SUBSIDIARY EMBEDDEDNESS, EXPECTED PERFORMANCE AND MNC COMPETENCE DEVELOPMENT

Ulf Andersson*
Department of Business Studies,
Uppsala University


Mats Forsgren
Institute of International Business,
Stockholm School of Economics

Ulf Holm
Department of Business Studies,
Uppsala University and
Institute of International Business,
Stockholm School of Economics

*Corresponding author. Ulf Andersson,
Department of Business Studies,
Uppsala University,
Box 513, 751 20 Uppsala, Sweden.

Telephone: +46 18 4711614.

E-mail: ulf.andersson@fek.uu.se

 

THE DIFFERENTIATED MNC

A usual conceptualization of the MNC today is that its competitive advantage is contingent on its different subsidiaries with their different business contexts (Ghoshal 1986, Goshal and Nohria 1989, Ghoshal and Bartlett 1990, Gupta and Govindarajan 1991, 1994, Rosenzweig and Singh 1991, Malnight 1996). These contexts differ because the subsidiaries operate in different local environments with different resources, business and legal systems, business networks etc. The MNC competitive advantage is based 1) on the breadth in the resources, experiences and capabilities of the subsidiaries as a group and 2) on the ability of the MNC to use the capability of one unit in other units' operations and vice versa. Diversity and transfer of capabilities is therefore at the heart of the differentiated MNC and constitutes its competitive advantage vis-à-vis domestic firms, or firms with less diversity in terms of subsidiary business contexts.

A basic feature of the differentiated MNC is that subsidiaries acquire knowledge from its external environment. This knowledge is later on institutionalized in the subsidiary and will constitute important building blocs for the capability of "how to do things". The ability of an organization to recognize the value of new, external, information has sometimes been labeled absorptive capacity (Cohen and Levinthal 1990). In the differentiated MNC the level of absorptive capacity at each subsidiary is crucial for the long-term competitive advantage of the whole company. Based on this capacity the MNC can be at the forefront in terms of product development, adaptation to changes in market structures, etc. This perspective on the importance of absorptive capacity is in line with recent trends in new FDI theory in which foreign investment is made also to acquire knowledge and not only to exploit knowledge (Cantwell 1990, Dunning 1996, Almeida 1996). Investments abroad are often carried out because the MNC wants to selectively tap knowledge linked to a specific local business context. The possibility to do so is dependent on the local units’ absorptive capacity.

Another feature of the differentiated MNC is that differences in absorptive capacity between subsidiaries have a profound impact on the strategic role of subsidiaries within the corporation. A subsidiary having high competence in identifying and assimilating new, external information will also be more important in the strategy process within the multinational firm than subsidiaries that have less of such competence (Birkenshaw and Morrison 1996, Taggart 1998, Andersson and Forsgren 1999).

A clarification of a subsidiary's absorptive capacity is needed. First, it is not simply the sum of the absorptive capacities of its employees, and we can therefore talk about an organization's absorptive capacity, similar to the distinction between individual learning and organizational learning. Secondly, absorptive capacity includes not only identification and assimilation of new information, but also the organization's ability to exploit or commercialize it. Therefore, absorptive capacity is dependent on both the subsidiary's direct interfaces with its environment and its ability to transfer and use the information inside the organization (Cohen and Levinthal 1990).

In the differentiated MNC a focused subsidiary's capability to absorb new knowledge from the environment influence the MNCs competitive advantage in mainly two different ways. First, there is reason to assume that it will strengthen the subsidiary's competitive strength in its own market place (Powell et al. 1996, Bierly and Chakrabarti 1996), and therefore, indirectly strengthen the whole corporation. Secondly, through transfer of externally acquired knowledge from the focused subsidiary to other units, the competence of the whole MNC will be upgraded (Zander and Kogut 1995, Szulanski 1995,1996, Hansen 1999).

This paper focuses on factors behind a subsidiary's absorptive capacity and the relationship between this capacity and its expected market performance and its importance for competence development in other MNC units, respectively. MNC competence development is here defined as development of products and production processes. In the next section the concept of embeddedness is introduced and its importance for absorptive capacity is discussed. In the following sections hypotheses are formulated concerning the relationship between embeddedness and expected subsidiary market performance and subsidiary importance for competence development. In the following sections the hypotheses are tested empirically by applying data from Swedish multinationals in a LISREL model. A final section contains a discussion of the results from the empirical testing and possible tracks for future research about the differentiated MNC.

DETERMINANTS OF ABSORPTIVE CAPACITY AT THE SUBSIDIARY LEVEL

What determines a subsidiary's ability to identify and assimilate new information from its environment?

There seems to be two, related, lines of reasoning about an organization's absorptive capacity. First, research about cognition indicates that learning is dependent on existing knowledge. The ability to identify and assimilate is dependent on the richness of the pre-existing knowledge. Learning is cumulative, and learning is greatest when the object of learning is related to what is already known (Cohen and Levinthal 1990, Mowery et al 1996). Acquisition of new information is not only dependent on the existence of the information as such, but also on the subsidiary's knowledge about areas or problems similar to, or related to, the new information.

Second, implicit in the conceptualization of absorptive capacity is the assumption that the organization learns from someone else, usually another organization. A subsidiary's absorptive capacity is therefore dependent on its involvement in inter-organizational learning (Lane and Lubatkin 1998, Mowery et al 1996, Hansen 1999, Kumar and Nti 1998). Consequently, an organization has not an equal capacity to learn from all organizations. A subsidiary's possibility to identify new information in another organization, and the ability to assimilate that information into the subsidiary, is heavily dependent on the structure of the existing dyadic relationship between the subsidiary and the other organization (Lane and Lubatkin op.cit.).

