Tatiana Kostova 
International Business Program Area 
College of Business Administration 
University of South Carolina 
Columbia, SC 29208 
Tel. (803) 777-3553 
Fax: (803) 777-3609 
E-mail: kostova@sc.edu 
____________ 

This study was supervised by the late Prof. Larry L. Cummings. 

The study was generously supported by the Carnegie Bosch Institute, the Graduate School of  Industrial Administration, Carnegie  Mellon 
and the Carlson School of Management at the University of Minnesota, 
and the company which served as research site.    

Abstract This study draws on research in international management and organization theory to examine the factors of success of transnational transfers of organizational practices within multinational companies. A theoretical model is developed proposing two groups of factors operating at the country-level and at the level of the organization, respectively. The effects on the success of transfer and on performance of the recipient unit are examined. The model is tested with data from 110 facilities of a US multinational company in 10 different countries. Theoretical contributions, implications for business practice, and future research are discussed. 

Key words: transfer of practices; institutional environment; inter-organizational relationships.    

Introduction 

This study examines the causes of success and failure of transnational transfers of organizational practices within multinational companies (MNC). While this particular project focuses on quality management practices that are being transferred from US parent companies to their foreign subsidiaries, the theoretical model on which it is based is more general in nature and could be applied to a variety of organizational practices other than quality management as well as to transfers that go in various directions -- from parent to subsidiary, from subsidiary to parent, and between subsidiaries. The study uses a cross-disciplinary approach integrating ideas from international and strategic management, organization theory, organizational behavior, and social psychology. 

Problem Definition 

For purposes of synergy and efficiency organizations often engage in cross-unit transfers of 'best' manufacturing and organizational practices, i.e., practices that reflect the core competencies and superior knowledge of the organization and that are believed to be a source of competitive advantage. While the internal transfer of practices is important for all business organizations, it is extremely salient in the context of the MNC as it is widely believed, in the theory of foreign direct investments, that the primary advantage a multinational firm brings to foreign markets is its possession of superior knowledge (Kogut & Zander, 1993). 

The strategic importance of these types of transfers is well recognized but there is overwhelming evidence that the process of transnational transfer is not always smooth and successful. On many occasions, managers in foreign affiliates of MNCs report frustration with headquarters' requests for implementation of 'yet another' new program. Quite often, they, intentionally or not, make decisions not to implement the practice while reporting to headquarters otherwise. Sometimes they implement the practices only partially by adopting those components that they feel 'people here will buy in' and ignoring the rest. In some extreme cases, local managers feel alienated from the parent company to the extent that they do not believe in parent's motives and do not even consider complying with the requests for transfer. 
Theoretical  Development 

Research on Transfer of Practices 

The phenomenon of practice transfer across national borders has been studied mainly in the field of international management. It has been shown that: practices differ and MNCs capitalize on transfer of practices across borders (Lincoln, Hanada & McBride, 1986); practices are socially embedded, i.e., they reflect the socio-cultural environments in which they have evolved and are being used (Janssens, Brett, & Smith, 1995); there are various barriers to transfer success that reflect the characteristics of the practice or are of cultural and organizational nature (Kedia & Bhagat, 1988; Zander & Kogut, 1995); MNCs are better vehicles for practice transfer than are markets across nation-states (Kogut,1991); practices vary in their transferability and integration based on their role for the alignment of global goals and their cultural determination (Robinson, 1994); transfer is facilitated through normative integration and verbal information networks as well as through coordinated informal structures (Pfeffer & Leblebici; 1973); and, finally, import of practices from parent companies affects subunit's performance (Zaheer, 1995). The actual practices used by a particular subunit of a multinational company are the result of the interplay of pressures for local isomorphism with the host country, on one hand, and pressures for global integration with the parent company, on the other hand. 

