Epistemological Problems of Economics

We call the method of scientific work that examines the effect, ceteris paribus, of change in one factor, the static method.74  Nearly everything that sociology and its hitherto best developed branch, economics, have thus far accomplished is due to the use of the static method. The assumption it makes, viz., that all other conditions remain perfectly unchanged, is an indispensable fiction for reasoning and science. In life everything is continually in flux, but for thought we must construct an imaginary state of rest.75  In this manner we conceptually isolate the individual factors in order to be able to study the effect of changes in them. The word “static” should not prevent us from seeing that the method in question is one whose goal is precisely the investigation of change.76

In the present state of the science, it is not yet possible to determine whether dynamic laws are feasible within the system of catallactics. A dynamic law would need to be able to show how changes would have to occur on the basis of forces acting within the static system even though no change in the data took place from without. It is well known that Ricardo and many epigones of the classical school—even Marx, for example—undertook such attempts, and that similar efforts have been made on the basis of modern science as well. We need not go into this more deeply at this time. Nor need we be concerned here with the question whether laws of sociological dynamics could be demonstrated to hold outside the narrow frame of economic theory. We must adhere to the notion of the dynamic law only in order to contrast it to the notion of the historical law.77

The formulation of historical laws, i.e., laws of historical change, has repeatedly been designated as the task of history. Many even set out to formulate such laws. Of course, these laws did not meet the demands one must make of a scientific law. They lacked universal validity.

In all these “laws,” as, for example, in Breysig’s, of which we have given an example above, the basis of this deficiency lies in the fact that ideal types were used in the construction of the law. Inasmuch as ideal types do not possess universal validity, propositions involving them must be similarly deficient. All the concepts encountered in the thirty-first law of Breysig, which has already been quoted, are to be viewed as ideal types. Not only are “rule of the Kaiser,” “rule of the people,” and “boom in trade and industry” to be understood in this way, but also “national economy” in the sense in which this term is employed by Breysig.

Laws of historical stages occupy a special position. Stages of historical development arranged in a series are delineated as ideal types, and then the statement is made that history consists in the progression from one stage to the next, and thence on to the third, and so on. It is obvious that as long as the necessity of such a progression cannot be established, this does not yet signify the demonstration of a conformity to law.78  If, however, the progression is maintained to be necessary, then this pronouncement, but not the ideal-typical constructions of the stages, would have to be regarded as a law, although only if its content were free of every reference to ideal types.

The laws of progress seek to satisfy this requirement. They trace the operation of one or several forces to whose permanent action they unequivocally attribute the direction in which social changes take place. Whether this development leads to good or evil, whether it signifies improvement or decline, is immaterial. Progress means here: progression on the necessary path. Now, it is, of course, true that all laws of progress hitherto formulated, in so far as they are not to be rejected from the outset as fictions in no way corresponding to reality, lose the strict character of law through their connection with ideal-typical constructions. Yet it would not be difficult to enucleate clearly the sociological law underlying each of them and to verify it. Even if we were then to deny that the historical law is a law, we should nevertheless find in it a law of sociological dynamics.

Work performed under the division of labor is more productive than isolated work. The same expenditure of labor and of goods of higher order produces a greater quantity of output and enables feats to be accomplished that an isolated worker would never be in a position to achieve. Whether or not this proposition of empirical technology and the physiology of labor is valid without exception—as far as we are at all warranted in speaking of absolute validity in the case of an empirical law—is of no importance for us, since, in any case, it is certain that only one or two instances, if any, can be cited, and then only with difficulty, for which it would not be valid. The increase in productivity brought about by the division of labor is what gives impetus to the formation of society and to the progressive intensification of social cooperation. We owe the origin and development of human society and, consequently, of culture and civilization, to the fact that work performed under the division of labor is more productive than when performed in isolation. The history of sociology as a science began with the realization of the importance for the formation of society of the increase in productivity achieved under the division of labor. However, sociology in general, and economics in particular, have viewed the law of the division of labor not as a constituent part of their own structure of thought, but as a datum, though one which is almost always—or, for all practical purposes, always—present. It is instructive to see how the Historical School sought to arrive at a “historical law” in this case.

