in international markets. Strategic Management Journal. 16(5). 333-351.

Li, S., & Scullion, H. (2006). Bridging the distance: Managing cross-border knowledge holders. Asia Pacific Journal of Management. 23(1). 71-92.

Lu, J. W., & Beamish, R W. (2001). The internationalization and performance of SMEs. Strategic Management Journal, 22(6/7). 565.

Lu, J. W., & Beamish, P. W. (2004). International diversification and performance: The S-curve hypothesis. Academy of Management Journal, 47(4), 598-609.

Lubatkin, M., Calori, R., Very, P., & Veiga, J. F. (1998). Managing mergers across borders: A two-nation exploration of a nationally bound administrative heritage. Organization Science, 9(6), 670-684.

Makino, S., & Delios, A. (1996). Local knowledge transfer and performance: Implications for alliance formation in Asia. Journal of International Business Studies, 27(5), 905-927.

Nadolska, A., & Barkema, H. G. (2007). Learning to internationalise: The pace and success of foreign acquisitions. Journal of International Business Studies, 38(7), 1170-1186.

Neter, J., Kutner, M. H., Nachtschiem, C. J., & Wasserman. W. (1996). Applied linear statistical models. Chicago: Irwin.

O'Grady, S., & Lane, H. W. (1996). The psychic distance paradox. Journal of International Business Studies, 27(2), 309-333.

Petersen, B., Pedersen, T., & Lyles, M. A. (2008). Closing knowledge gaps in foreign markets. Journal of International Business Studies. 39(7), 1097-1113.

Ronen, S., & Shenkar, O. (1985). Clustering countries on attitudinal dimensions: A review and synthesis. Academy of Management Review, 10(3), 435-454.

Rugman, A. M., & Verbeke, A. (1992). A note on the transnational solution and the transaction cost theory of multinational strategic management. Journal of International Business Studies, 23(4), 761-771.

Shaver, J. M, Mitchell, W., & Yeung, B. W. (1997). The effect of own-firm and other-firm experience on foreign direct investment survival in the United States, 1987-1992. Strategic Management Journal, 18(10), 811-824.

Sitkin, S. (1992). Learning through failure: The strategy of small losses. In B. M. Staw & L. L. Cummings (Eds.), Research in organizational behavior (pp. 231-266). Greenwich: JAI.

Thomas, D., Eden, L., Hitt, M., & Miller, S. (2007). Experience of emerging market firms: The role of cognitive bias in developed market entry and survival. Management International Review, 47(6), 845-867.

Tsang, E. W. K., & Zahra, S. A. (2008). Organizational unlearning. Human Relations, 61(10), 1435-1462.

Uhlenbruck, K. (2004). Developing acquired foreign subsidiaries: The experience of MNEs in transition economies. Journal of International Business Studies, 35(2), 109-123.

Vermeulen, F., & Barkema, H. (2001). Learning through acquisitions. Academy of Management Journal, 44(3). 457-476.

Yamaguchi, K. (1991). Event history analysis. Newbury Park: Sage.

Yelle, L. E. (1979). The learning curve: Historical review and comprehensive survey. Decision Sciences. 10(2), 302-328.

Zollo, M. (2009). Superstitious learning with rare strategic decisions: Theory and evidence from corporate acquisitions. Organization Science, 20(5), 894-908.

Prof. Y. P. Zeng ([mail])

School of Business, Southern Illinois University Edwardsville, Edwardsville, IL, USA

e-mail: yzeng@siue.edu

Prof. O. Shenkar

Fisher College of Business, Ohio State University, Columbus, OH, USA

Prof. S. Song

Haub School of Business, Saint Joseph's University, Philadelphia, PA, USA

Prof. S.-H. Lee

Jindal School of Management, University of Texas at Dallas, Richardson, TX, USA

Endnotes

(1) Grouping an MNE's experience is necessary in order to capture the two types of cultural differences simultaneously, since it is empirically impractical to examine the effect of individual prior FDI experiences on a subsequent focal FDI.