Research Question 

This study attempts to forward our understanding of the phenomenon of transfer of organizational practices across national borders by addressing the following research question: 

Based on the work in 'new' institutionalism (Meyer & Rowan, 1991) and 'old' institutionalism (Selznick, 1957), organizational practices (successful and 'best' practices, in particular) are defined as institutions that have evolved over time under the influence of organization history, people, interests, and actions. Practices reflect the established ways of conducting organizational functions; they reflect the shared knowledge and competence of the organization; they are 'infused with value' -- have meaning for organizational members that goes beyond technical efficiency; they tend to be externally and internally legitimate -- accepted and approved by the external legitimating environment and by employees as 'the way of doing certain tasks'. This conceptualization of the term organizational practice has its implications for the understanding of the process of transfer of a practice between organizational units. It also affects the definition of success of transfer of an organizational practice, a dependent variable in the study. Finally, it allows the development of a theory-based model of factors of success. 

The process of transfer of organizational practices from parent companies to foreign recipient organizational units and its objectives and motives can too be conceptualized and examined from the 'new' and 'old' institutional theory perspectives. Following the 'new institutionalists' tradition, it could be viewed as a process in which an organizational unit adopts a certain practice for legitimating purposes including legitimating itself within the company, i.e., by the headquarters or the parent company. Employing an 'old' institutional theory perspective would emphasize another aspect of the transfer process -- the objective to transfer institutionalized meaning in addition to the formal rules and procedures implied by the practice. The process of transfer is viewed here as comprised of two elements -- diffusion of a set of rules as well as transmission or creation of an 'infused with value' meaning of these rules among the employees at the recipient unit. This approach suggests that the transfer process does not end with the adoption of the specified rules but continues until these rules become internalized at the recipient unit, i.e., until the employees there attach to the practice a similar symbolic meaning and value as the employees from the home of the practice. 

The old institutionalism perspective and the symbolic interaction perspective are useful for distinguishing between the transfer of different types of organizational systems such as, for example, organizational practices from technological or product innovations. Technological and product innovations are mainly knowledge-based and the success of their transfer is determined to a great extent by the transferability of knowledge (Zander & Kogut, 1995). Organizational practices, on the other hand, are meaning- as well as knowledge-based and the success of their transfer is determined by the transferability of meaning in addition to the transferability of knowledge. 

Based on the above theorizing, success of transfer of organizational practices is defined as the degree of institutionalization of the practice at the recipient unit, where institutionalization is conceptualized at two levels - implementation and internalization. Implementation is the degree of formal adoption of a practice expressed in external and objective behaviors and actions in the organization. Internalization is defined as the internalization of the practice by employees at the recipient unit, i.e., 'infusion with value' of the practice at the recipient unit. A practice becomes 'infused with value' when it is accepted and approved by employees, when it becomes the default way of conducting a certain function, and also serves as a source of satisfaction and self-identity for employees. Internalization is thus conceptually proximal to traditional constructs from the field of organizational behavior such as organizational commitment, job satisfaction, and psychological ownership which are used here to better define and operationalize the concept of internalization. 

Organizational commitment to the practice (adapted from Mowday, Steers, & Porter, 1979) is defined as the relative strength of an individual's identification with and involvement in a particular organization practice that can be characterized by a strong belief in and acceptance of the practice's goals and values, a willingness to exert considerable effort for the implementation of the practice, and a strong desire to continue the use of the practice. Satisfaction with an organizational practice (adapted from Locke, 1976) is defined as an individual's positive attitude towards the organizational practice. Psychological ownership of an organizational practice (Pierce, Van Dyne & Cummings, 1996) is defined as that state where individuals feel as though the target of ownership, an 'extended selves'. 

A Model of Transfer Success 

A theoretical model of the factors and consequences of transfer success was developed based on the above conceptualizations and information gathered in interviews with managers from the headquarters and foreign subsidiaries of two large US multinational companies. The model proposes two distinct sets of factors of transfer success - country-level effects and organization-level effects. 