These two lines of reasoning lead to two statements regarding the determinants of a subsidiary's absorptive capacity. First, pre-existing knowledge at the subsidiary level is crucial for identification and assimilation of new knowledge. Second, the identification and assimilation of new knowledge is mainly carried out within established relationships with other organizations. The structure and nature of these relationships influence the level of subsidiary absorptive capacity.

The discussion of embeddedness in business networks has much in common with the reasoning above (Polyani 1957, Granovetter, 1985, 1992, Zukin and Di Maggio 1990, Uzzi 1997). Most important is the emphasis on the network of interorganizational relationships as an important part of every organization's business life. The network constitutes an organization's most important interface with the environment. It is mainly through these relationships that the exchange of information with the environment is handled. Therefore, a subsidiary's relationships can be looked upon as an investment that will influence its future capacity to identify and assimilate new information. The concept of embeddedness, as it has been used by several scholars, (Grabher 1993, Uzzi 1997, Andersson and Forsgren 1996, 1999 etc) combines the two notions above; the importance of prior knowledge and the importance of relationships. Embeddedness in networks is assumed to develop over time, from more arm’s-length relationships to relationships based on adaptation and trust (Johanson and Mattsson 1987, Håkansson 1989, Ford 1997,Uzzi 1997). This is also in line with Cohen and Levinthal's notion that absorptive capacity is something that develops over time, is path dependent and therefore builds on prior knowledge of another organization's capabilities (Cohen and Levinthal 1990).

Knowledge about other organizations’ capabilities is therefore dependent on the degree of embeddedness. The higher the degree of embeddedness, the greater the pre-existing knowledge about environmental factors of importance for the absorptive capacity. And the higher the degree of embeddedness, the higher is also the similarities between the focal organization and its counterparts in terms of knowledge bases and dominant logic (Lane and Lubatkin 1998). Therefore, embeddedness has a positive impact on both pre-existing knowledge and knowledge congruence between partners in the relationships, which in their turn both have a positive impact on an organization's absorptive capacity (Mowerty et al 1996).

This conclusion is of course also relevant for subsidiaries belonging to MNCs. The more the focal subsidiary has deep, extensive relationships with other organizations in its business environment the higher its ability to identify and assimilate new information from that environment. The depth of the relationship has to do with factors such as mutual adaptation of resources and activities as well as the existence of trust in the relationship. The breadth of the relationship has to do with the number of areas or functions in which the counterparts are related.

The depth and the breadth of an interorganizational relationship can vary depending on the type of embeddedness. For instance, a subsidiary can have old and mutual relationships with a certain customer or supplier in terms of e.g. sales or purchase but not in terms of R&D or other activities. In the following we will make a distinction between business embeddedness and technical embeddedness. The former type of embeddedness is related to the subsidiary's general business activities. It reflects to what extent the subsidiary's relationship with a certain external organization deviates from arms-length relationships in its ordinary business operations (Uzzi 1997). The higher the degree of business embeddedness the less inclined is the subsidiary (and its counterpart) to change partners even if economic conditions change. The partners have adapted their behavior to one another in terms of business routines, information exchange, planning systems etc. The interface between the subsidiary and the other organization is also broad in terms of number of functional areas and people involved.

Technical embeddedness is related to one specific functional area, development of products and production processes. It reflects the fact that some of the counterparts in the subsidiary's business network are more important than other counterparts for the subsidiary's technical development. A high degree of technical embeddedness means that the two organizations are highly interdependent in terms of technical development activities. For instance, the development of new products in the subsidiary is based on intensive information exchange with another organization, and the development activities is adapted to similar processes in the latter organization.

Through a subsidiary's business embeddedness technical embeddedness evolve. By exchanging with a counterpart new possibilities to use, for example, a product may emerge. This can in turn lead to common development projects and adaptations in production processes. Therefore, to the extent that a subsidiary is deeply involved in a system of buyer and seller relationships which deviates from ordinary arm’s-length relations, business embeddedness can be seen as an antecedent to technical embeddedness. This reasoning lead us to formulate the following hypothesis:

H1: A subsidiary's degree of business embeddedness is positively related to its technical embeddedness.

In the following two sections we will discuss how embeddedness influences the subsidiary performance in its own marketplace and its role as a contributor to competence development in other corporate units, respectively.

EMBEDDEDNESS AND MARKET PERFORMANCE

It is a common thought that an organization's performance is contingent on its ability to obtain resources from its environment. For instance, resource dependence theory stresses the relationship between performance and the ability to cope with strategic interdependencies in the environment (Pfeffer and Salancik 1978). In contingency theory, survival and success are dependent on the organization's responses to diverse environments (Lawrence and Lorch 1967, Stopford and Wells 1972, Egelhoff 1988). The importance of the ability to obtain resources from the environment is also apparent in theories dealing with intra-organizational power (Crozier 1964, Hickson et al 1971, Provan et al 1980, Pfeffer 1981). Theories focusing on geography in an organizational context also emphasize the importance of the firm's ability to selectively tap the environment of knowledge (Piore and Sabel 1984, Amin and Thrift 1994, Porter 1990, Sölvell and Zander 1995).