(2) https://www.cia.gov/library/publications/the-world-factbook/.

(3) Since both host-proximate and home-proximate experiences have an inverted U-shaped effect on the mortality of a subsequent FDI in a dissimilar country, one may argue that the underlying mechanism of the detrimental effect of these two types of experience may be the same. However. if the detrimental effect of host-proximate experience is caused by experience misapplication as opposed to a low learning ability of the MNE that we suggested in the paper, we should also expect a similar detrimental effect of double-proximate experience since there is no evidence showing that an MNE may more likely misapply experience in dissimilar cultures. Furthermore. if the detrimental effect of home-proximate experience is caused by a low learning ability of the MNE as opposed to experience misapplication that we suggested, we should also see a detrimental effect of double-proximate experience since it is also a type of home-proximate experience. Our finding of a beneficial effect of double-proximate experience does not validate these two possibilities, supporting our arguments that the mechanisms underlying the detrimental effects of host and home-proximate experience are different.
Table 1: Descriptive statistics and correlations for the subsample
of FDIs located outside the home country's cultural cluster

 Variables Mean S.D. 1 2

1. Home-proximate 4.18 6.85
 experience

2. Host-proximate 1.15 2.13 0.64 *
 experience

3. Double-distant 7.51 13.14 0.90 * 0.61 *
 experience

4. Host country spe- 0.89 2.02 0.62 * 0.56 *
 cific experience

5. South Korea-Host 3.27 1.61 -0.08 * 0.11 *
 country cultural
 distance

6. Subsidiary 0.67 0.47 0.00 0.11 *
 ownership (a)

7. Internationaliza- 0.17 0.38 -0.11 * -0.21 *
 tion path (b)

8. Finn age 3.51 0.29 0.04 -0.03

9. Finn size 7.56 2.03 0.53 * 0.42 *

10. Finn profitability 0.03 0.15 0.21 * 0.14 *

11. Firm overall FDI 13.73 22.17 0.96 * 0.71 *
 experience

12. South Korean 4.35 7.72 0.11 * 0.09 *
 MNEs' host coun-
 try experience

13. South Korean FDI 0.10 0.15 0.00 0.02
 exit from the host
 country

14. Host country GDP 0.00 0.01 -0.01 0.00
 growth

15. Industry R&D 0.01 0.01 0.23 * 0.23 *
 intensitv

16. Industry advertise- 3.03 2.46 0.31 * 0.23 *
 ment intensity

 Variables 3 4 5 6

1. Home-proximate
 experience

2. Host-proximate
 experience

3. Double-distant
 experience

4. Host country spe- 0.51 *