==================== Insert Figure 1 About Here ===================== 

Country-Level Effects 

International management research has typically conceptualized country effects as cultural effects and has operationalized culture using Hofstede's (1980) four cultural dimensions. In this study we have followed a different approach. The country-level effects are conceptualized with the concept of institutional environment rather than culture and are operationalized accordingly. The institutional environment of an organization is defined as the set of regulatory, cognitive, and normative institutions or 'pillars' as suggested by Scott (1995). The regulatory component of an institutional environment reflects the existing laws and rules in a particular national or regional environment which promote certain types of behaviors and restrict others. The cognitive component reflects the widely shared cognitive categories used by people in a given social entity (Markus & Zajonc, 1985). The normative component reflects culture, i.e., values, beliefs, norms, assumptions about human nature and human behavior. 

It is argued in the study that using the concept of institutional environment rather than culture is a better way to capture the country-level effects on organizations. The concept of institutional environment is more encompassing than the concept of culture as it captures regulations and cognitive structures in addition to norms and values. It is also more proximal to organizations and therefore has a more direct effect on them. The concept of institutional environment is used to argue that: (1) practices that become established in a given organization are shaped after the particular external institutional environment in which this organization functions; (2) similar institutional environments result in similar practices used by organizations; (3) when organizations spread across multiple institutional environments and try to transfer practices across units, the degree of similarity between the institutional environments of these units affects the ease and ultimate success of the transfer. The concept of institutional distance is introduced and defined as the difference/ similarity between two institutional environments, in the context of this study - between the environments of the parent company and the recipient unit. 

==================== Insert Figure 1 About Here ===================== 

Two types of effects of the institutional environments on success of transfer are possible - direct and distance effects. The direct effect is determined by the extent to which a recipient environment is supportive of the particular practice. The distance effect is determined by the degree to which the home and the host institutional environments are similar or different: 

Quality of Relationships Between Parent Company and Subsidiary. While the concepts of institutional environment and institutional distance reflect the social embeddeddness of the processes of transfer, the quality of relationships between the parent and the subsidiary reflects their organizational embeddeddness. It is argued here that the quality of these relationships will affect the success of transfer because it affects the motivation of the managers and employees at the recipient unit to engage in implementation. Three particular types of relationships are proposed as affecting transfer success -- commitment of employees at the recipient unit to the parent company (Mowday, Steers, & Porter, 1979), identity of employees at the recipient with the parent company (O'Reilly III & Chatman, 1986), and trust of employees at the recipient unit in the parent company (Bromiley & Cummings's, 1995). The quality of relationships between recipient unit and parent company might be an especially salient factor for transfer success when the practice is complex and ambiguous and its value for the subsidiary is difficult to assess. In this case, recipient unit's decision whether to engage in implementation of a particular practice and how much effort to put into this could be affected to a greater extent by less direct factors, such as commitment to, trust in, and identity with parent company. 

Thus we propose: 

In accordance with the strategic management literature on core competencies and resource-based view of the firm (Barney, 1991), as well as the work of 'old' institutionalists as discussed above, we propose that the success of transfer of a particular practice, i.e., the degree to which the practice has been successfully implemented and institutionalized at the recipient unit, will have effects on that unit's performance: 

The general theoretical model described above was tested in this study for a particular type of MNC, a particular practice, and a particular type of transfer - the transfer of a quality management practice from a US parent company to its foreign subsidiaries. The practice of quality management was chosen because of its strategic importance for US multinational companies and because it is among the most frequently transferred organizational practices. Moreover, the problems experienced during transfer of quality practices seem to be representative and typical of the problems associated with the transfer of other practices in US MNCs. 

Unit of Analysis. This is a multi-level study that falls into the category of ecological studies (Leung & Bond, 1989) that work with aggregate measures of variables collected from N individuals belonging to n different cultures. Variables measured at the individual level are aggregated by computing means and standard deviations across individuals constituting each culture. In addition, there might be external variables that exist only on the cultural level. The choice of the level to which the individual data should be aggregated in this study was based on the theoretical levels of analysis as follows: propositions 1A and 1B - subsidiary; proposition 2, 3, and 4 - location. 

