FOREIGN SUBSIDIARY COMPENSATION STRATEGY: AN AGENCY THEORY PERSPECTIVE

 Kendall Roth and  Sharon O'Donnell 
The University of South Carolina 
College of Business Administration 
International Business Program Area 
Columbia, SC 29208 

ABSTRACT

 In addition to the unit of analysis issue, as noted by Balkin and Gomez-Mejia (1987), the contingencies or conditions under which different reward structures may be more or less appropriate are not well understood. Concerning the management of multinational corporations, a dramatic change in the extent of integration and coordination of activities occurring on a worldwide basis is resulting in a very different industry condition. Prior to this industry "globalization" the multinational could view its subsidiaries from essentially a portfolio perspective. While transfers of tangible and intangible assets necessarily occurred to support foreign subsidiaries, the role of each subsidiary was focused dominantly on local country markets. Correspondingly, the primary concern of the foreign subsidiary manager was the performance of the subsidiary in the local market. However, globalization is resulting in a significant redefinition of the role of the subsidiary manager as the subsidiary becomes more tightly integrated with other subunits within the corporation. The subsidiary manager remains accountable for market or country-based performance but must increasingly be concerned with the contribution made by the subsidiary to the competitive position of the corporation as a whole. 

 As a result of globalization, subsidiary performance becomes multifaceted, defined by both local and worldwide corporate objectives. Presumably, this fundamental change in the role or focus of the subsidiary should be supported by a properly configured reward structure. Thus, the primary research objective of this study is the determination of "proper" compensation strategies to support managers of foreign subsidiaries within the global industry context. More specifically, it is argued that the agency problem associated with foreign subsidiary characteristics is a critical influence in the determination of compensation strategies necessary to produce desired organizational outcomes. 

THEORY AND HYPOTHESES

 The principal can further limit the agency problem by incurring monitoring costs or through the design of appropriate incentives for the agent (Jensen and Meckling, 1976: 308). In simple situations, resources may be invested into directly monitoring agent's actions (Holmstrom, 1979). However, in more complex situations, difficulty in monitoring or verifying agent behavior results from "information asymmetries" where the agent has information that is not available to the principal (Gomes-Mejia & Balkin, 1992b). Information asymmetries are created as the agent has greater specialized knowledge than the principal regarding task performance and where the agent has a high level of managerial discretion (Gomez-Mejia & Balkin, 1992b; Rajagopalan & Finkelstein, 1992). Managerial discretion accompanies environmental and strategic complexities which create operational contexts with multiple decision options, low task programmability and ambiguous cause-effect relationships (Eisenhardt, 1988; Gerhart & Milkovich, 1990; Rajagopalan & Finkelstein, 1992). Agency theory prescribes the use of compensation based on the performance of the agent as resultant information asymmetries make the behavior of the agent costly or difficult to observe (Conlon & Parks, 1990). 

Agency Problem in the Headquarters-Foreign Subsidiary Relationship

As the principal, the headquarters cannot effectively make all the decisions in the MNC since it does not possess and must, therefore, depend on the unique knowledge of subsidiaries. At the same time, the headquarters cannot relinquish all decision-rights to the subsidiaries since the local interests of subsidiaries may not always be aligned with those of the headquarters or the MNC as a whole (1994: 492).

 The second factor that increases the agency problem in the headquarters-foreign subsidiary relationship concerns the strategic and operational role of the foreign subsidiary. Within a global industry, the role of a foreign subsidiary ranges from global rationalization to lateral centralization (Crookell, 1990; Roth & Morrison, 1992; Rugman & Bennett, 1982). Global rationalization occurs when the foreign subsidiary is a single part of a system rationalized worldwide, with the responsibility for system coordination residing at headquarters. The foreign subsidiary may perform only a subset of the value-adding processes comprising the system. While the subsidiary is interdependent with other entities within the firm, there is little agency problem at the senior management level as actions and outcomes of the foreign subsidiary must be relatively visible within the networked system and the specialized knowledge to manage the system is headquarters-based. Similarly, managerial discretion is low at the subsidiary level given that the coordinative decision-making responsibilities reside at headquarters. 

 At the other extreme, with lateral centralization, a foreign subsidiary has worldwide responsibility for a complete set of value activities associated with a specific product or product line. The subsidiary controls the research and development, production and marketing activities of a product or product line on a global basis. Thus, in response to competing within the global industry context, strategic and operation responsibilities are centralized and coordinated worldwide. However, rather than centralization of global decision-making being located solely at headquarters, the responsibilities are dispersed laterally throughout the organization, with the set of value activities for different products or product lines being centralized at different subsidiary locations. Referred to as "transnational" or "differentiated networks," there is an increasing amount of empirical support for these dispersed and differentiated roles within the multinational corporation (Bartlett & Ghoshal, 1989; Ghoshal & Nohria, 1989; Nohria & Ghoshal, 1994; Roth & Morrison, 1992). 

 Lateral centralization requires subsidiary management to have both specialized knowledge and considerable managerial discretion. Specialized knowledge exists as information, concerning the product/market and across unit linkages, is managed globally by the subsidiary. The subsidiary needs this information to facilitate decision making concerning the coordination of activities of other subsidiaries residing in multiple country locations. Headquarters is thereby dependent on the subsidiary for this information. Concerning managerial discretion, with lateral centralization foreign subsidiary management has direct responsibility for a multi-faceted and proactive global function. Such a role increases the latitude of options available to subsidiary management (Hambrick and Finkelstein, 1987). With the increased decision options of subsidiary management that accompany managerial discretion, management behavior is essentially non-programmable. Rajagopalan and Finkelstein suggest that, as managerial discretion increases, "managers are less constrained in decision making, and monitoring managerial work is more difficult" (1992: 128). In addition, strategic roles conferring managerial discretion are characterized by high ambiguity in behavior-outcome relationships as the number of factors influencing outcomes increases with discretion (Rajagopalan and Finkelstein, 1992). Thus, it is expected that the specialized knowledge and managerial discretion associated with lateral centralization result in information asymmetries which increase the agency problem in the headquarters-foreign subsidiary relationship. 

