Structural Contingency Theory Revisited: Main Effects Versus Interactions in Child’s National Study of Manufacturing and Service Kenneth L. Kraft George M. Puia zyxwvutsr The University of Tampa Jerald Hage University of Maryland zyxwvutsrq Abstmci The paper suggests the value in returning to the original idea in contingency theory-direrent structural contin- gencies produce digerent kinds of performance. A com- prehensive multi-industry data set was reanalyzed com- paring structural and contextual main eflects with context-structureinteractions for each of six performance variables. A number of significant main eflects and in- teractions were found to correlate with economic and so- cial performances in both manufacturing and service in- dustries. It zyxwvutsrq ,% concluded that there zyxwvut is support for both universal and contingent relationships. Rt?SW?@? zyxw Le bulletin suggire l’intkrgt du retour a l’idke originale de la thkorie de contingence-diflirentes contingences structurelles produisent dflkrentes zyx sortes de perfor- mances. Les donnks comprkhensives d’une industrie mul- tipleftirent reanalyskes en comparant les rksultats actuek structuraux et leurs contextes avec les interactions struc- ture-contexte de chacune des six diflirentes variables performances. Un nombre dkflets d ’importance princi- pale et &interactions ftit dkcouvert Ctre en corrklation avec les performances tconomiques et sociales aussi bien dans les industries &fabrication que dam les services. D’ou zyxwv oC la conclusion d’une kvidence supportant le rap- port universe1 et contingent. The objective of zyxwvutsr this paper is to suggest that it is worth returning to the original idea in contingency theory, namely, that different structural contingencies-regardless of the names given them, such as mechanistic or or- ganic-produce different kinds of performances. Organ- izational theory oscillated away from contingency per- spectives during the 1970s and 1980s (see Astley & Van de Ven, 1983; Hannan & Freeman, 1975; Pfeffer, 1981) before some of the most promising ideas were adequately researched. Much of the empirical research on contin- gency theory focused on the relationships between context and structure, neglecting measures of performance (Dal- ton, Todor, Spendolini, Fielding, & Porter, 1980). In ad- Portions of this paper were presented at the Southern Management Association Meeting, 1988. Address all corespondence to Kenneth L. Kraft, Department of Management, College of Business, The University of Tampa, Tampa, FL 33606, USA. OASAC 1995 182 dition, in the few studies that have examined contingency theory, there has been a confusion of the unit of analysis, e.g., corporate headquarters vs. the division. There has also been considerable variation in the definition of con- tingencies, with few studies focusing on the central issue of structural contingencies. For some major exceptions, see Donaldson (1985) and Khandwalla (1977). These studies were limited to the corporate headquarters level. Furthermore, there has been a major conceptual problem in the research on s p e d k organizations zy within a particular market, an error that has not been discussed, to our knowledge. The social and economic outcomes of an organization are strongly influenced by the kinds of machinery and human capital that are used, that is, the knowledge base of the organization. Clearly, some industries have much higher economies of scale than oth- ers, and this has a powerful impact on their performances. This kind of control has not been included in prior re- Revue canadienne des sciences de l’administration Canadian Journal of Administrative Sciences u3), 182-194