Some scholars have more directly related performance to the ability to absorb new knowledge from the environment through the network of specific inter-organizational relationships. For instance, Powell et al. (1996) found empirical support for a positive relationship between rates of growth and network- relationships among biotechnology firms. In a study of 23 entrepreneurial firms it has been argued that organizational performance increases with the use of embedded ties with network partners (Uzzi 1997). Zaheer et al. (1998) found a positive relationship between interorganizational trust and performance in a sample of 107 buyer-supplier interfirm relationships in the electrical equipment manufacturing industry.

Becoming embedded in business relationships is a long-term process characterized by mutual investments leading to different kinds of adaptations by the interacting parties. This does not only involve level and types of specific technical investments, but also the development of systems for carrying out business exchange efficiently. Within the rationale of a business actor - the subsidiary – of becoming embedded in a system of business relationships, lies an expectation of profitability. As shown by Blankenburg-Holm et al (1997), mutual dependence and commitment in a setting of business relationships, lead to value creation. Thus, we can expect that present embeddedness relates to future prospects. If goal attainment is at the heart of a subsidiary’s market performance we should also maintain that it is the expected market performance rather than the present market performance, which should be assessed (Anderson, 1990). Thus expected performance, rather than present performance, is focused in this paper. This probably means that perceptual measures provide a better yardstick than objective ones since performance should reflect the corporate managers’ view on the subsidiary’s goal attainment. Further, perceptions (of subsidiary expected performance) can be considered "reality" (Kelly and Thibaut 1978) and are therefore likely to reflect corporate managers’ view and behavior in the relationship with the subsidiary.

If we apply this reasoning to the context of the differentiated MNC we will argue that the expected performance of the individual subsidiary is positively related to the degree of embeddedness in its business relationships. This conclusion is relevant for both types of embeddedness. Business embeddedness can be expected to improve the subsidiary's possibility to obtain new information about changing market conditions, etc., which will improve its market performance in the future. Further, if we assume that development of new products and production processes are key factors for economic growth and competitive advantage (Mansfield 1968, Bartlett and Ghoshal 1990, Chesnais 1986, Dosi et al 1988, De Mayer 1992) we can conclude that technical embeddedness is positively related to the subsidiary's expected market performance. The following hypotheses can therefore be formulated:

H2: A subsidiary's degree of business embeddedness is positively related to its expected market performance.

H3: A subsidiary's degree of technical embeddedness is positively related to its expected market performance.


SUBSIDIARY EMBEDDEDNESS AND COMPETENCE DEVELOPMENT
WITHIN THE MNC

In the differentiated MNC the subsidiaries have important, but different, roles in the competence development, and therefore competitive advantage, of the whole MNC. Even though the performance of the individual subsidiary in its own marketplace also contribute to the performance of the whole MNC, a crucial question is to what extent the subsidiary contributes to the competence development in other corporate units through knowledge transfer.

On a general level it is reasonable to argue that the higher the performance of the individual subsidiary the higher the probability that it also plays a significant role in developing other MNC units' competence. High performance will attract sister units' interest and work as an incentive to learn from that subsidiary, for instance considering product development, marketing efforts etc. Consequently, the following hypothesis can be formulated:

H4: The market performance of an individual subsidiary is positively related to its importance for competence development in other MNC units.

It is often argued that an organization develops a common understanding by which to transfer competence from ideas to production and products (Kogut and Zander 1995). Transfer of knowledge within organizations is therefore easier to accomplish than transfer of knowledge between organizations. Consequently, one of the main competitive advantages with MNCs is the possibility to transfer knowledge between subsidiaries, located in different countries and business contexts. We would therefore expect that within an MNC, knowledge flows from subsidiaries with high competence to subsidiaries with lower competence.

We have argued above that network embeddedness is a driving force for the subsidiary’s competence development, because the higher the degree of embeddedness the higher the absorptive capacity of the subsidiary. High absorptive capacity lead to high competence development in a subsidiary, and consequently a transfer of knowledge from that subsidiary to other subsidiaries. Embeddedness is therefore one important underlying variable that explains which of the subsidiaries will be givers, and which will be receivers.

Although this reasoning could be applicable to both business embeddedness and technical embeddedness, the general character of the former is probably of less relevance for influencing competence development in terms of product and production process development for other units inside the MNC. If we would have considered other dimensions of competence development such as market knowledge, sales techniques or other "non-technical" competencies, the influence of business embeddedness on sister units’ competence development would have been appropriate. Consequently, the following hypothesis can be formulated:

H5a: A subsidiary's technical embeddedness is positively related to other MNC units' competence development.

However, several authors have pointed out that knowledge transfer within organizations is difficult to accomplish. The reason for this is connected to the knowledge itself as well as to characteristics of the sender, the recipient and the relationship between them. For instance it has been argued that idiosyncratic, specific, tacit, and/or non-codified knowledge is difficult to transfer from one unit to another, due to problem of separating such knowledge from the unit that carries the knowledge, and add it to another actor's knowledge base. (Grant 1996, Jensen and Meckling 1992, Kogut and Zander 1992, 1995, Hansen 1999, Spender 1996, Szulanski 1995, 1996) Other problems have been related to the recipients' ability or willingness to absorb new information (Cohen and Levinthal 1990, Szulanski 1996, Allen 1977, Hayes and Clark 1985, Porter 1985), or to the willingness of a unit to share information with other units (Porter 1985, Szulanski 1996, Forsgren 1997).