Research Site. The study was conducted at two large US MNCs. One of the companies was used as a site for conducting the preliminary model-building interviews. The second company was used as a site for testing the model with empirical data. This is a large US agricultural company (called here Globe) with more than 73,000 employees and presence in about 60 countries in North America, South America, Asia, Australia, Europe, Africa, and the Pacific Rim. The company has a product divisional structure with several major divisions and multiple lines of business within each division. Among other things, Globe was interested in the drivers and barriers to transfer success - the question addressed in this study. 

Sample Description. The sample consisted of all employees in all Globe's plants, offices, and sites located in nine foreign countries all of which had been involved in importing quality management practices from the US parent company. Countries included Canada, Argentina, the United Kingdom, the Netherlands, France, Spain, Australia, Portugal, and Malaysia. In addition to the foreign locations, three US-based plants were also included in the sample. Countries were included based on practical, as well as theoretical and methodological considerations. Practically, subsidiaries were included from those countries where the company had significant presence and the implementation of quality management practices was considered to be of strategic importance for the whole organization. Theoretically and methodologically, we included countries we thought would provide variance in the independent variables, and institutional profile, in particular. With few exceptions, all company locations from the selected countries and all employees at these locations were included in the study. The total number of countries included was 10, the total number of locations invited to participate was 111, and the total number of surveys sent to individual employees was greater than 8,350. 

Operationalization and Measurement The dependent variable, success of transfer, was operationalized with two sets of variables, respectively - Implementation and Internalization. Implementation was operationalized as Reported implementation (the degree to which the various components of the quality management program have been implemented in a location according to the official and/or formal reports given by that location back to the headquarters) and Actual implementation (the degree to which the various components of the quality management program have been actually implemented in a location). Two sources of data on Actual implementation were used - location's managers (Actual implementation - managers) and all location's employees using a 30-item scale designed after the Chairman's Quality Award criteria (Actual implementation - corporate standards). Internalization was operationalized by Employee commitment to the practice, Employee satisfaction with the practice, and Employee psychological ownership of the practice. Validated measurement instruments or adapted versions were used where possible. Location's performance was used in the model as a correlate of transfer success. Performance was operationalized with five general performance measures that are widely used in the company and that apply across divisions, countries, and locations including annual growth rates, average annual growth rates and five-year growth rates for Earnings per employee, Output per production employee, Return on assets, Gross profit per unit produced, and Safety index. 

The independent variables were operationalized and measured as follows. Institutional environment was operationalized as a three dimensional construct including regulatory, cognitive, and normative dimensions, respectively. It was measured with an original 18-item instrument developed and validated in this study. The instrument assesses the institutional profile of a country. Institutional distance between two countries was operationalized also in three dimensions consisting of the algebraic differences between the institutional profiles of the two countries. Quality of relationships between the parent company and the foreign location was operationalized with three constructs -- Commitment of the location to the parent company, Identity of the location with the parent company and Trust of the location in the parent company. Power/ Dependence between a location and the parent company was operationalized as perceptions of the senior management people at the location about that location's relative Autonomy and Influence on decision-making and parent company's Control over it. A brief summary of the measurement instruments used in the study is given in Table 1 in the Appendix. 

=================== Insert Table 1 About Here =================== 

Data Collection. Data collection methods included field surveys and interviews. Aimed at acquiring data from the most relevant sources, as well as avoiding potential problems of common-source biases, the study used four different questionnaires administered to three groups of respondents from the company and one group of respondents from outside of the company: Survey 1 - completed by Divisions' central offices at the corporate headquarters; Survey 2 - completed by the senior managers at each location; Survey 3 -- completed by all non-top-management managerial, professional and other employees at each location; and Survey 4 - completed by economic, trade, and agricultural officers from the embassies of the USA in the countries from the sample. A description of the sources of data and the content of the surveys used is given in Table 2 in the Appendix. 