Bücher’s theory of stages wants to comprehend “all economic development, at least that of the Central and Western European nations, where it can be historically traced with sufficient accuracy” under a “principle of central significance for understanding the essential phenomena of the economy.” The theory finds this principle in the relation in which the production of goods stands to their consumption. Specifically, it is discernible in the length of the route that goods must travel in passing from producers to consumers. Hence follows his division into the three stages of the self-sufficient household economy, the town economy, and the national economy.79

We shall not dwell on the fact that each one of the three stages is delineated, and can be delineated, only as an ideal type. This is a shortcoming characteristic of all these “laws.” What is noteworthy is only that the freedom with which the historian may construct ideal types enables Bücher to reject the obvious idea, apparently displeasing to him for political reasons, that “mankind is on the point of rising to a new stage of development, which would have to be contrasted to the three previous stages under the name of world economy.”80  However, it cannot be our task to point out all the minor weaknesses and flaws in Bücher’s schematization. What concerns us is exclusively the logical form, and not the concrete content, of the theory. All that Bücher is in a position to state is that in the course of historical development up to the present three stages are to be distinguished. He is unable to give us any information about the causa movens of the changes that have occurred hitherto or about future developments. One cannot understand how Bücher, on the basis of his theory, comes to call every succeeding stage the “next higher” in relation to the preceding one, or why he assumes without hesitation that “the transition from the national economy to the next higher stage . . . will come,” while expressly adding that one cannot know how “the economic future will look in detail.”81  The metaphorical use of the term “stage” need Dot have led him to say “higher” stage instead of merely “succeeding” stage; and on the basis of his theory nothing can warrant his predicting that any further change will take place, much less his confident assurance that such a change could not consist in a regression to one of the previous stages. Consequently, it is impossible to see a “law” in a theory of this kind; and Bücher rightly avoids designating it as such.82

A question, however, which is in any case much more important than whether or not one is dealing with a “law” here is whether the construction of such schemata is useful for the enlargement and deepening of our knowledge of reality.

We must answer this question in the negative. The attempt to force economic history into a concise schema is not only without value for cognition, as we see from the remarks above; it has an effect nothing short of detrimental. It was responsible for Bücher’s failure to see that a shortening of the route that goods traveled in passing from producers to consumers occurred in the later Roman Empire precisely as a result of the decline in the division of labor. The dispute about whether or not the economy of the ancients is to be viewed as a self-sufficient household economy may appear idle to us when we reject Bücher’s, as we do every similar, schematization. Yet if one does not wish to close one’s mind to the possibility of understanding one of the greatest changes in history, the decline of ancient civilization, one must not fail to appreciate the fact that antiquity had gone further in the division of labor—or, to use Bücher’s own words, in “the length of the route that goods travel in passing from the producers to the consumers”—than the first centuries of the Middle Ages. The realization of the higher productivity of work performed under the division of labor places at our disposal the indispensable means for the construction of the ideal types necessary for the intellectual comprehension of this event. In this respect, the concepts of the self-sufficient household economy (production solely for one’s own consumption, the exchangeless economy), the town economy (production for a clientele), and the national economy (commodity production) may prove their usefulness as ideal types appropriate to the subject matter. The decisive and fateful error lies not in their construction, but in the attempt to connect them with a schema of stages and to base this schema on the law of the division of labor.