 cific experience

5. South Korea-Host -0.21 * 0.27 *
 country cultural
 distance

6. Subsidiary 0.09 * 0.03 -0.01
 ownership (a)

7. Internationaliza- -0.24 * -0.17 * 0.15 * -0.12 *
 tion path (b)

8. Finn age -0.06 * 0.04 0.06 * 0.04

9. Finn size 0.59 * 0.34 * -0.04 0.06 *

10. Finn profitability 0.23 * 0.18 * 0.04 * 0.01

11. Firm overall FDI 0.98 * 0.64 * -0.11 * 0.06 *
 experience

12. South Korean -0.04 0.40 * 0.50 * -0.02
 MNEs' host coun-
 try experience

13. South Korean FDI -0.01 -0.03 0.05 * -0.05 *
 exit from the host
 country

14. Host country GDP 0.02 -0.01 -0.08 * -0.05 *
 growth

15. Industry R&D 0.29 * 0.18 * -0.04 * 0.21 *
 intensitv

16. Industry advertise- 0.34 * 0.24 * -0.09 * 0.1 *
 ment intensity

 Variables 7 8 9 10

1. Home-proximate
 experience

2. Host-proximate
 experience

3. Double-distant
 experience

4. Host country spe-
 cific experience

5. South Korea-Host
 country cultural
 distance

6. Subsidiary
 ownership (a)

7. Internationaliza-
 tion path (b)

8. Finn age 0.07 *

9. Finn size -0.31 * 0.04 *

10. Finn profitability -0.13 * -0.02 0.23 *

11. Firm overall FDI -0.21 * -0.02 0.59 * 0.23 *
 experience

12. South Korean 0.09 * 0.15 * -0.02 0.05 *
 MNEs' host coun-
 try experience

13. South Korean FDI 0.04 0.01 0.02 0.02
 exit from the host
 country

14. Host country GDP 0.05 * -0.03 -0.08 * -0.03
 growth

15. Industry R&D -0.12 * 0.3 * 0.26 * 0.11 *
 intensitv

16. Industry advertise- -0.16 * 0.05 * 0.21 * 0.09 *
 ment intensity

 Variables 11 12 13 14 15

1. Home-proximate
 experience

2. Host-proximate
 experience

3. Double-distant
 experience

4. Host country spe-
 cific experience

5. South Korea-Host
 country cultural
 distance

6. Subsidiary
 ownership (a)

7. Internationaliza-
 tion path (b)

8. Finn age

9. Finn size

10. Finn profitability

11. Firm overall FDI
 experience

12. South Korean 0.06 *
 MNEs' host coun-
 try experience

13. South Korean FDI 0.03 -0.01 0.02
 exit from the host
 country

14. Host country GDP -0.02 -0.04 0.05 *
 growth

15. Industry R&D 0.28 * 0.10 * -0.05 * -0.03
 intensity

16. Industry advertise- 0.34 * -0.06 * -0.05 * 0.02 0.04 *
 ment intensity

(n=2,392)

* p<0.05

(a) Wholly owned subsidiaries = 1

(b) An MNE started FDls from the cultural cluster of its
home country = 1

Table 2: Descriptive statistics and correlations for the subsample
of FDIs located NN within the home country's cultural cluster