=================== Insert Table 2 About Here =================== 

Development of the Surveys. All survey instruments were developed with the active participation of the Corporate Quality Department in Globe. Another issue of consideration in this study was the use of a cross-cultural sample of respondents who speak at least seven different basic languages. Accordingly, surveys 2 and 3 were double-reverse translated. Altogether, twelve different translations or adjustments of these surveys were done. 

Survey Administration: Procedures and Results. First, Surveys 2 and 3 were distributed to all locations in the ten countries included in the study. The distribution scheme was complex and was carried out by the researcher and a team of company employees. Most countries administered the surveys on-site. Various strategies were used aimed at increasing the response rate and ensuring confidentiality and anonymity. Survey 4 was administered by fax. Survey 1 was administered last. A total of 3,772 surveys 2 and 3 (from Globe's managers and employees) from 104 locations were completed. The response rate, as a whole, was good given the nature and the scope of the study -- 50% for managers (Survey 2), 43% for employees (Survey 3), and 44% total. A major hurdle in getting even higher response rate was the geographical dispersion of the participants especially where surveys were mailed to employees' home addresses. Surveys were literally distributed to hundreds of geographically distant locations in the participating countries. The results of the survey administration are summarized in Table 3 in the Appendix. 

=================== Insert Table 3 About Here =================== 

Analytical Procedures. This study employed a variety of analytical procedures to test the psychometric properties of the measurement instruments and the propositions. The instruments measuring the independent and dependent variables were tested for internal consistency and face and content validity using standard statistical techniques such as reliability analysis with Cronbach's Alpha and exploratory and confirmatory factor analyses. The propositions were tested with various techniques including ANOVA, MANOVA, correlation analysis and multivariate multiple regression. For the purpose of achieving cross-cultural construct equivalence we employed a variety of design and analytical strategies as recommended in the literature (Rosenzweig, 1994; Mullen, 1995; Cheng, 1994).  Results and Discussion 

Institutional Profile and Institutional Distance 

We consider the theoretical and the operational development of the constructs of institutional profile and institutional distance to be among the major contributions of this study. A significant amount of time and effort were spent for the development and validation of the measurement instrument for institutional profile. Using data on the regulatory, cognitive, and normative institutional profiles in 10 countries, that was collected here, and conducting some post-hock analyses, we were able to improve the scales measuring institutional profile to an acceptable level in terms of face and content validity and reliability. Only after that did we use the results (institutional profile indexes and institutional distances) for testing the substantive propositions. 

The analysis showed that the instrument is capable of discriminating between countries in regard to the degree to which their institutional environments are favorable for implementing quality management practices. ANOVA procedures yielded statistically significant between country differences in all measures of institutional profile, i.e., overall, regulatory, cognitive, and normative (p<0.0001). Overall, France, the USA, Canada, and the UK had the most favorable institutional environments while Malaysia, Portugal, and Argentina had the least favorable ones. The institutional profile indexes computed for the 10 countries are presented in Table 4 in the Appendix. Institutional distance is. Institutional distances (another important construct in the context of this study) were computed for all institutional profile measures between the US (home-country of the practice) and each of the other countries in the sample (recipient host-country of the practice). Results on institutional profiles and institutional distances are presented in Table 5 in the Appendix. 

======================== Insert Tables 4 & 5 About Here ======================== 

Psychometric properties of the measurement instruments 

The psychometric properties of the measurement instruments used in the study were tested thoroughly using all relevant statistical methods: reliability analysis and exploratory and confirmatory factor analyses. The results demonstrated high reliability and internal consistency of the scales and satisfactory overall fit of the measurement model. The exploratory factor analysis also supported the theoretically derived dimensionality of the constructs. An exception were the constructs of Psychological ownership of practice and Autonomy of subsidiary both of which yielded two dimensions instead of one. All analyses were performed also by country and showed satisfactory levels of the cross-cultural equivalence of the measures. 