 The third contingency affecting the agency problem concerns commitment or psychological alignment at the individual level. Agency theory assumes that the principal-agent relationship is a social relationship. The fundamental issue is the divergence of principal and agent interests or goal incongruence. Organizational commitment, as an attitude, has been defined as identification with and willingness to embrace organizational goals (Mowday, Porter & Steers, 1982). It follows, therefore, that as an agent accepts and works toward organizational goals, goal incongruence between the principal and agent is reduced and, as a result, the agency problem is low. As articulated by Eisenhardt (1989: 62), "if there is no goal conflict, the agent will behave as the principal would like, regardless of whether his or her behavior is monitored." 

 Applied to the foreign subsidiary manager as an agent, his or her values or identification may vary in the degree to which they are attached to the "principal" organization. We argue that the notion of parent commitment, the psychological identification of the foreign manager with headquarters, is particularly important in determining the agency problem within the global industry context. The importance of parent commitment in this context results from the effects of international interdependence within the organization. Competition in a global industry implies increased international interdependence within firms as activities and resources are integrated regionally or worldwide (Bartlett & Ghoshal, 1989; Kobrin, 1991; Roth, Schweiger & Morrison, 1991). Complex interdependence requires extensive collaboration and mutual adjustments among the participants involved (Galbraith, 1987; Saavedra, Earley & Van Dyne, 1993; Thompson, 1967). For the foreign subsidiary manager, while commitment to local operations remains important, a high level of commitment to a corporate-wide perspective is necessary as he or she may need to embrace decisions or adjustments that are suboptimal at the subsidiary level. Thus, researchers have suggested the need for a "common world view" or a "company way" of addressing these complex interdependencies within the global industry (Bartlett & Ghoshal, 1989; Prahalad and Doz, 1987). Essentially, this is an application of control through socialization, where the desire is to "deemphasize national cultures and to replace them with an integrating company culture" so that managers at different foreign subsidiary locations are willing to work together and accept a common way of doing things (Edstrom and Galbraith, 1977: 256). Thus, it is argued that as the parent commitment of the foreign subsidiary manager increases, goal conflict or incongruence between the manager and headquarters is less and, as a consequence, the agency problem is reduced. 

 In summary, we have argued that cultural distance and lateral centralization influence information asymmetries and parent commitment influences goal incongruence within the headquarters-foreign subsidiary relationship. These two dimensions, information asymmetries and goal incongruence, are fundamental to determining the extent to which an agency problem potentially exists. To address the agency problem, in contrast with devising monitoring systems, the principal may also counter information asymmetries and goal incongruence by establishing reward incentive systems for the agent (Eisenhardt, 1988; Jensen & Meckling, 1976). As described by Gomez-Mejia and Balkin (1992b: 923), "when an agent has high autonomy, independence and highly specialized knowledge, monitoring becomes very difficult and expensive, so principals will rely on incentives to reward agents for appropriate outcomes." Essentially, as the difficulty and costs of monitoring increase, incentive alignment becomes a more important means by which to address the agency problem. This "incentive alignment" is defined as the extent to which the reward structure is designed to induce managers to make decisions in the best interests of the principal (Tosi & Gomez-Mejia, 1989: 171). Properly designed, the reward structure promotes self-monitoring as it is provides incentives that "impel agents to minimize opportunistic behavior and promote their compliance with principals' interest" (Kosnik & Bettenhausen, 1992). 

Hypothesized Compensation Strategy

 While the preceding arguments focused on defining the agent as the top management of the foreign subsidiary, individuals at lower levels of the subsidiary may also contribute to the agency problem. As will be argued later, this is a particularly important consideration in the global industry context. Thus, the pay mix for the entire subsidiary was considered an important compensation component to include in this study. The final component, senior management salary adjustment criteria, was included because, at the subsidiary level, "nonincentive" compensation often comprises a major portion of the reward structure. Therefore, the basis by which performance is assessed for salary adjustments was considered to be a potentially critical compensation component to examine in the foreign subsidiary context. 

 Senior management pay mix. The discussion in the preceding section suggested that, due to increased goal incongruence and information asymmetries, the agency problem is potentially greater as a foreign subsidiary has: (1) greater cultural distance from headquarters, (2) a lateral centralization form and (3) less commitment by the senior subsidiary manager to the parent organization. The arguments regarding the design of compensation strategy in response to agency problem are well developed in the literature (readers are referred to Fama & Jensen, 1983; Holmstrom, 1979; Shavell, 1979; Tosi & Gomez-Mejia, 1989). Given the potential agency problem, the basic objective is to specify the most efficient contract by which to govern the principal-agent relationship. Two basic contract forms are delineated in the literature: behavior-oriented and outcome-oriented (Eisenhardt, 1989; Nilakant & Rao, 1994; Rajagopalan & Finkelstein, 1992). Behavior-oriented contracts include non-incentive or fixed salary compensation whereas outcome-oriented contracts would include incentive or commission salary. As summarized by Tosi and Gomez-Mejia (1989: 172), theory and research both lead to the conclusion that the agency problem is minimized when compensation for senior managers is incentive-based, i.e. tied to their performance. Incentive-based compensation is the more efficient contract when appropriate managerial behaviors are difficult to monitor (Jensen and Meckling, 1976). Alternately, if the agent is expected to perform a behavior and that behavior can be observed directly, then the agent should be paid directly for that behavior via salary (Eisenhardt, 1988: 493). Extending these arguments to the foreign subsidiary context, the following hypotheses are suggested: 

• H1a: As the cultural distance of a subsidiary increases, senior subsidiary management will receive a greater proportion of their compensation through incentives.

 

• H1b: As the foreign subsidiary role is characterized by increased lateral centralization, senior subsidiary management will receive a greater proportion of their compensation through incentives.

