Abstract
Cross-sectional studies of organisation have yielded a number of relationships between contingencies and structure. However, some diachronic studies of changes in contingency and structure have not produced the results expected. The suggestion of this paper is that this may reflect inappropriate forms of data analysis based on a theoretical orientation of contingency determinism rather than contingency-fit. The former posits that a change in contingency directly causes a change in structure. The latter states that it is the lack of fit, a mismatch between contingency and structure, which leads to structural change. An illustration of the two alternative approaches is given by reference to longitudinal data on changes in strategy and structure. Two hypotheses derived from the contingency-fit model are confirmed, whereas the one derived from the contingency determinism approach is not.