Propositions testing 

Propositions 1A, 1B, 2, and 3 were tested at two different levels of analysis - with aggregated data at the level of the location and with individual-level data, respectively. As a whole, these two sets of tests produced similar results. There were some differences related to the power of the tests such as the number of significant relationships, for example, but the general nature of the findings and the direction of the relationships were the same. Given the similarity of the findings and for purposes of clarity, we do not separate the discussion of the findings by level of analysis in this paper. Proposition 4 was tested at the level of the location only. 

Propositions 1A, 1B, 2, and 3 were tested with multivariate multiple regression where three sets of dependent variables (Reported implementation and Actual implementation-managers; Actual implementation-corporate standards; and all Internalization variables) were separately regressed on the set of independent variables relevant to the proposition. For purposes of brevity, in this paper, we first describe the main findings and then present some of the main results in a summary table (see Table 6 in the Appendix). A detailed report on the statistical procedures and results is available from the author upon request. 

=================== Insert Table 6 About Here =================== 

Proposition 1A and 1B. The results for these two propositions about the effects of institutional environments on the success of transfer were mixed. The propositions were supported in general. Institutional environments were found to be significantly associated with all seven dependent variables measuring success of transfer. However, the dependent variables fell into two groups in regard to the direction (sign) of the relationships. The first group included Reported implementation and Actual implementation-managers. The second group included Actual implementation-corporate standards, Commitment to practice, Satisfaction with practice, and Psychological ownership of the practice. 

Reported implementation and Actual implementation-managers behaved exactly as proposed in the theoretical model. They were positively associated with Institutional profile as suggested in proposition 1A (direct effects), i.e., the degree of implementation of a practice at a recipient unit is higher when the institutional environment of the recipient unit is supportive of that practice. Reported implementation and Actual implementation-managers were negatively associated with the institutional distance between the parent company and the recipient unit. Countries that were more distant from the home country of the practice, i.e., the USA, had lower levels of implementation as assessed by these two measures than the countries that were more similar to the USA. Based on these results, it can be concluded that transfer of practices is more successful (regarding these two measures of implementation) when the institutional environment of the recipient country is supportive of the particular practice and/or when the recipient country is institutionally similar to the home country of the practice. 

The other measures of success of transfer -- Actual implementation-corporate standards and the four Internalization measures - behaved contrary to what was suggested in proposition 1A. and proposition 1B. They were negatively associated with the institutional profile of the recipient country and were positively associated with the institutional distance between the home country of the practice and the recipient country. This finding suggests in essence that internalization of a practice at a recipient unit is higher in countries where the institutional climate is not so supportive of the particular practice and in countries which are institutionally distant from the parent company. Thus, the less favorable the environment in a recipient country was in regard to quality, the higher were employee Commitment to, Satisfaction with, and Psychological ownership of the practice. Employees from Argentina, Portugal and Malaysia, the countries that had the least favorable institutional environments, reported the highest levels of Internalization of the quality practice. On the other hand, Internalization was the lowest in France and the USA, the two countries with the most favorable institutional climate for quality. 

The results thus suggested an interesting distinction between the Implementation and the Internalization variables measuring transfer success. These two sets of variables seemed to have opposite relationships with the Institutional profile and Institutional Distance variables. Potential reasons for these results are discussed and future research on this question is proposed. 

Proposition 2 was fully supported. It was shown that practices tend to be more successfully transferred to organizational units whose employees are more committed to the parent company, have a stronger sense of identity with the parent company, and trust the parent company. This finding is consistent with previous research in organizational behavior and intra-organizational relationships. It adds to this literature by applying the concept of intra-organizational relationships to a particular phenomenon, the transfer of practices, and showing a direct relationship between commitment, identity, and trust on one hand, and success of transfer, on the other hand. All three quality of relationships variables (commitment, identity, and trust) had significant effects on the success of transfer of practices. Moreover, all seven dependent variables measuring success of transfer were affected by the quality of relationships variables. Thus, this study strongly suggests that interactions between organizational units are affected by the quality of past relationships between these units. The higher the mutual commitment, identity, and trust, the more likely that the transfers will succeed. The poorer the relationships, the lower the chances of successful transfers. 