In a differentiated MNC one source of barriers to knowledge transfer are especially interesting. The complex, idiosyncratic, interaction processes between the subsidiary and its counterparts create a competence that cannot easily be used in other corporate units' business contexts. Knowledge development is context-specific, or even relation specific. This specificity is positively correlated with the ability to create new knowledge. For instance, an extensive, long term co-operation with a specific customer or supplier will improve the subsidiary's problem-solving capacity, and its ability to create new knowledge. But the more context-specific the solutions are, the more difficult it will be apply the knowledge in another corporate unit's business context. Consequently, if we assume that subsidiary network embeddedness is positively related to competence development, but also to context specificity, we can also argue that there is a trade off between embeddedness and the possibility to transfer knowledge to other corporate units.

One can argue that the context-specificity problem is reflected in the degree of embeddedness of the subsidiary's network. The more the subsidiary is engaged in deep and intensive interaction with specific counterparts in its business context the higher the context specificity of the relationships, and the more resources are deployed in relation-specific activities. It is reasonable to assume that this trade off problem is especially relevant for the subsidiary’s development activities. For instance, the more the local subsidiary is involved in product development cooperation with specific external customers or suppliers the more context specific are these activities.

Therefore, the following alternative hypothesis can be formulated:

H5b: A subsidiary's degree of technical embeddedness is negatively related to its importance for other MNC units' competence development.

 

THE SAMPLE

Data have been collected from 97 subsidiaries of 20 Swedish MNC divisions with the divisional management positioned in Sweden. The subsidiaries are situated in Europe and North America, 92 and 5 respectively. The sample represents a wide spectra of Swedish industry and includes large and well known companies in industries such as gas applications, hard material tools, industrial equipment, management training, petrochemicals, power distribution, pulp and paper, software and telecommunications equipment. The subsidiaries included are among the largest and most important in their respective divisions. On average the subsidiaries in the sample account for over 50 per cent of the divisions’ combined operations, measured in terms of number of employees. In 25 per cent of the divisions, the investigated subsidiaries account for more than 80 per cent of the divisions’ total operations, whilst they account for between 10 and 60 per cent in the remaining divisions. The size of the subsidiaries varies between 50 and over 5,000 employees.

In conjunction with the divisions’ headquarters we have selected subsidiaries that can be considered representative for the divisions’ business activities, in order to increase the possibility for general conclusions from the data gathered. The average number of employees in the divisions was 5850 with a variation between 300 to more than 27,000. Turnover ranged from 0.6 to 23 billion SEK with an average of about 6 billion SEK. All divisions are highly international with, on average, more than 50 per cent of their employees outside Sweden. Taken together, the divisions had more than 100,000 employees and an annual turnover exceeding 100 billion SEK.

The investigated subsidiaries have their own production and sales. The development of products and of the production process is therefore an important activity in every subsidiary. The subsidiaries studied have a mixture of business relationships with both external and corporate suppliers and customers.

DATA GATHERING

To increase validity and reliability in the assessment of the subsidiaries’ relationships with customers and suppliers interviews have been made with two different managers in each subsidiary, the sales manager and the manager responsible for purchasing. They were asked to describe and assess the three most important relationships with customers and suppliers respectively. The managers interviewed were asked to characterize the relationships in focus according to a standardized questionnaire.

After interviewing the subsidiary managers in one division, we turned to the headquarters and made a personal interview with the divisional manager, based on the same type of standardized questionnaire. Through these interviews we gathered information about the headquarters view of each subsidiary’s future market performance and its importance for sister units’ development. The headquarters knowledge about the subsidiary’s specific business relationships was also checked. This can be seen as a further validation of the questionnaire and increased reliability of the respondents’ answers. The study involved personal interviews with more than 200 managers from leading positions in both the subsidiaries and the divisional headquarters. Each personal interview lasted for about one and a half-hours, with problems with concepts and interpretations in the questionnaire being discussed and explained. The discussion time and possibilities to further explain the specific questions clearly improves the reliability of the data collected compared with for example a mail survey.

A significant feature of this research is that the global divisional managers judge the two dependent variables, expected market performance and importance for competence development, while the subsidiary managers assess the two independent variables business- and technical embeddedness. This design means that we avoid the potential bias that would arise in the data if the same person estimated both the dependent and the independent variables. The idea behind separating the assessment is also to get a more reliable measure of subsidiary expected performance and importance for competence development by asking somebody outside the subsidiary in question to evaluate this. There is reason to believe that the divisional managers are in a better position to judge both the subsidiary’s performance in the market place, and its relative importance in the division’s competence development, than are the subsidiaries themselves.

CONSTRUCT ANALYSIS

The hypothesized model in (Figure 1) was empirically tested in a LISREL model (Figure 2). The validity of LISREL models is estimated by the validity of the entire model, i.e. the nomological validity. But before estimating the nomological validity of the model, with the specified causal relations, it is important to judge the homogeneity of the constructs included in the model, i.e. convergent validity, and the discriminant validity, that is, to what extent the constructs are separated from each other. First, we will describe the operationalization of the included constructs, then we will evaluate the different forms of validity.

Expected Market Performance

Expected performance, rather than present performance, is focused in this paper. As, first of all, market performance should reflect the corporate managers’ view on the subsidiary’s goal attainment, perceptual measures provide a better yardstick than objective ones. Therefore in our analysis we use the subsidiary’s market performance in terms of expected profitability, sales and market shares, as perceived by divisional managers. Apart from the advantage of separating the responses relating to technical- and business embeddedness from those concerning performance, the measure also has the virtue of giving the divisional headquarters the possibility to make comparisons across subsidiaries and countries. A 5-point Likert type scale (1 = very small to 5 = very high) has been used to separate the answers (see Table 1).

