THE CONTRIBUTION OF STATISTICS
The review of current theories in Chapter I barely mentions the type of work upon business cycles which is most characteristic of the present and seems most promising for the future--analysis of statistical data. A few of the theorists--notably Henry L. Moore-- make skilled and elaborate use of quantitative methods, and almost all cite statistical evidence upon occasion. But most theories of business cycles are still built up by methods which would have seemed familiar to Sismondi and Ricardo.
On the other hand, there has recently appeared a group of business-cycle statisticians who as yet have sought, not to construct general theories, but to establish more precisely the facts concerning cyclical fluctuations in particular economic processes. By their detailed researches, the statistical workers are building up a literature more like the current literature of the natural sciences than like that of economic theory. It contains few treatises, but a multitude of technical papers; it is mathematical in form and empirical in spirit; it deals with restricted problems, lays stress upon measurements, and aspires to prediction.
Between these two groups of workers, the theorists and the statisticians, there has been less communion than their mutual interests require. Many of the statisticians pay little heed to current theories of business cycles, and many of the theorists make little use of statistical methods. A similar divergence of outlook, associated with a similar division of labor, seems not uncommon in modern science. Experimentalists and pure theorists often have difficulty in understanding each other; but in the long run each group provides grist for the other's mill, and scientific progress is a joint product of the____________________