Proposition 3 attempted to test the general idea of power/dependence in the context of transfer of practices in multinational companies. The power/dependence argument suggested that organizational units that perceive themselves to be dependent on the parent company will put more effort into the transfer initiative because of the risks of unfavorable consequences in case of non-compliance. If the power/dependence motive was true, higher dependence would be associated with higher levels of transfer success. The results for this proposition were mixed. Control of location by parent was significant predictor of two of the Internalization variables -- Commitment to practice and Satisfaction with practice but was insignificant for Psychological ownership and for all Implementation measures. Therefore, the degree of implementation was not affected by the perceptions of the location of being controlled, i.e., dependent on the parent company. The other variable, Autonomy of subsidiary in decision-making, yielded completely opposite results to what was proposed by the power/dependence argument. It was found that Autonomy, which is a measure of relative power rather than dependence, affected positively all Implementation variables and Satisfaction with practice. Thus, in this sample, locations that considered themselves to be relatively autonomous and independent from the parent company, had achieved higher success in the transfer of the practice from the parent company. Although inconsistent with the proposition based on power/dependence logic, this finding is interpretable theoretically if the organizational behavior literature on participation in decision-making and empowerment is employed. 

Proposition 4 was partially supported. Some of the performance measures were significantly and positively correlated with some of the variables measuring implementation and internalization of the quality process. The average annual growth rate of Output per production employee was positively correlated with Actual implementation-corporate standards (r=0.5218; p<0.05), Employee commitment to the quality process (r=0.5224; p<0.05), and Psychological ownership 'WE' (r=0.5876; p<0.01). The average annual growth rates of Earnings per employee and Return on assets were positively correlated with Psychological ownership 'WE' (r=0.4785; p<0.05 and r=0.5704; p<0.05, respectively). The five-year growth rates of performance measures had only one significant relationships with measures of transfer success - Output per production employee was positively correlated with Psychological ownership 'WE' (r=0.5613; p<0.05).  Significance and Contribution 

This study is significant for several reasons. First of all, it addresses a question of practical relevance and strategic importance for multinational companies that has not been examined previously, i.e., the transfer of organizational practices. As argued in the study, organizational practices are different from technology and product innovations that have been the major focus of past research. Transfer of practices goes beyond transfer of information and knowledge, to include transfer of interpretation and meaning. Companies that know how to do this well have better chances of achieving synergy and overall efficiency by using the practices that have been developed successfully at one unit on an organization-wide basis. Finding the factors that contribute to the successful transfer or cause transfer failure and examining the performance effects of the institutionalization of the practices in the recipient subsidiaries results in specific recommendations for practicing managers. This study has advanced our knowledge of the effects of institutional environments and intra-organizational relationships on the success of transfer. This study is also important because it looked at a practice of strategic significance for US companies, i.e., quality management. This study provided evidence for several country level and organizational level determinants of the success of transfer of the quality practice. It also showed that there is a direct positive relationship between the institutionalization of the quality practice at a particular organizational unit and the growth of that unit's performance. 

The study also has significant theoretical contributions. It draws from and integrates several disciplinary areas -- international management, organization theory and organizational behavior to develop a cross-level model of transfer success. It uses past research from these areas to better conceptualize and operationalize the key constructs. In doing so it links traditional concepts coming from relatively independent established fields. The study crosses levels of analysis, applying for example, the institutional theory framework to an intra-organizational setting, and using organizational behavior constructs to operationalize concepts from institutional theory. A key construct that was conceptualized and operationalized in this way was the concept of success of practice transfer. In addition to the more traditional measures of Actual implementation of the components of the practice at the recipient unit, this study looked at the Internalization of the transferred practice by the employees at the recipient unit. It was found that these two measures of success may differ significantly and are affected in a different way by the institutional environments. 