 

• H1c: As the senior manager of the foreign subsidiary has lower parent commitment, senior subsidiary management will receive a greater proportion of their compensation through incentives.

 

• H2a: As the cultural distance of a subsidiary increases, the higher the compensation level for senior subsidiary management relative to the market.

 

• H2b: As the foreign subsidiary role is characterized by increased lateral centralization, the higher the compensation level for senior subsidiary management relative to the market.

 

• H2c: As the senior manager of the foreign subsidiary has lower parent commitment, the higher the compensation level for senior subsidiary management relative to the market.

 

 Multinational corporations pursuing global operational approaches to competing in global industries have been found to employ extensive coordination of subsidiary activities across locations and exhibit a high level of shared values or management philosophy (Bartlett & Ghoshal, 1989; Roth, Schweiger & Morrison, 1991). Global strategy implementation requires an increased capacity for integrated actions within the multinational. The coordination of activities at the functional and task levels creates operational interdependence. As described by Bartlett and Ghoshal, a critical implementation issue in managing global interdependence is developing a "shared vision and personal commitment to integrate the organization at the fundamental level of individual members" (1989: 66). This is because the information and tasks to be integrated occur at operational levels within the multinational. As a result, specialized knowledge is also developed at operational levels within the foreign subsidiary. Furthermore, to facilitate cooperative exchanges across borders it becomes critical for these individuals at lower levels of the subsidiary to embrace the corporate-wide goals. Thus, extending the previous arguments regarding pay mix at the senior management level, it is also expected that as the agency problem increases, increased use of incentive-based compensation applied at the subsidiary level will be the more efficient contract form. This form would provide motivational incentive at the individual level as required by Bartlett and Ghoshal's description of effective global strategy implementation, consistent with and reinforcing such incentives at the senior management level. Thus, the following hypotheses are forwarded: 

• H3a: As the cultural distance of a subsidiary increases, subsidiary personnel will receive a greater proportion of their compensation through incentives.

 

• H3b: As the foreign subsidiary role is characterized by increased lateral centralization, subsidiary personnel will receive a greater proportion of their compensation through incentives.

 

• H3c: As the senior manager of the foreign subsidiary has lower parent commitment, subsidiary personnel will receive a greater proportion of their compensation through incentives.

 

• H4a: As the cultural distance of a subsidiary increases, the greater the percentage weight given to regional and/or corporate performance as criteria for determining the salary adjustments of senior subsidiary management.

 

• H4b: As the foreign subsidiary role is characterized by increased lateral centralization, the greater the percentage weight given to regional and/or corporate performance as criteria for determining the salary adjustments of senior subsidiary management.

 

• H4c: As the senior manager of the foreign subsidiary has lower parent commitment, the greater the percentage weight given to regional and/or corporate performance as criteria for determining the salary adjustments of senior subsidiary management.

 

Implications for Subsidiary Effectiveness

 The preceding hypotheses are bivariate in nature, focusing on the design of a particular compensation component in response to the agency problem. Because bivariate relationships are essentially reductionistic, previous studies have found that using a systems or configurational approach is important when theory suggests that normative outcomes are associated with the alignment of a system (Govindarajan, 1988; Gresov, 1989; Venkatraman & Prescott, 1990). Such an approach attributes increased effectiveness to the internal consistency or fit among multiple contextual and/or design factors (Doty, Glick & Huber, 1993). 

 The hypotheses (H1-H4) specify a unified compensation strategy, the design of which depends on the agency problem. When the agency problem is "high" it is posited that foreign subsidiary compensation should emphasize incentive-based pay at both the senior management and subsidiary levels, pay above market compensation levels and utilize regional/corporate performance as a criteria for determining salary adjustments for senior subsidiary management. Incentive alignment, the degree to which the compensation system is designed to induce the subsidiary management to make decisions that are in the best interest of headquarters, should result in the foreign subsidiary realizing the outcomes or objectives designated by headquarters. 

 In this study, the desired outcome is defined as perceived subsidiary effectiveness. Perceived subsidiary effectiveness is examined because subsidiaries may be evaluated on a wide range of performance dimensions. As described by Govindarajan and Fisher (1990: 269), "it is not possible to use the same set of criteria to evaluate every SBU since, by definition, different SBU strategies imply quite different goals and priorities." This issue is particularly important in the context of subsidiaries residing across different country locations within organizations that vary subsidiary roles and responsibilities globally. For example, a subsidiary may operate in a location for the purpose of establishing a presence in a sophisticated market thereby providing access to leading edge consumers and technological developments. Evaluation of the subsidiary would necessarily include dimensions beyond traditional profitability indicators. In addition, implementing global strategies implies interdependence between units within the firm. Gomez-Mejia and Balkin (1992a: 44) note that as units within a firm are more interdependent, "the more difficult it is to identify the contribution [to performance] of any one unit." Thus, the definition of subsidiary effectiveness is determined by the performance dimensions expected by the principal. Based on the compensation strategy configuration articulated in the preceding hypotheses, the following systems hypothesis is forwarded: 

• H5: Incentive alignment will be positively associated with perceived subsidiary effectiveness.

 

METHODS

Industry Selection and Sample

 Within the selected global industries, foreign subsidiaries located in the U.S., U.K., Canada, Japan and Germany were identified using three directories: International Directory of Corporate Affiliations, America's Corporate Families and International Affiliates and Directory of Foreign Capital Affiliated Enterprises in Japan. For a foreign subsidiary to be included in the sample the parent company had to own at least 51% of the subsidiary and be headquartered in one of the other four countries. For example, in the U.K., foreign subsidiaries of U.S., Canadian, Japanese, and German-based corporations were identified. This process resulted in an initial set of 427 foreign subsidiaries for which a contact name could be identified. Further verification of the senior manager of each subsidiary and correct addresses and industry information resulted in a final set of 372 subsidiaries. Using a mail questionnaire, responses were received from the senior manager of 100 foreign subsidiaries. The country location distribution of the respondents was as follows: 40 foreign subsidiaries within the U.S., 22 in the U.K., 14 in Canada, 12 in Japan and 12 in Germany. The sales of the subsidiaries ranged from $2 million (U.S. dollars) to $1.8 billion with an average sales level of $144 million (s.d. = 265 million). The subsidiaries represented 73 different corporate parents with the sales of the parents ranging from $14 million to $60 billion with an average parent sales of $3.5 billion (s.d. = 8.6 billion). 