Subsidiary Importance for Competence Development in the MNC

To receive valid and reliable measures of the subsidiary importance for competence development we have used indicators reflecting the subsidiary’s importance for technical development in terms of product and production development for other sister units. We have chosen to use the divisional management’s recognition of to what extent each subsidiary is important to its sister units’ product and production development. One benefit of using the divisional management’s responses to these questions is that it separates them from the responses relating to embeddedness. It also gives the divisional management possibilities to compare different subsidiaries contribution to competence development. A 5-point Likert type scale (1 = very small to 5 = very high) has been used.

Business Embeddedness

Business embeddedness should mirror the subsidiary’s capacity to understand changing business conditions and its ability to adapt to these conditions through its business relationships. Business embeddedness is related to the subsidiary’s general business activities, and should resonate to what extent the subsidiary’s relationships with external suppliers and customers depart from arm’s-length relations in its ordinary business operations (Uzzi 1997). Therefor we need sound indicators of how embedded the subsidiary is in its relationships in terms of how they have adapted their business behavior and the breadth of its contact pattern with its external business partners. We have asked the subsidiary sales manager and the manager responsible for purchasing in the subsidiary to estimate to what extent the subsidiary have adapted their way of doing business because of the relationships to their most important external customers and suppliers. The responses will indicate the degree of external business embeddedness in these relationships i.e. the relationship’s interchangeability. We have also asked them to assess the number of different functional areas that are involved in direct contacts with the customers and suppliers. The higher the number of functional areas involved the greater the investment in the relationship, and also, the higher the possibility to absorb new knowledge about the general business conditions. Further the greater the possibility will be to understand in what specific areas, e.g. marketing, purchasing, product development etc., increased involvement and adaptation can be commercially viable.

 

Technical Embeddedness

Technical embeddedness should reflect the value of a business relationship in terms of the subsidiary’s capacity to absorb new technology. It is often argued that development of technology is reflected, above all, in a company’s development of new products and/or production processes (see, e.g., Mansfield 1968). We have therefore chosen the development of new products and new production processes as our two indicators of external technical development. The subsidiary sales- and purchasing managers have been asked to assess to what extent a specific external customer- or supplier relationship, respectively, has caused adaptation of the subsidiary’s product- and production development. A 5-point Likert- type scale from 1=not at all to 5=very much has been used for each indicator. By simply adding the scores of each of the subsidiary’s relationships and then dividing the score by the number of relationships the subsidiary have, two indicators have been created reflecting the average technical embeddedness of the subsidiary’s external network. It should also be pointed out that the emphasis has been on the subsidiary’s most important product or group of products in the interviews with the subsidiary managers. This means that all questions about business relationships, adaptation, product development and production development refers to a specific product/market area rather than to the subsidiary’s total activity. This will certainly increase the relevance of our indicators and also improve the reliability of the answers given by the subsidiary managers.

In Table 1 below, convergent validity is judged by the R2-values, measuring the strength of the linear relationships, the t-values, a significance test of each relationships in the model, and the factor loading for each indicator (Jöreskog and Sörbom 1993). The constructs in this LISREL model all have good convergent validity i.e., are homogenous constructs. As can be seen in Table 1 all R2-values are above 0.20 showing strong linear relations between the construct and the indicator ( Jöreskog and Sörbom 1993). The t-values for each indicator are highly significant (lowest t-value 3.36) and the factor loadings are strong.

To assess discriminant validity a measurement model is created, testing that not more than one construct measures a single indicator. Our set of latent constructs has high discriminant validity as key statistical estimates show lack of uni-dimensionality. In one case though, the variation of one of the indicators, increase in sales volume, which is indicated by the construct Expected Market Performance, is to a limited degree also explained by the latent construct Importance for Competence Development. This affects the discriminant validity of these two constructs negatively. However, between the two constructs, which is built up by five indicators, only one indicator has this problem. Further, as the convergent validity is good and each of the indicators assigned to the construct Importance for Competence Development has a stronger relationship with the construct than the indicator increase in sales volume has, we will proceed with the analysis without omitting the indicator increase in sales volume. Moreover, the indicator is the strongest one in the construct Expected Market Performance.

The second step in the analytical progress is to form the structural model by specifying the causal relations in accordance with the five hypotheses. In the following we test single causal relations with t-values and factor loadings between the constructs in the model. We assess the entire model by chi-squares and degrees of freedom and a probability estimate (p-value), which is a test of a non-significant distance between data and model, i.e. nomological validity (Jöreskog and Sörbom, 1993).

RESULTS

Through repeated iterations a LISREL analysis proceeds with the model being further fine-tuned and a more coherent representation of the empirical data. For our model, the first step is to test all five hypothesized causal relationships simultaneously. Although the overall model is significant (Chi2 =21.54 (df=22), p= 0.49), the results reveals only two significant paths between the four constructs, the one between business embeddedness and technical embeddedness and between technical embeddedness and importance for competence development. Thus, the overall model is not good. However, the statistics indicated that eliminating some insignificant paths could improve the model. First we omitted the most insignificant relation, i.e. between business embeddedness and expected market performance, (H. 3), which had a factor loading of -0.02 and a t-value of -0.07. Therefore, in the next step we tested hypotheses 1, 2, 4 and 5. The statistical output for the overall model was again significant (Chi2(23) =21.54, p=0.55). However, the relation between expected market performance and importance for competence development was still unchanged and insignificant, which led to another re-estimation of the model after excluding this relation.

In the final estimation, which included a test for confirmation of hypotheses 1, 4 and 5, the statistical output revealed a significant fit for the overall model (Chi2(24) =23.30, p=0.50). The paths between the constructs were also significant: business embeddedness and technical embeddedness (factor loading = 0.83, t=2.54), technical embeddedness and market performance (factor loading =0.27, t=2.00), technical embeddedness and importance for competence development (factor loading =0.46, t=2.55).