 Variables Mean S. D. 1 2

1. Home-proximate 2.65 3.79
 experience

2. Double-distant 4.83 8.17 0.90 *
 experience

3 Host country 0.76 1.65 0.62 * 51
 specific experience

4. South Korea-Host 2.22 0.79 0.08 * 0 21 *
 country cultural
 distance

5. Subsidiary 0.47 0.50 -0.002 0.08 *
 ownership (a)

6. Internationalization 0.58 0.19 0.05 * 0.22 *
 path. (b)

7. Firm age 3.53 0.28 0.04 0.06 *

8 Firm size 6.83 1.88 0.53 * 0.59 *

9. Firm profitability 0.02 0.13 0.21 * 0.23 *

10. Firm overall FDI 8.24 11.83 0.96 * 0.98 *
 experience

11 South Korean 6.34 9.47 0.11 * -0.04 *
 MNEs' host coun-
 try experience

12 South Korean FDI 0.11 0.10 -0.001 -0.012
 exit from the host
 country

13. Host Country GDP 0.00 0.01 -0.01 0.02
 growth

14. Industry R&D 0.01 0.01 0.23 * 0.3 *
 intensity

15 Industry 7.13 4.02 0.31 * 0.34 *
 advertisement
 intensity

 Variables 3 4 5 6

1. Home-proximate
 experience

2. Double-distant
 experience

3 Host country
 specific experience

4. South Korea-Host 0.27 *
 country cultural
 distance

5. Subsidiary 0.029 0.004
 ownership (a)

6. Internationalization -0.10 * 0.10 * -0.07 *
 path. (b)

7. Firm age 0.04 0.07 * 0 04 * 0 11 *

8 Firm size 0.34 * 0.05 * 0.05 * -0.16 *

9. Firm profitability 0.17 * 0.04 * 0.006 -0.02

10. Firm overall FDI 0.64 * 0.11 * 0.06 * -0.17 *
 experience

11 South Korean 0.40 * 0.50 * -0.01 0.24 *
 MNEs' host coun-
 try experience

12 South Korean FDI -0.03 0.05 * -0.05 * 0.01
 exit from the host
 country

13. Host Country GDP 0.01 -0.08 * 0.05 * 0.01
 growth

14. Industry R&D 0.18 * -0.05 * 0.21 * 0.09 *
 intensity

15 Industry 0.24 * -0.09 * 0.09 * -0.15 *
 advertisement
 intensity

 Variables 7 8 9 10

1. Home-proximate
 experience

2. Double-distant
 experience

3 Host country
 specific experience

4. South Korea-Host
 country cultural
 distance

5. Subsidiary
 ownership (a)

6. Internationalization
 path. (b)

7. Firm age

8 Firm size 0.04

9. Firm profitability -0.02 0.14 *

10. Firm overall FDI -0.02 0.59 * 0.23 *
 experience

11 South Korean 0.15 * -0.03 0.05 * 0.05 *
 MNEs' host coun-
 try experience

12 South Korean FDI 0.01 0.02 0.02 0.01
 exit from the host
 country

13. Host Country GDP 0.02 0.08 * -0.03 0.02
 growth

14. Industry R&D 0.30 * 0.26 * 0.11 * 0.29 *
 intensity

15 Industry 0.04 * 0.22 * 0.09 * 0.34 *
 advertisement
 intensity

 Variables 11 12 13 14

1. Home-proximate
 experience

2. Double-distant
 experience

3 Host country
 specific experience

4. South Korea-Host
 country cultural
 distance

5. Subsidiary
 ownership (a)

6. Internationalization
 path. (b)

7. Firm age

8 Firm size

9. Firm profitability

10. Firm overall FDI
 experience

11 South Korean
 MNEs' host coun-
 try experience

12 South Korean FDI 0.02
 exit from the host
 country

13. Host Country GDP -0.04 0.05 *
 growth

14. Industry R&D 0.11 * 0.05 * -0.03
 intensity

15 Industry -0.07 * -0.05 * 0.02 0 04 *
 advertisement
 intensity

(n=2.729)

* p<0.05

(a) Wholly owned subsidiaries = 1

(b) An MNE started FDIs from the cultural cluster
of its home country = 1

Table 3: FDI experience and the mortality of FDIs located outside
the home country's cultural cluster (Robust standard errors in
parentheses. Industry dummies are not shown in the table. Likelihood
tests showed that compared to model 1, the variables added in
Model 3 (p<0.001). Model 5 (p<0.1). Model 7 (p<0.05), Model 8 (p<0.1).
and Model 10 (p<0.001) improve the overall fit of
the model significantly)

Variables Model 1 Model 2 Model 3

Home-proximate 0.06 0.08
experience (0.05) (0.04) *

(Home-proximate -0.004
experience) (2) (-0.00) ***

Host-proximate
experience

(Host-proximate
experience) (2)

Host country
specific experience

(Host country
experience) (2)

Double-distant
experience

(Double-distant
experience) (2)

South Korea-Host 0.00 0.01 0.02
country cultural (0.03) (0.03) (0.03)
distance

Subsidiary -0.16 -0.14 -0.13
ownership (a) (-0.14) (-0.14) (-0.13)

Internationalization 0.28 0.18 0.12
path (b) (0.24) (0.24) (0.25)

Firm age -0.49 -0.54 -0.58
 (-0.41) (-0.42) (-0.41)

Firm size 0.09 0.09 0.06
 (0.08) (0.07) (0.08)

Firm profitability 0.45 0.40 0.49
 (0.46) (0.43) (0.48)

Firm overall FDI 0.00 -0.02 -0.01
experience (c) (0.00 (-0.01) (-0.01)

