Abstract
With the advent of the widely heralded “new economy”, organizations are undergoing profound changes. The traditional vertically integrated business units seem to serve as the springboard to launch the new organizational forms (e.g., virtual organization, network organization). The basic tenets of the organizational form, viz., strategy, structure and context, have gained renewed vigor as building blocks for the new organizational forms. Structural contingency has served as the foundation concept for organizational design for a long time. This paper offers a methodological path to extend the concept of ‘fit’ where multiple, possibly conflicting contingencies and structural alternatives are holistically examined. The solution is illustrated by: (1) examining the multivariate fit between four context variables and five structure dimensions; and then (2) testing the impact of the derived configurations on performance. The source of data for this illustration of methodological extension to the assessment of ‘fit’ is Umanath & Kim (1992). A pair of operational configurations depicting theorized formulations of organizational designs is derived; this pair can serve as the ‘ideal profiles’ in the assessment of target configurations. Since several target configurations may qualify as equally effective, this methodology supports the concept of equifinality in organizational design solutions.
Motivation
With the advent of the widely heralded “new economy”, organizations are undergoing profound changes. The very nature of the business model which steers organizational thinking is changing. New frames of reference and new stakeholders are emerging. Amidst all this, corporate management is wrestling with new organizational forms (e.g., virtual organization, network organization) to effectively harness the market forces to yield favorable business solutions. Markets and contracts, rather than hierarchies, are becoming essential co-ordination tools. Information systems and electronically mediated work groups seem to catalyze boundary spanning among traditional vertical organizations. Nonetheless, the traditional vertically integrated business units continue to thrive and seem to serve as the springboard to launch the new organizational forms. The basic tenets of the organizational form, viz., strategy, structure and context, have gained renewed vigor in the role of building blocks for the new organizational paradigm as managers grapple with organizational des ign i s sues for inter – organizational contexts in the decision making environment of contemporary business practice.
The concept of structural contingency has held the attention of researchers in the organizational design discipline for a couple of decades. The basic theme in structural contingency is that context and structure must somehow ‘fit’ together in order for an organization to perform well (Drazin & Van de Ven, 1985: 514). However, it is argued that there are several perspectives of the concept of ‘fit’ (e.g., Fry & Smith, 1987; Van de Ven & Drazin, 1985; Venkatraman, 1989). Traditional approaches assess ‘fit’ on the basis of bivariate relationships, as if context and structure variables covaried in independent pairs. In contrast, advocates of the ‘systems approach’ to structural contingency (Miller, 1981; Van de Ven & Drazin, 1985) contend that the perspectives that focus on how single contextual factors affect single structural characteristic and how these pairs of context and structure factors interact to explain performance are reductionist in nature. The anatomy of an organizational unit, asserts Drazin & Van de Ven (1985), is complex and its richness can be better captured as a ‘gestalt’, that is, by simultaneously addressing the many contingencies, structural alternatives and performance criteria in a holistic fashion. Simultaneous consideration of multiple, sometimes, conflicting contingencies, also rejects the notion of a ‘single best structure’. ‘Equifinality’, that is, the existence of several feasible, equally effective structural design options for any given context is, thus, central to the systems approach (Van de Ven & Drazin, 1985).
The focus of this paper is on the systems approach for the assessment of fit, which advocates holistic assessment of structural contingency. Examining pairs of representative variables from the theoretical domains of, say, context and structure, and their impact on performance are fine, as long as the inferences made are confined to the pairs of variables studied. However, studying several such pairs in isolation and extrapolating the findings to inferences associating the root domains (e.g., context and structure) is a ‘leap of faith’ that is problematic on grounds that the resulting conclusions ignore the multivariate/holistic aspects of relationships (Miller, 1986; Van de Ven & Drazin, 1985). Miller (1981: 3) argues that dynamic interactions among variables of environment, organizational structure and strategy generate organizational configurations or adaptive patterns that are “… expected to have tightly interdependent and mutually supportive parts, the significance of which can be best understood by referring to the whole.” This idea of the whole is more than the sum of the parts was earlier expressed by Khandwala (1973: 493) when he contended that “… gestalt or configuration of an organization is likely to be a more potent determinant of its effectiveness than any of the individual components of this configuration, and particularly so if the configuration is the “right” one; that is, if it fits the firm’s situation.”
Drazin & Van de Ven (1985) proposed the pattern analysis technique as a way to evaluate contextstructure configurations under the auspices of systems approach. Procedures like cluster analysis and q-factor analysis are capable of formulating multivariate fit. However, these are essentially inductive approaches, exploratory in nature and only provide an implicit notion of fit; whereas, the deductive pattern-analytic procedure(s) provide an explicit specification and testing of a particular conceptualization of fit (Venkatraman & Prescott, 1990: 5). While certainly a step towards implementation of the systems approach, the pattern analysis technique is not without limitations. In this paper, we propose a methodological extension that can supplement the pattern analysis approach.