The resulting model is based on three relations. The first is a significant positive relation from subsidiary business embeddedness to subsidiary technical embeddedness, thus supporting hypothesis 1. Second, there is a significant and positive relation from subsidiary technical embeddedness to subsidiary performance. Thus, hypothesis 4 is supported. Hypothesis 3 - the relation between subsidiary business embeddedness and expected subsidiary performance - is not supported. Among the two competing hypotheses, 5a and 5b, hypothesis 5a is supported, as there is a positive and significant relation between subsidiary technical embeddedness and competence development. Finally, hypothesis 2 is not supported, though the relation between expected subsidiary performance and importance for competence development is positive and has a relatively strong t-value (1.53). This may have to do with the two constructs not being totally discriminant, which means the result regarding hypothesis 2 is somewhat tentative.

The overall result indicates that technical embeddedness is a relevant predictor for a subsidiary's market performance, as well as for its role as a contributor to competence development in other MNC units. In line with hypothesis 4 technical embeddedness seems to have a positive and significant influence on the subsidiary's expected market performance. The data support the view that close relationships with external customers and suppliers in terms of product and production process development have a positive impact on the subsidiary's success in its own market place. To the best of our knowledge this is the first time it has been possible to demonstrate empirically that an MNC subsidiary's external network constitutes an important part of its core competence with direct consequences for its expected performance.

However, the result also indicates that a subsidiary's technical embeddedness also predicts its role within the MNC. Hypothesis 5a, not 5b, is supported, which indicates that there is a positive relationship between a subsidiary's technical embeddedness and its importance for other units' competence development. Or expressed differently, the subsidiary's ability to contribute to other units' competence development is not counterbalanced by the problem of transferring knowledge from one subsidiary to other subsidiaries due to context specificity.

Hypothesis 3 is not supported, which indicates that business embeddedness has no direct impact on the subsidiary’s expected performance. However, the LISREL analysis suggests business embeddedness has an indirect impact on subsidiary performance and competence development through its impact on technical embeddedness. This result confirms that the two types of embeddedness play different roles for performance and competence development in the MNC. Close business relationships with external customers and suppliers will lead to relation-specific adaptation of resources and activities directed to product and process development. This, in turn, will improve the subsidiary's position in its own market place, as well as its possibility to contribute to other units' development of their products and production processes.

Finally, Hypothesis 2 is not supported. Expected market performance does not seem to have a significant impact on importance for competence development in other units. Or expressed differently, a subsidiary’s future prospects in its own market place can be good without the subsidiary being influential on other units' competence development.

CONCLUDING REMARKS

The model is only partial in the sense that competence development in MNC units depends on many factors. In the context of the differentiated MNC there are three modifying factors that should be pointed out specifically; type of technology/product at the subsidiary level, the importance of corporate embeddedness and the corporate managers' use of mechanisms to co-ordinate across organizational and country borders.

It is obvious that the relationships in the model are heavily dependent on the type of operations in the subsidiary. The problem of knowledge transfer has always been related to the problem of transfer of tacit, idiosyncratic, knowledge. The less the operations in the subsidiary, including its relationships with external counterparts, are built on tacit knowledge, the less the problem of knowledge transfer to other units. That is, Hypothesis 5b will be of less significance in such cases, especially due to the fact that context specificity will be of less importance. The data includes a fair amount of low technology subsidiaries. It is possible that the result would have been different if the sample only had contained high technology subsidiaries.

On a general level it has been argued that the type of relationship between two actors has a profound impact on the possible knowledge transfer between the two. This is actually one of the most important determinants for absorptive capacity (Lane and Lubatkin 1998). Even if external relationships usually dominate a subsidiary's network we have to consider the role of relationships with corporate units for our model. In principle, these relationships can have been built up over a long time and have the same characteristics as the external ones. Therefore, we have reason to assume that the more the subsidiary is involved in business relationships with other corporate units, the better the possibility that competence developed in the subsidiary, due to external relationships, can be transferred and used in these units. Consequently, corporate embeddedness can be seen as a variable that modifies the negative influence due to context specificity and lack of motivation to participate in knowledge transfer within the MNC.

This line of thought can be carried a bit further. Often an MNC subsidiary functions as a link between competence development in the external environment and in the corporate environment. Assimilation and commercialization of new knowledge is carried out through the relationships with both external and corporate counterparts. In such a subsidiary there is no immediate contradiction between deploying resources in relationships with external customers and suppliers compared to relationships with corporate units. New competence is not first of all developed in exchange with external actors and then (hopefully) transferred to other corporate units. It is developed in a constant and simultaneous interplay with external and corporate units. In these cases Hypothesis 5b has no relevance.

The model deals mainly with the relationships between structure and performance. But we have to realize that different forms of conduct can function as an intermediary variable. One conduct variable is the co-ordination mechanisms used by the corporate headquarters. Several authors have pointed out that there are integrative devices that can be used by the headquarters to stimulate co-operation and co-ordination between subunits, including competence development (Galbraith 1973, Edström and Galbraith 1977, Hedlund 1986, Ghoshal and Bartlett 1988, 1990, Grant 1996). The more successful these devices are in stimulating co-operation and knowledge sharing between subsidiaries with different business contexts and interests, the less powerful will Hypothesis 5b be. Or expressed differently, the negative impact of external embeddedness on the subsidiary's importance for competence development in other units can be reduced or eliminated due to successful knowledge management by the corporate or divisional headquarters.