 Two procedures were used to examine nonresponse bias. Kazanjian and Drazin (1989) detail an extrapolation procedure for evaluating nonresponse bias. Based on the work of Armstrong and Overton (1977), they maintain that the profile of nonrespondents is likely to be more similar to late respondents than early respondents. This study used an initial mailout followed by two additional mailouts to nonrespondents in a period of four weeks and eight weeks after the initial mailing. Comparing all variables included in the models, the means of the last set of respondents did not differ significantly from the first set of respondents. A second procedure used to examine whether the responding firms differed from the nonresponding firms involved gathering secondary data for a subset of the nonresponding firms. Randomly selecting fifty-two subsidiaries for which data on sales and number of employees were available, a comparison of the means of these two variables indicated no significant differences existed. 

Measures

where CDjk = the cultural distance between countries j and k, Dij = the score for parent country j on cultural dimension i, Dik = the score for subsidiary country k on cultural dimension i, and Vi = the variance of the index for cultural dimension i. Following Kogut and Singh (1988), this formula corrects for the variance of each cultural dimension and averages across the four dimensions. 

 The second dimension, lateral centralization, was measured with an index patterned after the Roth and Morrison (1992) subsidiary scale. Executives were asked to indicate how characteristic eight statements were in describing the responsibilities of their subsidiary. The statements were: (1) the subsidiary is primarily an implementor of headquarters developed strategy, (2) the subsidiary has worldwide responsibility for production activities of a product or product line, (3) product expertise within the corporation resides within this subsidiary, (4) the subsidiary maintains control over the export marketing of products, (5) production process innovations are developed by the subsidiary, (6) the subsidiary has worldwide responsibility for marketing activities of a product or product line, (7) the subsidiary controls product research and development activities and (8) international market development costs are incurred by the subsidiary. A five point scale was used where 1 = "not at all characteristic" and 5 = "extremely characteristic" and the responses were averaged to create an overall lateral centralization score. 

 The construct reliability and validity of this measure was assessed three ways. First, the internal reliability (Cronbach's coefficient alpha) was examined and found acceptable (a = 0.90). Second, the critical implication of lateral centralization from an agency perspective is increased subsidiary autonomy. Using a measure adapted from Egelhoff's (1988) autonomy/centralization index, executives were asked to indicate for twenty-three decision areas the extent to which the decision was made independently by the subsidiary. As expected, the lateral centralization measure was found to be correlated with the autonomy index (r =0.71, p <.001). The final validity assessment was based on contacting the Senior V.P. or V.P. International at headquarters for twenty-eight subsidiaries. The identification of these individuals was based on phone interviews with subsidiary and/or headquarters personnel to determine the direct report relationship of the subsidiary. Using a mail questionnaire, the headquarters manager was provided the name of the specific subsidiary included in the study and asked to respond to the lateral centralization measurement items for the subsidiary. The correlation coefficient between the headquarters and subsidiary measures of lateral centralization was 0.80 (p < .001). 

 The third dimension of agency problem is parent commitment. Parent commitment was measured by the senior manager's commitment to headquarters using a scale adapted from Gregersen and Black (1992). The Gregersen and Black scale was developed from previous studies measuring organizational commitment and has been found to have a high level of reliability and validity (Mowday, Porter & Steers, 1982; O'Reilly & Chatman, 1983). The internal reliability for the parent commitment scale (a = 0.83) was generally consistent with reliabilities reported by Gregersen and Black (1992). 

 Compensation Strategy. Four components of compensation strategy were measured. Senior management pay mix was operationalized based on the compensation of senior subsidiary management contributed by: (1) short-term bonus plan, (2) long-term bonus plan, (3) short-term equity plan, and (4) long-term equity plan, as a percentage of total senior management compensation. This scale was based on similar measures used in the Balkin and Gomez-Mejia (1987) and Galbraith and Merrill (1991) studies and was self-reported by the subsidiary senior manager. To assess the validity of the information provided by the senior manager, for forty-one subsidiaries, the human resources (HR) manager was identified at the subsidiary or headquarters location. The HR manager was asked to provide the pay mix compensation information for subsidiary senior management. The average correlation between the pay mix reported by the subsidiary senior managers and the HR managers was 0.62 (p <.001). 

 Market positioning was defined as the extent to which the pay level of the subsidiary is above or below that of competitors. This component was measured using a scale adapted from Balkin and Gomez-Mejia (1990). For both salary and incentives, respondents were asked to rate the compensation level of their subsidiary as compared to that of competitors. 

 To operationalize subsidiary pay mix, the senior management pay mix scale was modified to capture the percentage of total compensation paid by the subsidiary through incentive-based compensation forms (using the four plans detailed above) to all employees. The average correlation between the pay mix reported by the subsidiary senior managers and the forty-one HR managers was 0.35 (p <.05). While significant, this correlation was lower than expected. Follow-up interviews indicated that headquarters respondents often had limited information about the use of incentives at lower levels within foreign subsidiaries. 

 The salary adjustment criteria concerns the basis for evaluating senior subsidiary management performance for salary adjustments. For the salary component of the compensation program the respondents were asked to indicate "the approximate weight given to each criteria" used as a basis for evaluating their performance. Respondents reported the percentage weight given to the following criteria: (1) individual performance, (2) subsidiary performance, (3) regional performance and (4) corporate performance. The salary adjustment criteria construct, reflecting the reward alignment for the nonincentive compensation component, was the aggregate percentage of the salary adjustment criteria given to regional and corporate performance. The correlation between the headquarters and subsidiary measures of salary adjustment criteria was 0.96 (p <.001). 

 Subsidiary Effectiveness. Given that performance expectations will vary across different performance dimensions as well as across different subsidiaries and country contexts, it was not possible to use a single performance criterion for all subsidiaries. Thus, a multidimensional scale was developed, with the performance dimensions taken from similar measures used in studies examining business unit or profit center performance (Govindarajan & Fisher, 1990; Gupta & Govindarajan, 1986). Ten performance dimensions were listed: sales volume, market share, profit, cash flow from operations, return on investment, new product development, market development, cost control, personnel development and political/public affairs. For each dimension, respondents were asked to rate the performance of their subsidiary, relative to superiors' expectations, on a 5-point scale ranging from 1="not at all satisfactory" to 5="outstanding." An average effectiveness index was created for each foreign subsidiary, based on the reported performance along each dimension. As noted by Gupta and Govindarajan, a high correlation between superiors' and self-ratings can be expected in situations where "subordinates are guaranteed anonymity and understand that the objective of data collection is scientific, not evaluation" (1986: 713). Both of these conditions were met in this study. In addition, Govindarajan and Fisher (1990) found a strong correlation between manager and superior assessments using this measurement approach. In this study, for twenty-eight subsidiaries, the Senior V.P. or V.P. International at headquarters also evaluated the subsidiary effectiveness using the ten performance dimensions. The correlation coefficient between the headquarters and subsidiary measures of subsidiary effectiveness was 0.65 (p <.001). Thus, while it is certainly acknowledged that reliance on a self-reported subjective measure has limitations, it would appear that the measurement approach used is both valid and appropriate for this study. 

 Control Variables. Studies suggest that corporate diversification influences compensation strategy. Gomez-Mejia (1992) argues that as corporations diversify, headquarters has less detailed knowledge about subunits. Interdependence between units may also decrease as a result of diversification. These conditions lead to greater use of incentive compensation and risk sharing (Gomez-Mejia, 1992). Thus, the extent of corporate diversification, measured by the number of industries (defined by four-digit SIC code) outside the firm's primary industry, was included as a control variable when estimating the compensation strategy relationships. Organization size has been found to be related to the design and positioning of compensation strategy (Lewellen & Huntsman, 1970; Fisher & Govindarajan, 1992). This influence occurs apparently at both the corporate and business-unit level (Balkin & Gomez-Mejia, 1990). Thus, corporate and subsidiary size, operationalized by the logarithm of annual sales, were included as control variables. Compensation strategy and perceived outcomes may also be influenced by the level of subsidiary profitability. The pretest indicated that managers would be hesitant to report objective performance information, so the profitability control variable, after tax return on investment, was operationalized using a 7-point range where: 1="negative return on total investment," 2="between 0%-5%," 3="between 5%-10%," 4="between 10%-15%," 5="between 15%-20%," ="between 20-25%," and 7="greater than 25%." 

 Several authors contend that the compensation strategy of foreign subsidiaries should reflect local cultural norms and standards (Gomez-Mejia & Welbourne, 1992; Hodgetts & Luthans, 1993). However, a study by Rosenzweig and Nohria (1994) found that while foreign subsidiary human resources practices closely paralleled local country contexts, executive bonuses did not vary. Thus, while the evidence is inconclusive, it is potentially important to examine the influence that the host country may have on compensation strategy design. The mean responses on the compensation variables were compared across host country locations to evaluate the appropriateness of pooling the data. This analysis indicated that foreign subsidiaries in the U.K. had a significantly lower use of incentive-based compensation for senior management and lower market positioning as compared to the other host country locations. Thus, a nationality dummy variable (coded for U.K.) was included in the senior management pay mix and market positioning equations. 

Analysis

 To derive the profiles, an aggregate "agency problem" index was created using the three agency dimensions. As suggested by Gresov (1989), observations scoring in the middle third were dropped to better distinguish differences between firms on the agency index. The top five performing firms were then identified in the "high" agency problem group and the "low" agency group. Means on the compensation components were computed for the high performing firms comprising each group and used to define the ideal profile. An alignment or fit measure was then calculated for the remaining firms within each group, based on the euclidean distance over the range of compensation components. In the final step, subsidiary effectiveness was correlated with the fit index. A significant negative correlation would support the alignment hypothesis, as deviations from the pattern would be associated with a decline in performance. 

RESULTS

TABLE 1

Descriptive Statistics and Correlations

 The first hypotheses (H1a-H1c) address the pay mix for senior management of the foreign subsidiary as it is related to the agency problem. As indicated in the first equation reported in Table 2, cultural distance was not found to be related to senior management pay mix, therefore H1a was not supported. Support was provided for H1b as incentive-based compensation for senior management increased as subsidiaries were characterized by lateral centralization. Opposite the hypothesized direction, the use of incentive-based compensation management was found to increase with higher levels of senior management commitment to the parent organization (H1c). 

 The second equation reported in Table 2 examines the posited relationship between the agency problem and the extent to which the pay level of the subsidiary is above or below that of its competitors. Market positioning was not found to be influenced by either cultural distance (H2a) or parent commitment (H2c). H2b was supported as market positioning was positively related to lateral centralization. 

 The relationship between the agency problem and the pay mix of the subsidiary is examined in the third equation in Table 2. Strong support exists for H3a; a greater proportion of incentive-based compensation was used for subsidiary personnel as the cultural distance between the subsidiary and headquarters increased. No support was found for H3b and H3c. The final equation in Table 2 examines the relationship between the agency problem and salary adjustment criteria for senior subsidiary management. The coefficients for both cultural distance and lateral centralization were not significant, offering no support for H4a and H4b. H4c was supported as the coefficient for parent commitment was significant and in the expected direction. Thus, consistent with the agency theory argument, a greater weight was attached to regional and corporate performance as criteria for senior management salary adjustments as the commitment of the senior manager to the parent organization was lower. 

TABLE 2

Results of Analysis for Compensation Strategy

 +p<.10 
*p<.05 
**p<.01 
***p<.001 

 The ideal profiles used in examining the incentive alignment hypothesis (H5) are reported in Table 3. The results of the pattern analysis examining the relationship between incentive alignment and subsidiary effectiveness is reported in Table 4. H5 was supported as subsidiary effectiveness was found to decline as subsidiaries moved away from the ideal profiles. 

TABLE 3

Incentive Alignment Ideal Profiles

TABLE 4

Correlations of Distance Measures and Subsidiary Effectiveness

DISCUSSION AND CONCLUSIONS

 The results of this study provide moderate support the application of an agency theory perspective to understanding compensation strategy within the headquarters-foreign subsidiary relationship. However, the results suggest that the agency problem within the foreign subsidiary context is multidimensional. This multidimensionality apparently results in different aspects of the agency problem influencing different compensation strategy dimensions. 

 Cultural distance and lateral centralization were considered important factors for understanding the degree to which the agency problem may exist for a headquarters-foreign subsidiary relationship. It was argued that these factors reflect the extent of the information asymmetry between headquarters and the subsidiary as a result of managerial discretion and subsidiary environmental and strategy-specific knowledge. Cultural distance was found to be associated only with subsidiary pay mix, with the use of incentive-based compensation increasing with distance. If a one-tailed test is used, cultural distance was also found to influence the use of regional and corporate performance criteria as a basis for salary adjustments (p<.03). In contrast, lateral centralization was found to influence the senior management pay mix and market positioning dimensions. 

 There are two basic explanations as to why cultural distance and lateral centralization may have different influences on compensation strategy. First, the two factors correspond to different analysis levels. Cultural distance is a nationality dimension that would be the same for a given particular home-host country combination, whereas lateral centralization is subsidiary specific. Thus, subsidiaries within a host country context may all have to respond to local contextual conditions in configuring certain components of their incentive structure, such as the use of incentives for all subsidiary personnel. However, for other compensation dimensions, such as the pay mix for subsidiary senior management, firm-specific pressures may dominate--such as attempting to provide equity among managers at relatively high levels of the organization--rather than host country pressures. This issue, the unique and interactive effects of local isomorphism, or responses to country-specific practices and firm-specific practices warrants additional research attention. Essentially, the results of this study suggest that future studies integrating institutional theory and agency theory explanations of compensation strategy, similar to Eisenhardt's (1988) approach, have the potential to make an important contribution to understanding compensation practices at the foreign subsidiary level. 

 The second explanation concerns the possible relationship between cultural distance and lateral centralization. For the subsidiaries in this study, cultural distance and lateral centralization were found to be negatively related (Table 1). This suggests that while multinational corporations are utilizing differentiated approaches in which subsidiaries are selectively given global responsibilities, the subsidiaries assuming such roles are "culturally close." Selecting culturally close subsidiaries may be a temporal or evolutionary state as firms initially employing lateral centralization may do so with subsidiaries more easily observed or trusted due to cultural- based understanding. Alternately, it simply may be that subsidiaries with the resources and competencies to assume a global role are the closer units, as many foreign investment activities historically have occurred in markets that are less culturally distant (Kogut & Singh, 1988; Johanson & Vahlne, 1990). 

 As stated previously, subsidiaries characterized by lateral centralization have an increased level of incentive compensation for senior management and a higher level of compensation relative to competitors as compared to subsidiaries that are part of a rationalized system. This is an important finding in that it not only confirms the applicability of agency theory in further understanding the headquarters-foreign subsidiary relationship, but it also verifies the need to realign compensation systems as MNCs reorganize and restructure global responsibilities. Specifically, as global strategies are pursued using "differentiated network" type implementation approaches, foreign subsidiary management compensation must include an incentive-based component for subsidiary senior management. Subsidiaries identified as "top" performers reported an average total incentive compensation for senior management of over thirty percent (Table 3). While this figure is only suggestive of an optimal level, it does indicate that a significant portion of the compensation is incentive-based. The inflection point in this relationship, where the outcomes no longer change given an increase in the incentive compensation structure, is certainly an area that needs to be examined by future research as this study does not establish the level where the incentive structure will optimize returns. 

 Lateral centralization does not appear to influence the subsidiary pay mix or the use of regional and corporate performance as a criteria for senior management salary adjustment. These results were not anticipated. It is possible that if the sample had greater representation of subsidiaries characterized by both cultural distance and lateral centralization, subsidiary pay mix would indeed be altered. Alternatively, lateral centralization may result in an agency issue predominantly at the senior management level, and thereby not requiring modification in the incentive structure for all subsidiary personnel. 

 Finding no relationship between lateral centralization and the salary adjustment criteria may be attributable to the type of international interdependence that could accompany both lateral centralization and global rationalization. While it was argued that lateral centralization affected the agency problem at the senior management level, operational interdependence at the functional or task level will exist with both the lateral centralization and global rationalization subsidiary forms. The distinction between the two forms, as related to the agency problem, concerns the strategic autonomy of senior management as a result of the locus of global decision-making. Regardless of the decision-making role, from a strategy implementation perspective, global responses require the foreign subsidiary to have integrated activities at an operational level with headquarters and other units within the multinational (Bartlett and Ghoshal, 1989; Roth, Schweiger and Morrison, 1991). As a result, for both lateral centralization and rationalization, some alignment of the salary performance criteria may be desirable. As international interdependence increases at an operational level, the use of regional and corporate performance criteria would occur irrespective of foreign subsidiary form. 

 The results regarding the influence of parent commitment are not straightforward. As expected by the agency arguments, with increased commitment to the parent a lower weight was attached to regional and corporate performance as a salary adjustment criteria. If goal congruence exists between the principal and agent there is little need for salary adjustments to be linked to the performance objectives of the principal. However, increased parent commitment was also found to be related to a higher use of incentive-based compensation, a finding counter to a agency theory view. Perhaps managers exhibiting a high level of commitment to the parent are perceived as having greater potential to impact organizational outcomes than managers with lower organizational commitment. Gerhart and Milkovich (1990) suggest that a strong perceived link between an employee behavior and organizational performance offers increased opportunity to use incentive-based compensation. Furthermore, the high commitment to the parent may be a result of considerable human capital investments such as training and experience. Extending human capital theory, Gerhart and Milkovich (1990) argue that these human capital investments are also likely to be related to a high potential to influence performance outcomes and therefore associated with the use of contingent pay. In the foreign subsidiary context, human capital theory and expectancy theory may provide a more useful theoretical link between individual characteristics and pay mix than agency arguments. 

 The final conclusion concerns the relationship between compensation strategy and organizational outcomes. As an integrated system, the design of the compensation strategy was found to have a strong impact on perceived subsidiary effectiveness. This result is rather tentative given that effectiveness was measured subjectively and self- reported by subsidiary management, however, it does suggest that reconfiguring the compensation strategy of the multinational corporation can be effective, particularly in response to global decision-making responsibilities, cultural distance and the commitment of the subsidiary management. 

 While the results of this study contribute to agency theory through applying the theory in the headquarters- foreign subsidiary context, it is acknowledged that the study examines the agency problem rather narrowly. Agency theory is also concerned with the levels and effect of monitoring. Thus, future research needs to incorporate other mechanisms by which the agency problem may be managed, in addition to the compensation strategy of the foreign subsidiary. Mechanisms, such as the use of expatriates, third-country nationals or local nationals with extensive headquarters work experience as well as management development programs providing international management rotation may provide corporate socialization which, in turn, influences the extent of the agency problem in the headquarters-foreign subsidiary relationship. Such mechanisms may thereby either substitute for or complement compensation strategy adjustments. In addition, for managers of foreign subsidiaries, the compensation package will typically include other benefits that extend beyond salary and incentives. Research designs examining a more comprehensive set of dimensions, both monitoring and motivational incentives, would make it possible to assess more completely the techniques and patterns of control that lead to successful headquarters-foreign subsidiary relationships. 

REFERENCES

 Balkin, D. B., & Gomez-Mejia, L. R. 1987. Toward a contingency theory of compensation strategy. Strategic Management Journal, 8: 169-182. 

 Balkin, D. B., & Gomez-Mejia, L. R. 1990. Matching compensation and organizational strategies. Strategic Management Journal, 11: 153-169. 

 Bartlett, C. A., & Ghoshal, S. 1989. Managing across borders: The transnational solution. Boston, MA: Harvard Business School Press. 

 Belsley, D. A., Kuh, E., & Welsch, R. E. 1980. Regression diagnostics. Cambridge, MA: Cambridge University Press. 

 Conlon, E. J., & Parks, J. M. 1990. Effects of monitoring and tradition on compensation arrangements: An experiment with principal-agent dyads. Academy of Management Journal, 33: 603-622. 

 Crookell, H. 1990. Canadian-American trade and investment under the free trade agreement. Westport, CT: Quorum Books. 

 Cvar, M. 1984. Competitive strategies in global industries. Unpublished dissertation. Harvard Business School. 

 Doty, D. H., Glick, W. H., & Huber, G. P. 1993. Fit, equifinality, and organizational effectiveness: A test of two configurational theories. Academy of Management Journal, 36: 1196-1250. 

 Drazin, R., & Van de Ven, A. H. 1985. Alternative forms of fit in contingency theory. Administrative Science Quarterly, 30: 514-539. 

 Eaton, J., & Rosen, H. S. 1983. Agency, delayed compensation, and the structure of executive remuneration. Journal of Finance, 38: 1489-1505. 

 Edstrom, A., & Galbraith, J. R. 1977. Transfer of managers as a coordination and control strategy in multinational organizations. Administrative Science Quarterly, 22: 248-263. 

 Egelhoff, W. G. 1988. Organizing the multinational enterprise. Cambridge, MA: Ballinger. 

 Eisenhardt, K 1988. Agency and institutional-theory explanations: The case of retail sales compensation. Academy of Management Journal, 31: 488-511. 

 Eisenhardt, K. 1989. Agency theory: An assessment and review. Academy of Management Review, 14: 57-74. 

 Erez, M. & Earley, P. C. 1993. Culture, self-identity and work. NY: Oxford University Press. 

 Fama, E. F., & Jensen, M. C. 1983. Agency problems and residual claims. Journal of Law and Economics, 26: 327-349. 

 Finkelstein, S., & Hambrick, D. C. 1989. Chief executive compensation: A study of the interaction of markets and political processes. Strategic Management Journal, 10: 121-134. 

 Fisher J., & Govindarajan, V. 1992. Profit center manager compensation: An examination of market, political and human capital factors. Strategic Management Journal, 13: 205-217. 

 Galbraith, C. S., & Merrill, G. B. 1991. The effect of compensation program and structure on SBU competitive strategy: A study of technology-intensive firms. Strategic Management Journal, 12: 353-370. 

 Galbraith, J. 1987. Organization design. In J. Lorsch (Ed.), Handbook of organizational behavior, 343-357. Englewood Cliffs, NJ: Prentice-Hall. 

 Gerhart, B., & Milkovich, G. T. 1990. Organizational differences in managerial compensation and financial performance. Academy of Management Journal, 33: 663-691. 

 Ghoshal, S., & Nohria, N. 1989. Internal differentiation within multinational corporations. Strategic Management Journal, 10: 323-337. 

 Gomez-Mejia, L. R. 1992. Structure and process of diversification, compensation strategy, and firm performance. Strategic Management Journal, 13: 381-397. 

 Gomez-Mejia, L. R., & Balkin, D. B. 1992a. Compensation, organizational strategy, and firm performance. Cincinnati: South-Western Publishing Co. 

 Gomez-Mejia, L. R., & Balkin, D. B. 1992b. Determinants of faculty pay: An agency theory perspective. Academy of Management Journal, 35: 921-955. 

 Gomez-Mejia, L. R., & Welbourne, T. 1992. Compensation strategies in a global context. Human Resource Planning, 14 (1): 29-42. 

 Govindarajan, V. 1988. A contingency approach to strategy implementation at the business unit level: Integrating administrative mechanisms with strategy. Academy of Management Journal, 31: 828-853. 

 Govindarajan, V., & Fisher, J. 1990. Strategy, control systems, and resource sharing: Effects on business-unit performance. Academy of Management Journal, 33: 259-285. 

 Gregersen, H. B., & Black, J. S. 1992. Antecedents to commitment to a parent company and a foreign operation. Academy of Management Journal, 35: 65-90. 

 Gresov, C. 1989. Exploring fit and misfit with multiple contingencies. Administrative Science Quarterly, 34: 431-453. 

 Gupta, A. K., & Govindarajan, V. 1986. Resource sharing among SBUs: Strategic antecedents and administrative implications. Academy of Management Journal, 29: 695-714. 

 Hambrick, D. C., & Finkelstein, S. 1987. Managerial discretion: A bridge between polar views of organizational outcomes. In B.M. Staw and L. Cummings (Eds.) Research in Organizational Behavior, vol. 8: 369-406. Greenwich, CT: JAI Press. 

 Hodgetts, R. M, & Luthans, F. 1993. U.S. multinationals' compensation strategies for local management: Cross-cultural implications. Compensation and Benefits Review, 25: 42-48. 

 Hofstede, G. 1980. Culture's consequences: International differences in work-related values. Beverly Hills: Sage Publication. 

 Hofstede, G. 1991. Cultures and organizations: Software of the mind. London: McGraw Hill. 

 Holmstrom, B. 1979. Moral hazard and observability. Bell Journal of Economics. 10: 74-91. 

 Jensen, M., & Meckling, W. H. 1976. Theory of the firm: Managerial behavior, agency costs, and ownership structure. Journal of Financial Economics, 3: 305-360. 

 Johanson, J., & Vahlne, F. 1990. The mechanism of internationalisation. International Marketing Review. 7: 11-24. 

 Kazanjian R. K., & Drazin, R. 1989. An empirical test of a stage of growth progression model. Management Science, 35: 1489-1503. 

 Kobrin, S. 1991. An empirical analysis of the determinants of global integration. Strategic Management Journal, 12 (special issue): 17-31. 

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

 Kosnik, R. D., & Bettenhausen, K. L. 1992. Agency theory and the motivational effect of management compensation: An experimental contingency study. Group of Organization Management, 17: 309-330. 

 Lewellen, W. G., & Huntsman, B. 1970. Managerial pay and corporate performance. American Economic Review, 60: 710-720. 

 Morrison, A. J., & Roth, K. 1992. A taxonomy of business-level strategies in global industries. Strategic Management Journal, 13: 399-418. 

 Mowday, R. T., Porter, L. W., & Steers, R. M. 1982. Employee-organization linkages: The psychology of commitment, absenteeism, and turnover. New York: Academic Press. 

 Nilakant, V., & Rao, H. 1994. Agency theory and uncertainty in organizations: An evaluation. Organization Studies, 15: 649-672. 

 Nohria, N., & Ghoshal, S. 1994. Differentiated fit and shared values: Alternatives for managing headquarters- subsidiary relations. Strategic Management Journal, 15: 491-502. 

 O'Reilly, C. A., III, & Chatman, J. 1983. Organizational commitment and psychological attachment: The effects of compliance, identification, and internalization on prosocial behavior. Journal of Applied Psychology, 71: 492-499. 

 Ouchi, W. 1979. A conceptional framework for the design of organizational control mechanisms. Management Science, 25: 833-848. 

 Prahalad, C. K., & Doz, Y. 1987. The multinational mission: Balancing local demands and global vision. NY: The Free Press. 

 Rajagopalan, N., & Finkelstein, S. 1992. Effects of strategic orientation and environmental change on senior management reward systems. Strategic Management Journal, 13: 127-142. 

 Rosenzweig, P. M., & Nohria, N. 1994. Influences on human resource management practices in multinational corporations. Journal of International Business Studies, 25: 229-251. 

 Roth, K. 1995. Managing international interdependence: CEO characteristics in a resource-based framework. Academy of Management Journal, 38: 200-231. 

 Roth, K., & Morrison, A. J. 1992. Implementing global strategy: Characteristics of global subsidiary mandates. Journal of International Business Studies, 23: 715-735. 

 Roth, K., & Ricks, D. 1994. Goal configuration in a global industry context. Strategic Management Journal, 15: 103-120. 

 Roth, K., Schweiger, D., & Morrison, A. J. 1991. Global strategy implementation at the business unit level: Operational capabilities and administrative mechanisms. Journal of International Business Studies, 22: 369-402. 

 Rugman, A., & Bennett, J. 1982. Technology transfer and world product manufacturing in Canada. Columbia Journal of World Business, Winter: 58-62. 

 Saavedra, R., Earley, P.C., & Van Dyne, L.V. 1993. Complex interdependence in task-performing groups. Journal of Applied Psychology, 78: 61-72. 

 Shavell, S. 1979. Risk sharing and incentives in the principal and agent relationship. Bell Journal Economics, 10: 55-73. 

 Thompson, J. D. 1967. Organizations in action. NY: McGraw-Hill. Tosi, H. L., & Gomez-Mejia, L. R. 1989. The decoupling of CEO pay and performance: An agency theory perspective. Administrative Science Quarterly, 34: 169-189. 

 Venkatraman, N., & Prescott, J. E. 1990. Environment-strategy coalignment: An empirical test of its performance implications. Strategic Management Journal, 11: 1-23. 

 Zajac, E. J., & Westphal, J. D. 1994. The costs and benefits of managerial incentive and monitoring in large U.S. Corporations: When is more not better? Strategic Management Journal, 15: 121-142.