South Korean MNEs' 0.00 0.00 0.00
host country (-0.04) (-0.04) (0.03)
experience

South Korean FDI -0.79 -0.80 -0.83
exit from the host (-0.42) * (-0.42)* (-0.43) *
country

Industry R&D -85.37 -80.37 -74.38
intensity (-20.4) *** (-20.7) *** (-19.6) ***

Industry -21.32 -19.57 -19.85
advertisement (-13.18) (-13.45) (-13.60
intensity

Host country 0.00 0.00 0.00
GDP growth (-0.02) (-0.02) (-(1.02)

Wald [chi square] 33750 *** 37214 *** 49985 ***

Variables Model 4 Model 5 Model 6

Home-proximate
experience

(Home-proximate
experience) (2)

Host-proximate 0.07 0.19
experience (0.02) *** (0.05) ***

(Host-proximate -0.02
experience) (2) (-0.01) ***

Host country 0.04
specific experience (0.04)

(Host country
experience) (2)

Double-distant
experience

(Double-distant
experience) (2)

South Korea-Host -0.02 -0.03 -0.01
country cultural (-0.04) (-0.04) (-0.04)
distance

Subsidiary -0.16 -0.18 -0.15
ownership (a) (-0.14) (-0.14) (-0.14)

Internalization 0.30 0.33 0.30
path (b) (0.24) (0.24) (0.25)

Firm age -0.54 -0.50 -0.52
 (-0.41) (-0.43) (-0.42)

Firm size 0.09 0.09 0.09
 (0.07) (0.07) (0.08)

Firm profitability 0.49 0.53 0.43
 (0.46) (0.47) (0.45)

Firm overall FDI 0.00 0.00 0.00
experience (c) (-0.00) (-0.00) (-0.00)

South Korean MNEs' 0.01 0.01 -0.02
host country (0.03) (0.03) (-0.04)
experience

South Korean FDI -0.79 -0.85 -0.78
exit from the host (-0.41) * (-0.40) ** (-0.42) *
country

Industry R&D -86.45 -85.14 -88.30
intensity (-20.9) *** (-19.9) *** (-22.1) ***

Industry -20.30 -18.33 -22.10
advertisement (-13.13) (-13.07) (-13.12) *
intensity

Host country 0.00 -0.01 -0.0
GDP growth (-0.02) (-0.02) (-0.02)

Wald [chi square] 33877 *** 37730 *** 41081 ***

Variables Model 7 Model 8

Home-proximate
experience

(Home-proximate
experience) (2)

Host-proximate
experience

(Host-proximate
experience) (2)

Host country 0.21
specific experience (0.07) ***

(Host country -0.03
experience) (2) (-0.01) ***

Double-distant -0.06
experience (-0.02) **

(Double-distant
experience) (2)

South Korea-Host -0.01 -0.03
country cultural (-0.04) (-0.04)
distance

Subsidiary -0.08 -0.13
ownership (a) (-0.14) (-0.15)

Internationalization 0.37 0.23
path (b) (0.26) (0.24)

Firm age -0.54 -0.62
 (-0.43) (-0.43)

Firm size 0.08 0.08
 (0.08) (0.07)

Firm profitability 0.40 0.41
 (0.46) (0.44)

Firm overall FDI 0.00 0.03
experience (c) (0.00) (0.01) **

South Korean MNEs' -0.03 -0.02
host country (-0.04) (-0.03)
experience

South Korean FDI -0.83 -0.78
exit from the host (-0.41) ** (-0.-31) *
country

Industry R&D -79.37 -85.28
intensity (-20.4) *** (-23.1) ***

Industry -19.80 -19.73
advertisement (-13.22) (-13.36)
intensity

Host country -0.01 -0.01
GDP growth (-0.02) (-0.02)