To sum up, the results indicate that network embeddedness matters when performance and competence development in MNCs are focused. However, this does not imply that all relationships in a subsidiary business network must be highly embedded, but rather that some of them must have these characteristics. The possibility that a business network becomes too embedded and tightly structures has also been pointed out by several scholars (for a discussion of this problem see e.g. Granovetter 1973, Cristensen and Bower 1996, Uzzi 1997, Håkansson and Snehota 1998). A crucial task for management will be to handle the possible trade off between flexibility connected to arm’s-length relationships and learning connected to embeddedness. But that does not contradict the fact that embeddedness is an important feature of the business network when performance and competence development in MNCs are focused.

 

References

Allen, T.J., 1977. Managing the Flow of Technology. Cambridge, Mass.: IMT Press.

Almeida, P., 1996. Knowledge souring by Foreign Multinationals: Patent Citation analysis in the U.S. Semiconductor Industry. Strategic Management Journal, Vol. 17, 155-165.

Amin, Ash and, Nigel Thrift, 1994. Living in the Global. In A. Amin and N. Thrift, editors, Globalizations, institutions and regional development in Europe. Oxford UK: Oxford University Press.

Anderson, Erin, 1990. Two firms, one frontier: On assessing joint venture performance. Sloan Management Review, Winter: 19-30.

Andersson, Ulf and Mats Forsgren, 1996. Subsidiary Embeddedness and and Control in the Multinational Corporation. International Business Review, 5(5), 487-508.

Andersson, Ulf and Mats Forsgren, 1999. In Search of Centre of Excellence: Network Embeddedness and Subsidiary Roles in Multinational Corporations. Management International Review (Forthcoming).

Bierly, P., Chakrabarthi, A. 1996. Generic Knowledge Strategies in the U.S. Pharmaceutical Industry. Strategic Management Journal, 17, 123-135.

Birkenshaw, Julian M. and A. M. Morrison, 1996. Configurations of strategy and structue in multinational subsidiaries. Journal of International Business Studies, 26(4): 729-794.

Cantwell, John 1990. A survey of theories of international production. In Pitelis, N., Sugden, R., The Nature of the Transnational Firm. London:Routledge.

Chesnais, F. 1986. Science, technology and competitiveness. Science Technology Industry Review, 1:85-129.

Cohen, W., Levinthal, D.A. 1990. Absorptive Capacity: A new perspective on learningnad innovation. Administrative Science Quarterly, 35:128-152.

Crozier, M. 1964. The bureaucratic phenomenon. Chicago: University of Chicago Press.

De Mayer, A.1992. Management of international R&D operations. In O. Granstrand, Håkanson, L., Sjölander, S.(eds), Technology management and international business. Chichester: John Wiley & Sons Ltd.

Dosi, G., Freeman, C., Nelson, R.N., Silverberg, G., Soete, L. 1988. Technical change and economic theory. London: Pinter Publishers Ltd.

Dunning, John, 1996. The geographical sources of the competitiveness of firms: some results of a new survey. Transnational Corporations, 5(3): 1-29.

Edström, A., Galbraith, J., 1979. Transfer of Managers as a Coordination and Control Device in Multinational Organizations. Administrative Science Quarterly, 22, pp 248-63.

Egelhoff, William G. 1988. Organizing the multinational enterprise - An information-processing perspective. Cambridge, MA: Ballinger Publishers Company.

Ford, David, editor, 1997. Understanding Business Markets. London: The Dryden Press.

Forsgren, Mats, 1997. The Advantage Paradox of the Multinational Corporation. In Björkman, I., Forsgren, M., The Nature of the International Firm. Nordic Contributions to International Business Research. Kopenhagen: Copenhagen Business School Press.

Forsgren, M., Johanson, J., Sharma, D., 1999. Development of MNC Centres of Excellence. In Holm, U., Pedersen, T., The Emergence and Impact of MNC Centres of Excellence: A Subsidiary Perspective. MacMillan (Forthcoming).

Galbraith, J. 1973. Designing Complex Organizations. Reading, Mass.: Addison Wesley.

Ghoshal, S., 1986. The Innovative Multinational: A Differentiated network of organizational roles and Management Processes. Unpublished doctoral dissertation. Harvard Business School.

Ghoshal, S., Bartlett, C. 1990. The Multinational Corporation as an Interorganizational Network. Academy of Management Review, 15, 4, pp. 603-625.

Ghoshal, S., Bartlett, C., 1988. Creation, adoption, and diffusion of innovations by subsidiaries of multinational corporations. Journal of International Business Studies, Fall, 365-387

Ghoshal, S., Nohria, N. 1987. Multinational Corporations as Differentiated Networks. Working Paper 87/13, Fontainebleau, France: INSEAD.

Ghoshal, S., Nohria, N., 1989. Internal Differentiation within Multinational Corporations. Strategic Management Journal, 10, 1989.

Grabher, G. 1993. Rediscovering the Social in the Economics of Interfirm Relations. In Grabher, G. (ed) The Embedded Firm. London:Routledge.

Granovetter, M. 1985. Economic action and social structure: The problem of embeddedness. American Journal of Sociology, 91(3): 481-510.

Granovetter, M., 1992. Problems of explanation in economic sociology. In Nohria, N., Eccles, R.(eds), Networks and organizations: Structure, form and action. Boston: Harvard Business School Press.

Grant, R. M. 1996. Toward a knowledge-based theory of the firm. Strategic Management Journal. 17, 109-122.

Gupta, A.K. & Govindarajan, V., 1991. Knowledge Flows and the structure of control within multinational corporations. Academy of Management Review, 16(4): 768-792.

Gupta, A.K., Govindarajan, V. 1994. Organizing for knowledge flows within MNCs. International Business Review, 4: 443-457.

Hansen, M.T., 1999. The search-transfer problem: The role of weak ties in sharing knowledge across organization subunits. Administrative Science Quarterly, 44(1), 82-111.

Hayes, R.H., Clark, K.B. 1985. Exploring the Sources of Productivity Differences at the Factory Level. New York: Wiley.

Hedlund, G., 1986. The Hypermodern MNC: A Heterarchy? Human Resource Management, 25(1), 9-35.

Hickson, D.J. Hinings, J.R., Lee, C.A., Schneck, R.E., Pennings, J.M. 1971. A strategic contingencies theory of intraorganizational power. Administrative Science Quarterly, 16(1): 216-229.

Håkansson, H., 1989. Corporate Technological Behaviour: Co-operation and networks. London: Routledge 1989.

Jensen, M.C., Meckling, W.H. 1992. Specific and general knowledge and organizational structure. In Werin, L., Wijkander, H.(eds.), Contract economics. Oxford: Blackwell.

Johanson, J., Mattsson, L-G., 1987. Internationalization in industry system: A network approach. In N. Hood and J-E. Vahlne, (eds.), Strategies in Global Competition. Routledge, London, UK.

Jöreskog, K. G. and Sörbom, D., 1993. LISREL 8: Structural equation modelling with the SIMPLIS command language. Lawrence Erlbaum, Hillsdale, NJ.

Kelly, H. H. and Thibaut, J. W. 1978. Interpersonal Relations: A Theory of Interdependence. Wiley, New York

Kogut, B. Zander, U., 1992. Knowledge of the firm, combinative capabilities and the replication technology. Organization Studies, 3, 383-397.

Kumar, R., Nti, K.O. 1998. Differential Learning and Interaction in Alliance Dynamics: A Process and Outcome Discrepancy Model. Organization Science, 9 (3), 356-367.

Lane, P.J., Lubatkin, M. 1998. Relative absorptive capacity and interorganizational learning. Strategic Management Journal, 19, 461-477.

Lawrence, P.R. Lorch, J.W. 1967. Organization and environment: Managing differentiation and integration. Boston: Harvard Business School Press.

Malnight, T. W. 1996. The transition from decentralized to network-based MNC-structures. An evolutionary perspective. Journal of International Business Studies, 27(1): 43-65.

Mansfield, E. 1968. Industrial research and technological innovation. NY: W.W. Norton & Company, Inc.

Mowerty, D.C., Oxley, J.E., Silverman, B.S. 1996. Strategic alliances and interfirm knowledge transfer. Strategic Management Journal, 17, 77-91.

Pfeffer, J., 1981. Power in Organizations. Marchfield MA: Pitman Books Ltd.

Pfeffer, J., Salancik, G.R. 1978. The external control of organizations. A resource dependence perspective. New York: Harper & Row.

Piore, M., Sabel, C. 1984. The second industrial divide. New York: Basic Books.

Polyani, K., 1957. The great transformation. Boston: Beacon Press.

Porter, M., 1985. Competitive Advantage: Creating and Sustaining Superior performance. New York: Free Press.

Porter, M.E., 1990. The competitive advantage of nations. Free Press.

Powell, W.W., Koput, K.W. Smith-Doerr, L. 1996. Interorganizational collaboration and the locus of innovation: networks of learning in biotechnology. Administrative Science Quarterly, 41(1), 116-137.

Provan, K.Y. Beyer, J.M., Kruytboch, C. 1980. Environmental linkages and power in resource- dependence relations between organizations. Administrative Science Quarterly, 25(2): 200- 225.

Rosensweig, P. and Singh, J., 1991. Organizational environment and the multinational enterprise. Academy of Management Review, 16(2): 340-64.

Spender, J.-C., 1996. Making knowledge the basis of a dynamic theory of the firm. Strategic Management Journal, 17, 45-62.

Stopford, J., Wells, L.T. 1972. Managing the multinational enterprise: organization of the firm and ownership of the subsidiaries. London: Longman.

Szulanski, G., 1995. Unpacking stickiness. An empirical investigation of the barriers to transfer best practise within the firm. Academy of Management Journal. Special Issue. 437-444.

Szulanski, G., 1996. Exploring internal stickiness: Impediments to the transfer of best practice within the firm. Strategic Management Journal, 17, 27-43.

Sölvell, Ö. and Zander, I., 1995. Organization of the dynamic multinational enterprise: The home- based and the heterarchical MNC. International Studies of Management and Organization. 25(1-2): 17.

Taggart, J.H. 1998. Strategy shifts in MNC subsidiaries. Strategic Management Journal, 19: 663- 681.

Uzzi, B., 1997. Social structure and competition in interfirm networks. The paradox of embeddedness. Administrative Science Quarterly, 42:35-67.

Zander, U. and Kogut, B., 1995. Knowledge and the speed of transfer and imitation of organizational capabilities: An empirical test. Organization Science, 6(1), pp. 76-92.

Zukin, S., Di Maggio, PJ(eds) 1990. Structure of Capital. Cambridge: Cambridge University Press.

Wexley, K.N., R.A. Alexander, J.P. Greenawalt, and M.A. Couch. 1980. Attitudinal congruence and similarity as related to interpersonal evaluation in manager-subordinate dyads. Academy of Management Journal, 23(2): 320-330.