It was therefore with good reason that Bücher refrained from any attempt to base his theory of stages on the law of the higher productivity of work performed under the division of labor. This law makes only one statement about the objective result that can be attained through the division of labor. It does not say that the tendency toward further intensification of the division of labor is always operative. Whenever and wherever an economic subject is confronted with the choice between a procedure employing a more intensive and one employing a less intensive division of labor, he will adopt the former, provided that he has also recognized the objectively greater output that he can thereby attain and provided also that he values this difference in output more highly than the other consequences which, perhaps, are bound up with the transition to a more intensive division of labor. However, the law as such can make no statement about whether or to what extent this recognition does in fact take place. It can teach us to comprehend and explain causally a change that has already taken place, whether it be in the direction of a more intensive or of a less intensive development of the division of labor, but the law cannot show us why or even that the division of labor must always be more intensively cultivated. We are able to arrive at this conclusion only on the basis of an historical judgment—that is, one formed with the conceptual means at the disposal of history—of what peoples, groups, and individuals want under the influence of the factors determining their existence: their inborn qualities (racial inheritance) and their natural, social, and intellectual environment.

However, we do not know how these external factors are transformed within the human mind to produce thoughts and volitions directed and operating upon the outer world. We are able to ascertain this only post factum, but in no way can we deduce it in advance from a known regularity formulated as a law. Hence, we cannot infer from the law of the division of labor that the division of labor must always make further progress. The division of labor may again be set back temporarily or even permanently. A government may be dominated by an ideology that sees its social ideal in the reversion to autarky. One may consider this quite improbable, but one cannot make a clear and definite prediction about it, for the reasons which have already been given. In any case, one must not overlook the fact that today an ideology hostile to the international division of labor is beginning to exercise a great influence upon the foreign economic policy of many nations.

The law of the division of labor does not belong to the universally valid system of a priori laws of human action. It is a datum, not an economic law. For that reason it appears impossible to formulate on its basis an exact law of progress, i.e., a law free of ideal-typical constructions. On this point the optimists among the liberal sociologists of the Enlightenment, who were confident of progress and who were always reproached with “defective historical intelligence,” were logically much more correct than their critics. They never denied that they based their firm belief in continual social progress not on “laws,” but on the assumption that the “good” and the “reasonable” must ultimately prevail.

The same shortcomings can be shown in every attempt to construct a theory of historical stages. Underlying all such theories are generally, though not always, observations and discoveries that are correct in themselves. But the use that these theories make of them is impermissible. Even where the experience to which they refer does not exhibit merely a nonrepeatable succession of phenomena, these theories go far beyond what is logically legitimate. Before the beginnings of an independent social science, historians were aware of the importance of proper location for productivity. Since the conditions that make locations appear more or less favorable undergo change, one acquires a means of historically explaining shifts of location and migratory movements. On the other hand, the theories of geographical stages, entirely apart from the fact that they present the law of location in the most crudely oversimplified and inadequate way, render access to the understanding of these problems only more difficult. Hegel maintained:

World history goes from East to West; for Europe is obviously the end of world history, and Asia, the beginning. While the “East” in itself is something quite relative, there exists for world history an East xατ εξοχην; for, although the earth is a sphere, history, nevertheless, does not travel in a circle around it, but has, on the contrary, a determinate East, viz., Asia. Here rises the external, physical sun, and in the West it sinks down; in compensation for which, however, the inner sun of self-consciousness, which diffuses a nobler splendor, rises here.83

And according to Mougeolle, there is a “law of altitudes,” namely, that in the course of history the city is increasingly forced down into flat land by the mountains; and a “law of latitudes,” to the effect that civilization has always moved from the tropics toward the poles.84  In these laws too we find all the shortcomings that attach to every theory of historical stages. The causa movens of the changes is not shown, and the accuracy of the geographical concepts that they contain cannot conceal the fact that for the rest they are based on ideal-typical constructions, and indeed on such as are uncertain and therefore unusable, like “world history” and “civilization.” But still more serious by far is the fact that without any hesitation they leap from the statement of the law of location to a volition uniquely determined by it.

