Peering at the Past Century's Corporate Strategy Through the Looking Glass of Time-Series Analysis : Extrapolating from Chandler's Classic Mid-Century American Firms?

This study has a dual thrust. Substantively, it revisits Chandler's pragmatic fit-performance directive (Chandler's 'efficiency thesis') or his assertion that firms whose structure matches their strategy become more effective than mismatched firms. It is important to revisit the empirical origin of this result in view of its gradual international extrapolation beyond the time and place in which it originated. By gathering the financial data relevant to the firms cited as examples by Chandler, we identify whether the change to the multidivisional structure did indeed lead to improved financial performance in the mid-twentieth century American firms he described. Methodologically, it explores a novel approach to the empirical validation of basic theories. Using historical replication, it undertakes a longitudinal, time-series analysis of a classic theory based on twentieth-century data, and thus investigates the testing of classic data with a modern tool. We undertake multiple replications. Three separate longitudinal studies are performed, consisting of two forecasting methods applied to 11 individual time series, and a comparison technique applied to the same. These are used with the three most common measures of performance at the time of Chandler's writing. All three methods reveal that the differences predicted by his theory are not borne out by the longitudinal analysis of a core group of Chandler's own exemplary firms. The three sets of longitudinal analyses we present raise some substantive questions regarding this cornerstone of classical theory, and carry positive methodological portents regarding the use of historical replication as a stepping-stone for twenty-first-century research