Wald [chi square] 36040 *** 40507 ***

Variables Model 9 Model l0

Home-proximate 0.06
experience (0.03) *

(Home-proximate -0.003
experience) (2) (-0.00) **

Host-proximate 0.16
experience (0.04) ***

(Host-proximate -0.02
experience) (2) (-0.01) ***

Host country 0.15
specific experience (0.08) *

(Host country -0.01
experience) (2) (-0.01) *

Double-distant -0.05 -0.01
experience (-0.02) ** (-0.01)

(Double-distant 0.00
experience) (2) (-0.00)

South Korea-Host -0.02 -0.03
country cultural (-0.04) (-0.04)
distance

Subsidiary -0.12 -0.09
ownership (a) (-0.14) (-0.14)

Internationalization 0.24 0.26
path (b) (0.24) (0.26)

Firm age -0.63 -0.66
 (-0.44) (-0.45)

Firm size 0.07 0.04
 (0.07) (0.07)

Firm profitability 0.40 0.52
 (0.44) (0.47)

Firm overall FDI 0.03
experience (c) (0.01) **

South Korean MNEs' -0.02 -0.02
host country (-0.03) (-0.04)
experience

South Korean FDI -0.78 -0.89
exit from the host (-0.42) * (-0.41) **
country

Industry R&D -84.7 -75.89
intensity (-22.3) *** (-19.7) ***

Industry -19.39 -17.01
advertisement (-13.42) (-13.51)
intensity

Host country -0.01 -0.01
GDP growth (-0.02) (-0.02)

Wald [chi square] 51575 *** -37135 ***

No. of observations = 2.392

No. of FDIs=493

No. of failures=264

* p<0.1: ** p<0.05: *** p<0.01

(a) Wholly owned subsidiaries = 2

(b) An MINE started FDIs from the cultural cluster of its home
country = 1

(c) Firm overall FDI was excluded in the final model because
it is the sum of the four types of experiences included in the model

Table 4: FDI experience and the mortality of FDIs located within
the home country's cultural cluster (Robust standard errors in
parentheses. Industry dummies are not shown in the table)

Variables Model 1 Model 2

Double-proximate experience -0.09
 (-0.03) ***

Host country specific
experience

Double-distant experience

South Korea-Host country 0.02 0.01
cultural distance (0.08) (0.08)

Subsidiary ownership (a) -0.27 -0.27
 (-0.15) * (-0.15) *

Internationalization path (b) 0.23 0.30
 (0.21) (0.21)

Firm age -1.22 -1.24
 (-0.36) *** (-0.34) ***

Firm size 0.15 (1.14
 (0.07) ** (0.07) **

Firm profitability -0.56 -0.56
 (-0.09) *** (-0.09) ***

Firm overall FDI experience (c) 0.00 0.04
 (0.01) (0.(11) ***

South Korean MNEs' host 0.09 (1.05
Country experience (0.05) * (0.05)

South Korean FDI exit from -0.23 -0.15
the host country (-0.82) (-0.84)

Industry R&D intensity -29.23 -31.12
 (-17.91) (-18.28) *

Industry advertisement 30.80 29.93
intensity (18.83) (19.12)

Host country GDP growth 0.00 0.00
 (0.02) (-0.02)

Wald [chi square] 32423 *** 43902 ***

Variables Model 3 Model 4

Double-proximate experience

Host country specific 0.10
experience (0.05) **

Double-distant experience 0.05
 (0.03)

South Korea-Host country 0.00 0.02
cultural distance (0.07) (0.08)

Subsidiary ownership (a) -0.27 -0.27
 (-0.15) * (-0.15) *

Internationalization path (b) 0.23 0.27
 (0.21) (0.21)

Firm age -1.22 -1.23
 (-0.36) *** (-0.35) ***

Firm size 0.16 0.14
 (0.07) ** (0.07) **

Firm profitability -0.56 -0.56
 (-0.09) *** (-0.09) ***

Firm overall FDI experience (c) -0.01 -0.02
 (-0.01) (-0.02)

South Korean MNEs' host 0.06 0.08
Country experience (0.05) (0.05)

