THE ROLE OF STRATEGIC CHOICE
Abstract This paper critically examines available theoretical models which have been derived from statistically established patterns of association between contextual and organizational variables. These models offer an interpretation of organizational structure as a product of primarily economicconstraints which contextual variables are assumed to impose. It is argued that available models in fact attempt to explain organization at one remove by ignoring the essentially political process, whereby power-holders within organizations decide upon courses of strategic action. This 'strategic choice' typically includes not only the establishment of structural forms but also the manipulation ofenvironmental features and the choice of relevant performance standards. A theoretical re-orientation of this kind away from functional imperatives and towards a recognition of political action is developed and illustrated in the main body of the paper.
SYSTEMATIC comparative investigation of the relationships between organizational structure and situational variables has been the guiding principlefor major research programmes both in the United States, under Blau, Hage and Aiken, Hall, Lawrence and Lorsch, and in Britain under Pugh and Woodward. In their work, referenced at the close of this paper, these researchers have attempted to discover the degree of empirical variation in organizational structures and to establish the conditions of such variation. Their findings, together with thosefrom other less extensive studies, provide the material from which models of structural determination have been constructed. This procedure is regarded as essential to the development of organization theory; as Blau has put it, 'only systematic comparisons of many organizations can establish relationships between characteristics of organizations and stipulate the conditions under which theserelationships hold, thereby providing the material that needs to be explained by theoretical principles and important guides for deriving these principles'. (1965: 338)
However, research designed to establish statistically the presence of associations between organizational characteristics usually leaves underlying processes to be inferred. An example is the attempt by Pugh and his colleagues toconstruct from factorial data a causal sequence of organization development (Pugh et al. 1969b). The difficulty here is that adequate explanation derives from an understanding of process, and in this regard the 'fact' of a statistically established relationship does not 'speak for itself'. At the very least, it may mask a more complex set of direct and
from the SAGE Social Science Collections.All Rights Reserved.
indirect relationships as Blalock (1969) points out. In addition, little understanding is afforded as to how the relationship was established and whether it is a necessary condition for the presence of other, perhaps desirable, phenomena. For these reasons, not only is research into organization of a processual and change-oriented type still required but soequally is an attempt to offer more adequate theoretical schemes in step with the advance of empirical research. At the present time, some of the most influential models of organization explicate little more than positively established associations between dimensions of organizational structure and 'contextual' (i,e. situational) factors such as environment, technology or scale of operation. Thesemodels proceed to the simplest theoretical solution which is that the contextual factors determine structural variables because of certain, primarily economic, constraints the former are assumed to impose.
It is the purpose of this paper to argue that this simple theory is inadequate, primarily because it fails to give due attention to the agency of choice by whoever have the power to direct...