A major theoretical contribution of the study is the conceptual development and the operational use of the concepts of Institutional environment and Institutional distance. Unlike most cross-cultural comparative research in the past, which has tried to account for country differences by using broad categories such as Hofstede's (1980) four cultural dimensions, this study takes a much more proximal approach by looking at institutional environments and in regard to a particular practice. Thus, although reflecting the broad differences in the regulatory, cognitive and normative components of the institutional environments, this construct is very close to peoples' behavior related to the particular practice. Data about ten countries collected from multiple sources demonstrated significant differences between countries in Regulatory, Cognitive, and Normative institutional profiles. Indexes for these variables were computed for all countries. 

Finally, this study is important because it sets a rich framework abundant with ideas for future research. Possible directions to generalize and expand this particular study include, for example, other organizational practices, non-US multinational companies, determinants of institutional distance, strategies for implementation and institutionalization of practices, further examination of some of the unexpected findings in this study, etc. 

   

Figure 2 Institutional Distance in the Context of Transnational Transfer of Practices  

Where: 

R - regulatory institutional environment 

C - cognitive institutional environment 

N - normative institutional environment 

Where: 

(Levels are computed on a 7-point Likert scale) 

Table 5 Institutional Distance Between the USA and the Other Countries 
 

Country	Institutional Profile	Regulatory	Cognitive	Normative
Canada	-0.08	-0.21	-0.46	0.29
Argentina	-0.74 *	-1.35 *	-0.72 *	-0.24
UK	-0.09	0.19	-0.53 *	0.05
Holland	-0.20	0.36	-0.99 *	0.01
France	0.06	0.60	-0.47 *	0.14
Spain	-0.39 *	-0.26	-0.62 *	-0.29
Australia	-0.14	0.10	-0.22	-0.16
Portugal	-0.97 *	-0.80 *	-1.35 *	-0.81 *
Malaysia	-0.79 *	-0.93 *	-0.29	-0.90 *

* - p < 0.05  Table 6. Summary of the Propositions Testing Results (Individual Level) 
 

Dep. Variable  Indep. Variable	Reported Implementation	Actual Implementation - managers	Actual Implementation - corporate standards	Commitment to Quality Practice	Satisfaction with Quality Practice	Psychol. Ownership 'I' of quality practice	Psychol. Ownership 'WE' of quality practice
Proposition 1.A
Regulatory profile	.1397**		-.1341**	-.2129**	-.0748**	-.1854**	-.1179**
Cognitive profile	.0923*	.1110**	-.0565**	-.0517**	-.3213**		-.0442*
Normative profile				-.0477*	-.0375*	-.0490*
Proposition 1.B
Regulatory distance			.0671**	.1181**
Cognitive distance	-.1209**	-.1209**			.3341**
Normative distance			.0978**	.1565**	.0526**	.1771**	.1385**
Proposition 2
Commitment to parent	.3864**	.4466**	.1011**	.0721**	.0753**	.0538**	.0453*
Identity with parent	.3237**	.3463**	.1067**	.1413**	.1000**	.0989**	.0879**
Trust in parent	.1319**	.3237**	.0846**		.1158**		.0546*
Proposition 3
Autonomy of location	.1276**	.1378**	.0596**		.0886**
Control by parent				.0477**	.1112**
Proposition 4
Earnings / employee							.4785**
Output/ prod. employee			5218**	.5224**			.5876**
Return on assets							.5704**
Gross profit/unit produced
Safety index

Only statistically significant relationships are shown in the table. Numbers represent multivariate multiple regression coefficients for Propositions 1-3 and correlation coefficients for Proposition 4. 

* p < 0.05 

** p < 0.01