South Korean FDI exit from -0.24 -0.19
the host country (-0.85) (-0.81)

Industry R&D intensity -28.06 -30.68
 (-17.80) (-18.18) *

Industry advertisement 29.27 30.98
intensity (19.14) (18.91)

Host country GDP growth 0.00 0.00
 (-0.02) (0.02)

Wald [chi square] 46843 *** 31575 ***

Variables Model 5

Double-proximate experience -0.05
 (-0.02) **

Host country specific 0.07
experience (0.05)

Double-distant experience 0.03
 (0.02)

South Korea-Host country 0.00
cultural distance (0.07)

Subsidiary ownership (a) -0.27
 (-0.15) *

Internationalization path (b) 0.28
 (0.22)

Firm age -1.24
 (-0.35) ***

Firm size 0.15
 (0.07) **

Firm profitability -0.56
 (-0.09) ***

Firm overall FDI experience (c)

South Korean MNEs' host 0.05
Country experience (0.05)

South Korean FDI exit from -0.17
the host country (-0.85)

Industry R&D intensity -30.19
 (-18.22) *

Industry advertisement 29.47
intensity (19.26)

Host country GDP growth 0.00
 (-0.02)

Wald [chi square] 42440 ***

No. of observations = 2.729

No. of subjects = 527

No. of failures = 284

* p<0.1: ** p<0.05: *** p<0.01

(a) Wholly owned subsidiaries=l

(b) An MINE started FDIs from the cultural cluster of its home
country cluster = 1 Firm overall FDI was excluded in the final
model because it is the sum of the three types of
experiences included in the model

Fig. 2: National cultural difference and the grouping of an MNE's
prior FDI experience ((1) previous-FDI-subsequent-FDI cultural
difference refers to the differences between the host countries of
the MNE's previous FDIs and the host country of the focal subsequent
FDI. Previous-FDI-home-country cultural difference refers to the
differences between the MNE's home country and the host countries
of its previous FDIs.)

Previous-FDI-
Subsequent-FDI
Cultural Difference

Large

 (2) (3)
 Home-Proximate Experience Double-Distant Experience

* Possible when a focal subsequent * Possible regardless the location
KDI Is located in a country of a focal subsequent FDI.
culturally dissimilar to the home
country * Example: for a Korean MNE
 initiating an FDI in the U.S.,
* Example: for a Korean MNR this experience includes the MNE's
initiating an FDI in the U.S.. previous FDIs in countries
this experience includes the MNE's culturally dissimilar to both
previous FDIs in countries the U.S. and South Korea (e.g.
culturally similar to South Korea Germany, Denmark, etc.).
but dissimilar to the U.S. (e.g..
China. Vietnam, etc.).

Small

 (1) (4)
 Double-Proximate Experience Host-Proximate Experience

* Possible when a focal subsequent * Possible when a focal
FDI is in a country culturally subsequent FDI is located in a
similar to the home country. country culturally dissimilar to
 the home country.
* Example: for a Korean MNH
initiating an FDI in China, this * Example: for a Korean MNE
experience includes previous FDIs initiating an FDI in the U.S.,
in countries similar to both this experience includes the MNE's
Korea and China (e.g. Vietnam. previous FDIs in countries
Singapore, etc.). culturally similar to the U.S.
 but dissimilar to South Korea
 (e.g. the UK. Canada, etc.)

Small Large

Previous-FDI-Home-Country
Cultural Difference
COPYRIGHT 2013 Gabler Verlag
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2013 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:RESEARCH ARTICLE
Author:Zeng, Yuping; Shenkar, Oded; Song, Sangcheol; Lee, Seung-Hyun
Publication:Management International Review
Article Type:Report
Geographic Code:9SOUT
Date:May 1, 2013
Words:16302
Previous Article:Rigor and relevance of IJV exit research.
Next Article:Erratum to: focused issue on indigenous management research.
Topics:

Terms of use | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters