The Determinants and Consequences of Subsidiary Initiative in Multinational Corporations

Beginning of article

This paper examines corporate entrepreneurship in multinational corporations through a detailed study of initiatives taken by foreign subsidiaries. We develop a theoretical model in which two levels of organizational context (corporate and subsidiary) promote or suppress subsidiary initiative, and initiative in turn has a feedback effect on both subsidiary and corporate context. Using a multi-method study (229 questionnaire returns plus 5 in-depth case studies), the key findings are as follows: Subsidiary initiative is promoted by a high level of distinctive subsidiary capabilities, and is suppressed by a high level of decision centralization, a low level of subsidiary credibility, and a low level of corporate-subsidiary communication. Over time, we find evidence that subsidiary initiative leads to an enhancement of credibility (vis-a-vis the head office), head office openness, corporate-subsidiary communication, and distinctive capabilities.

There is a long tradition of research that views corporate entrepreneurship at lower levels of the organization as an important ingredient in the recipe for long-term success. Practitioner oriented books by Peters and Waterman (1982), Kanter (1985), and Pinchott (1986), and a host of academic studies (e.g. Biggadike, 1979; Burgelman, 1991; Guth & Ginsberg, 1990; Hornsby, Naffziger, Kuratko, & Montagno, 1993; Zahra, 1993) have all picked up on the importance of corporate entrepreneurship as a way of challenging the status quo in large organizations, and thereby as a stimulus for corporate renewal and adaptation to environmental change.

Corporate entrepreneurship takes many forms, but one useful distinction is between focused and dispersed corporate entrepreneurship (Birkinshaw, 1997). Focused corporate entrepreneurship occurs in specially created "new venture divisions" whose mandate is to identify and nurture new business opportunities for the corporation (Burgelman, 1983a; Kuratko, Montagno, & Homsby, 1990; Sykes, 1986). Such divisions have little formal structure but provide "patient money" and support for risk taking and creativity (Galbraith, 1982; Kanter, 1985; Kuratko et al., 1990; Quinn, 1985; Sathe, 1985). Dispersed corporate entrepreneurship (or intrapreneurship) occurs throughout the firm. Rather than hiving off separate groups or divisions to be entrepreneurial, the dispersed approach sees the development of an entrepreneurial culture or posture as the key antecedent to initiative (Covin & Slevin, 1991; Ghoshal & Bartlett, 1994; Kanter, 1985; Zahra, 1993). The challenge for corporate management is to instill in its employees th e personal involvement and commitment that drives entrepreneurship.

This paper focuses on the dispersed form of corporate entrepreneurship. Specifically, it examines cases of subsidiary initiative in the context of large multinational corporations. Initiative refers to a discrete, proactive undertaking that advances a new way for the corporation to use or expand its resources (Kanter, 1982; Miller, 1983). [1] A subsidiary initiative is one that begins within a foreign subsidiary unit, rather than the headquarters operations, of a large corporation.

Notwithstanding the obvious importance of environmental and individual-level factors, our approach in this study is to model subsidiary initiative primarily as a function of its organization context. As Burgelman (1983a) and others show, corporate-level elements of context will frequently suppress the occurrence of subsidiary initiative. But there are also subsidiary-level elements of context, which seem more likely to promote initiative. Thus, the first research question can be broadly stated as follows: What facets of organizational context promote or suppress subsidiary initiative? In the second part of this paper we take a longer-term perspective on subsidiary initiative and explore the possibility that over time subsidiary initiative is likely to effect changes in the corporation' s organizational context such that subsequent initiatives are more easily pushed through. The guiding research question can thus be stated as follows: What is the impact of subsidiary initiative on organizational context?

To answer these questions, we first review the literature on organizational context and put forward an integrative model that takes into account the various definitions of context as well as the multiple levels of context evident in large corporations. We then apply this general model to the specific case of subsidiary initiative, and develop a series of hypotheses relating various facets of context to subsidiary initiative. The latter parts of the paper then report on an empirical examination of the hypothesized relationships.

THEORETICAL BACKGROUND

What are the drivers of initiative? Recent integrative models in the field of corporate entrepreneurship have suggested that there are individual, organizational, and environmental components (Covin & Slevin, 1991; Hornsby, et al., 1993; Zahra, 1993; Zahra & Covin, 1995), but even such a simple categorization as this has problems. Individual propensity to act entrepreneurially is a function of motivation (Kets de Vries, 1977; McLelland, 1967), which in turn is a function of both an individual's innate personality and the context in which he or she is working. Moreover, the boundary between organization and environment is far from clear when the organization is large and complex. The "environment" for a subsidiary company, for example, may represent all operations with which the subsidiary interacts, including those that are legally part of the same corporation (Ghoshal & Bartlett, 1990).

The approach taken in this paper is to view subsidiary initiative primarily as a function of organizational context. We define organizational context as the set of administrative and social mechanisms that shape the behaviors of actors in the organization, over which top management have some control. The essence of this definition is that initiative, like any other behavior, is a function of the setting in which it occurs, and that within an organization many of the critical facets of that setting are under the direct or indirect control of top management. Reporting relationships, access to financial resources, reward systems, development programs, and a host of other mechanisms all influence the way an individual behaves. These mechanisms together constitute organizational context (Bower, 1970; Prahalad & Doz, 1981). But the term "context" can also be viewed more broadly, in that the behavior of individuals in a subsidiary is shaped by more than just administrative and social mechanisms within the firm. One important set of cues for a subsidiary firm comes from its local environmental context, i.e. the set of customers, suppliers, competitors, and institutional bodies with which it interacts (Westney, 1994; Ghoshal & Nohria, 1989). Another can best be labelled the subsidiary's resource context (Pierce & White, 1999), which is the configuration of internal or external resources on which the subsidiary depends for its survival.

The Organizational Context Literature

A key part of the definition of organizational context is the recognition that both "administrative" and "social" mechanisms can be used to shape behavior. Two very distinct lines of thinking can be discerned in the literature representing these two sets of mechanisms. They will be referred to as structural context and behavioral context respectively.

Structural context. The term structural context was brought into common use by Bower (1970) in his study of the resource allocation process in a large diversified corporation. Structural context was defined as:

the set of organisational forces that influence the processes of definition (the economic characteristic of a project) and impetus (the force that moves it towards funding). These forces were found to be the elements of corporate structure: the formal organisation, the system of information and control used to measure performance of the business, and the systems used to measure and reward performance of managers (p. 71)

Bower's focus, in other words, was on relatively direct mechanisms for "controlling" behavior, such as the availability of resources and reward and punishment systems. Many subsequent studies extended Bower's original concept of context. Prahalad and Doz (1981), for example, used the broader term "organizational context" to refer to the direct mechanisms identified by Bower as well as some of the more indirect contextual mechanisms such as management development programs and socialization. Burgelman (1983a, 1983b) followed Bower's usage of structural context to develop a model of initiative propogation in the case of an internal corporate venture unit. The influence of structural context on initiative has also been widely studied. Beginning with Burns and Stalker's (1961) study of mechanistic and organic work environments, academic research has frequently drawn a link between the structure of the organizational unit and the presence of corporate entrepreneurship. One popular approach is to create a new ventur e division, which is a semi-autonomous unit with relatively little formal structure and good access to new product development funds (Burgelman, 1983a; Galbraith, 1982; Ginsberg & Hay, 1995; Sykes, 1986). This approach is built on the premise that entrepreneurship is stimulated by very different structural arrangements than ongoing managerial activity, so it needs to be kept separate from the rest of the organization.

Behavioral context. The term behavioral context was recently employed by Bartlett and Ghoshal (1995, p. 12) to refer to the "carefully nurtured, deeply embedded corporate work ethic that triggers the individual-level behaviours of entrepreneurship, collaboration and learning." The central observation in Bartlett and Ghoshal's research, as well as many related studies of organizational culture and climate, is that some organizations manage to instil in their employees an enthusiasm or level of involvement above and beyond that justified by economic rewards alone. Unlike structural context, which manipulates employees through a system of reward and punishment, behavioral context appears to encourage their involvement at an emotional level. Both shape employee behavior, but they do so in rather different ways.

Behavioral context research does not have a clearly defined academic heritage. Ghoshal and Bartlett (1994) were influenced by Weber's thinking on the emergence of a moral order in Western society and Barnard's (1938) Functions of the Executive, one function of which is inspiring the "faith" of employees as a means of generating their willing co-operation. In similar fashion, Kogut and Zander (1997) saw the firm's capacity for inspiring identity and belonging in its employees as a reaction to the disintegration of traditional patterns of social order described by Durkheim (1933) and others. Others have also acknowledged the sociological origins to our thinking on context (e.g. Burns & Stalker, 1961; Kanter, 1985; Thompson, 1968).

The implication of early writings, especially Barnard (1938), is that behavioral context exists as a set of guiding values and beliefs, the development of which is the responsibility of top management. The parallel literature on organizational culture (Denison, 1990; Ouchi, 1981; Pettigrew, 1979; Saffold, 1988; Schein, 1985) takes a similar position. As defined by Denison (1990, p. 2), culture is "the underlying values, beliefs, and principles that serve as a foundation for an organisation's management system as well as the set of management practices and behaviours that exemplify and reinforce those basic principles" (emphasis added). [2]

As with structural context, there is strong support in the corporate entrepreneurship literature for a link between behavioral context and initiative. Kuratko, Montagno, and Hornsby (1990) measured the "entrepreneurial environment" of an organization; Covin and Slevin (1991) referred to "entrepreneurial posture"; and Kanter (1985), Pinchott (1986), and Peters and Waterman (1982) all discussed various facets of organizational culture as a driver of initiative. [3] In sum, behavioral context can be understood as that part of the organization's culture that promotes the willing co-operation of employees (Barnard, 1938). It is, of course, possible for behavioral context to suppress co-operation as well--Bartlett and Ghoshal (1995), for example, discuss both the pathologies of context (control, constraint, compliance, and contract) as well as its desirable elements (support, stretch, trust, discipline). In both cases, behavioral context would appear to be a function of the sum of managerial actions over a long pe riod of time. Unlike structural context, which can be modified very rapidly, behavioral context can only be changed through consistent and deliberate effort on the part of management.

Structural vs. Behavioral Context

While structural and behavioral context are substantially different concepts with different intellectual roots, there have been some attempts at combining the two. Kanter (1985) discussed both the structural and cultural facets of "integrative" organizations; Burns and Stalker's (1961) "organic" organizations, likewise, had structural and behavioral dimensions; and some of the writing on multinational context management has discussed both structural and behavioral elements (e.g. Bartlett & Ghoshal, 1989; Prahalad & Doz, 1981). It would appear that structural and behavioral context are complementary, rather than competing, management approaches. In the context of corporate entrepreneurship, for example, companies such as 3M and Hewlett Packard are well known both for their entrepreneurial culture and for their use of new venture divisions for managing product development.

Another similarity between the two schools of thought is that both see the possibility of initiative having a feedback impact on context. As discussed above, Ghoshal and Bartlett (1994) saw behavioral context evolving in part through the actions of management and in part through the emergence of entrepreneurship, co-operation, and learning as desirable organizational outcomes. The implication here is of a tight reciprocal relationship between context and entrepreneurship. In terms of structural context, the linkage between the two constructs was suggested by Burgelman (1983a): he proposed that autonomous action (i.e. initiative) could, in the long term, bring new ventures within the overall concept of corporate strategy, which in turn would lead to changes in the structural context of the corporation. The implication, in this case, is that the influence of initiative on structural context is relatively tenuous and relatively long term, but present nevertheless.

Other Forms of Context

The literature on multinational corporations frequently emphasizes the point that a subsidiary unit faces competing pressures: for responsiveness to host country demands, and for conformity to corporate norms (Bartlett & Ghoshal, …