The Ultimate Foundation of Economic Science
Ludwig von Mises
Necessity and Volition
Statistics is the description in numerical terms of experiences concerning phenomena not subject to regular uniformity. As far as there is discernible regularity in the succession of phenomena, no recourse to statistics is needed. The objective of vital statistics is not to establish the fact that all men are mortal, but to give information about the length of human life, a magnitude that is not uniform. Statistics is therefore a specific method of history.
Where there is regularity, statistics could not show anything else than that A is followed in all cases by P and in no case by something different from P. If statistics show that A is in X% of all cases followed by P and in (100-X)% of all cases by Q, we must assume that a more perfect knowledge will have to split up A into two factors B and C of which the former is regularly followed by P and the latter by Q.
Statistics is one of the resources of historical research. There are in the field of human action certain occurrences and events characteristic features of which can be described in numerical terms. Thus, e.g., the impact of a definite doctrine upon the minds of people does not permit of any numerical expression. Its "quantity" can be ascertained only by the method of the specific understanding of the historical disciplines. But the number of people who lost their lives in struggles to arrange, by means of wars, revolutions, and assassinations, social conditions in agreement with a definite doctrine can be precisely determined in figures if all the documentation required is available.
Statistics provides numerical information about historical facts, that is, about events that happened at a definite period of time to definite people in a definite area. It deals with the past and not with the future. Like any other past experience, it can occasionally render important services in planning for the future, but it does not say anything that is directly valid for the future.
There is no such thing as statistical laws. People resort to the methods of statistics precisely where they are not in a position to find regularity in the concatenation and succession of events. The most celebrated statistical achievement, mortality tables, does not show stability, but changes in the mortality rates of the population. The average length of human life changes in the course of history, even if no changes were to emerge in the natural environment, because many factors that affect it are the result of human action, e.g., violence, diet, medical and prophylactic measures, the supply of foodstuffs, and others.
The concept of "statistical law" originated when some authors, in dealing with human conduct, failed to realize why certain statistical data change only slowly and, in blind enthusiasm, hastily identified slowness of change with absence of change. Thus, they believed themselves to have discovered regularities—laws—in the conduct of people for which neither they themselves nor anybody else had any other explanation than the—as must be emphasized, baseless—assumption that statistics had demonstrated them. From the shaky philosophy of these authors physicists borrowed the term "statistical law," but they gave to it a connotation that differs from that attached to it in the field of human action. It is not our task to deal with the meaning these physicists and later generations of physicists attached to this term or with the services statistics can render to experimental research and to technology.
The orbit of the natural sciences is the field in which the human mind is able to discover constant relations between various elements. What characterizes the field of the sciences of human action is the absence of constant relations apart from those dealt with by praxeology. In the former group of sciences there are laws (of nature) and measurement. In the latter there is no measurement and—apart from praxeology—no laws; there is only history, including statistics.
 See below, p. 65.
 About the most eminent instance of this doctrine, that of H. Th. Buckle, see Mises, Theory and History, pp. 84 ff.
Previous Page * Next Page
Table of Contents