Becher accounts as follows for his opinion that the possibility of historical laws cannot be denied in principle:

One did not want to admit historical laws as such because they are of a secondary, reducible, and derivative nature. This rejection rests upon an unsuitable, narrowly conceived notion of law, which, if applied consistently to the natural sciences, would compel us to deny the title of natural laws to many relationships that everyone designates as such. For most of the laws of natural science—e.g., the laws of Kepler, the laws of wave theory concerning resonance, interference, and so on, and the geometric-optical laws of the effect of concave mirrors and lenses—are of a secondary and derivative character. They can be traced back to more fundamental laws. The laws of nature are no more all ultimate, irreducible, or fundamental than they are all elementary, i.e., laws of elementary, not complex phenomena. . . . However, if this designation is quite generally conferred on numerous “laws” of natural science which are neither fundamental nor elementary, then it will not do to deny it to historical laws simply because they are not fundamental or elementary in character.85

In my opinion, this argument does not get to the heart of the matter. The question is not whether the designation “law” is to be applied only to fundamental or elementary regularities. This, after all, is an unimportant question of terminology. In and of itself, it would not be impossible, although inexpedient in the greatest measure and disregardful of all economy of thought, to formulate the laws of acoustics as statements about concerts rather than sound waves. However, it would certainly not be possible to include in these laws, if they are to retain the character of laws of natural science, statements about the quality and expression of the musical performance. They would have to confine themselves to what can be described by the methods of physics. We are unable to include the entire course of historical phenomena in laws, not because they are complicated and numerous or because factors and conditions independent of one another are involved in them, but because they include also factors whose role we are unable to determine precisely. The concepts of sociology extend as far as exactness is possible in principle. On the other side of these boundaries lies the domain of history, which, by means of ideal types, fills with the data of historical life the frame provided by sociology.

• 74The distinction between statics and dynamics as I conceive it differs from the distinction as Amonn conceives it. This difference cannot be gone into more thoroughly here. However, I must, of course, call particular attention to what Amonn says regarding the entirely different meaning that attaches to these conceptual correlates in mechanics and in economics. The concepts of statics and dynamics in economics do not involve the application of an analogy drawn from mechanics, but represent a mode of thinking appropriate to the character of economic science, for which only the name employed by mechanics was borrowed. Cf. Alfred Amonn, Grundzüge der Volkswohlstandslehre (Jena, 1926), Part I, 275 ff.
• 75Cf. J.B. Clark, Essentials of Economic Theory (New York, 1907), pp. 130 ff.
• 76It is a serious misunderstanding to believe, as Flügge does (“Institutionalismus in der Nationalökonomie der Vereinigten Staaten,” Jahrbücher für Nationalökonomie und Statistik, New Series, LXXI, 339) that the construction of a static state would not be suited to lead to the understanding of economic changes.
• 77Cf. above pp. 77 f.
• 78Cf. Georg Simmel, Die Probleme der Geschichtsphilosophie (4th ed.; Munich and Leipzig, 1922), pp. 107 ff.
• 79Cf. Karl Bücher, Die Entstehung der Volkswirtschaft, Series I (10th ed.; Tübingen, 1917), p. 91. Bücher’s theory of historical stages is taken here as representative of an entire class of such theories, among which, for example, we may number that of Schmoller. The dispute over precedence connected with Bücher’s theory is without importance from our point of view.
• 80Ibid., p. 149.
• 81Ibid., p. 150.
• 82On the other hand, Erich Becher, Geisteswissenschaften und Naturwissenschaften (Munich and Leipzig, 1921), pp. 131, 171f. is inclined to see in these theories of historical stages “universal laws, or, if one wishes to speak more reservedly, principles of historical economic development.”
• 83Hegel, Vorlesungen über die Philosophie der Weltgeschichte, pp. 232 f.
• 84Cf. Mougeolle, Les problèmes de l’histoire, pp. 98 ff., 121 ff.
• 85Cf. Becher, Geisteswissenschaften und Naturwissenschaften, p. 175.