JOSEPH A. SCHUMPETER, Harvard University
To protect the following comments from a not unnatural misunderstanding, I want to make it quite clear right away that I have no wish to advocate the historical approach to business cycles at the expense, still less to the exclusion, of theoretical or statistical work. As my own attempts in the field amply prove, I am as much convinced as anyone can be of the necessity of bringing to bear upon the study of business cycles the whole of our theoretical apparatus and not only aggregative dynamic schemata but also our equilibrium analysis. As should be evident from the last sentence, by theory I do not only mean explanatory hypothesis -- as our lamented Mitchell did -- but also and even principally the tools that theory puts at our disposal. Not less am I convinced that statistical investigations and methods (including all the methods that cluster around the handling of time series) are essential means for making headway. Further on I shall indicate how theory and statistics fit in with the historical approach as visualized in this paper. For the moment I only wish to emphasize that nothing is further from my mind than any desire to start the kind of methodological discussion which the logic of the modern scientific situation has definitely made a thing of the past.
Economic life is a unique process that goes on in historical time and in a disturbed environment. For this and other reasons there is an argument for historical or institutional study in almost any department of economics. But I mean something more specific than that. We talk about business cycles as a scourge (some of us call them the main scourge of capitalist life) and discuss possibilities of eliminating them altogether as if this were an unchallengeable end in itself, whereas it does not seem to me open to doubt:____________________