Abstract
This study presents econometric estimates of the effects of unionism on productivity in 83 West Virginia coal mines in the early 1920s. The size, detail, and panel structure of the data set permit investigation of many possible links between unionism and productivity, in contrast to the summary measures reported by most studies. The results show that the union effect was not uniform across mines and cannot be represented by a simple shift parameter. Rather, unionism significantly reduced productivity at small mines but not at large mines. Drawing on historical evidence, the author ascribes this differential effect to systematic differences between small and large operations in the quality of management and union leadership.