Books / Digital Text
4. Ludwig von Mises
Perhaps the most noteworthy thing about Böhm-Bawerk's statement is that its methodological prescription was repudiated in almost every point by his intellectual heir, Ludwig von Mises. Following his teacher on many facets of theory, Mises nonetheless developed an entirely different epistemological defense for his views. A neo-Kantian, he denied the possibility of arriving at laws by induction and defended the possibility of a purely a priori system of economic theory which he labeled "praxeology." In doing so he meant to free economics from a reliance on "psychological" considerations and sought only logical sanction for economic laws.
The early Austrians formulated their theories to contain actual psychological content. They did not intend that the terms they used be interpreted in such a way as to be free of such content.46 The third generation of Austrians, however, divided on this issue.
Hans Mayer, Leo Schonfeld, Paul N. Rosenstein-Rodan, and some other "realists" continued the Wieserian tradition and attempted to retain psychology as the basis of economic theory. These economists stressed the deliberateness of economic decisions. Their investigations into certain complex topics, such as utility calculations with complementary goods and the difference between ex ante and ex post utility, led them to reformulate utility theory into more sophisticated, but still psychologically based statements.47
Ludwig von Mises and Richard Strigl, retaining the ontological nature of Austrian theory but placing it on new epistemological foundations, led the "formalist" branch in associating the subjective valuations of individuals only with their actual choices.48 To Menger's concern with needs and Böhm-Bawerk's and Wieser's with psychology, Mises objects that economics as a science is not concerned with the motives behind human actions but only the implications of action itself.49 Such terms as "utility" and "satisfaction" are used by economics in a purely formal way devoid of psychological or hedonic content. "Concrete value judgments and definite human actions, he declares, "are not open to further analysis."50 That is, economics is not concerned with second-guessing the rightness or wrongness of purposes or actions. Mises points out that this neutrality follows from the subjectivist approach of viewing action through the eyes of the actor.51
Praxeology, according to Mises, is not concerned with why individuals pursue the specific purposes they do, but only with what can be deduced from the axiom that they do act purposively. From this fundamental axiom of human action, with the aid of certain subsidiary assumptions, the praxeologist deduces the entire body of economic theory.52 Mises presents this vision of economics comprehensively in Human Action. Its first seven chapters, devoted to methodology, keynote and book and provide the groundwork for all that follows.53
Mises is concerned not only with methodology in economics, but in the entire range of human studies. Praxeology, concerned with purposeful action and its ramifications whatever they may be, encompasses more than economics, though economics is its most developed branch.54
The formalistic approach of Mises is well illustrated by his derivation of the law of marginal utility without recourse to psychology or physiology. An agent having n units of a homogeneous good will employ one unit in a way (called the marginal employment or least urgent want) that he would choose to forego had he only n-1 units. The utility (preference ranking) assigned the marginal employment is called marginal utility. As his supply of want-satisfying good increases, the actor by definition chooses to apportion the increments to successively less urgent wants. The law of decreasing marginal utility, Mises thus affirms, is independent of any psychological or physiological statements regarding sensuous enjoyment, such as Gossen's law of the saturation of wants, upon which Wieser had based decreasing marginal utility: "Under the conditions no other result is thinkable. Our statement is formal and aprioristic and does not depend on any experience." The law "is already implied in the category of action. It is nothing else than the reverse of the statement that what satisfies more is preferred to what gives smaller satisfaction."55
Mises not only claims that praxeology provides aprioristic truth, but also that it "conveys exact and precise knowledge of real things." As Wieser attempted to do, Mises must forge a bridge from his deductions to the real world. His bridge consists of the argument that "the subject matter of praxeology, human action, stems from the same source as human reasoning. Action and reason are congeneric and homogenous; they may even be called two different aspects of the same thing."56 The "logical structure of action" is "linked to the logic of our thought," because we act on the basis of rational thought.57
While this argument may explain why human action is comprehensible to us, it is insufficient to anchor the chains of praxeological deduction in the real world. Praxeology, like Euclidian geometry, would make explicit what was once only implicit in its axioms, but without some tangible anchor its axioms would remain just as arbitrary. To provide the necessary sanction for the fundamental axiom of human action, Mises returns (in somewhat different fashion) to the foundation claimed by Wieser, namely introspection: "The starting point of praxeological thinking is not arbitrarily chosen axioms, but a self-evident proposition, full, clearly and necessarily present in every human mind.... The starting point of praxeology is a self-evident truth, the cognition of action, that is, the cognition of the fact that there is such a thing as consciously aiming at ends."58
Mises further develops a theme of the older Austrians from their debates with the Historical School when he takes pains to distinguish the knowledge provided by praxeology from that provided by history. Praxeology and history form the "two main branches of the sciences of human action."59 History, which includes economic statistics and descriptive economics, "cannot teach us any general rule, principle, or law." Indeed, the interpretation of statistics and other complex historical evidence presupposes praxeological knowledge in isolating causal relationships and grouping related events.60 Thus history, which is to say experience or empirical research, can neither prove nor disprove praxeological laws.
What experience can do for economic theory in Mises' view is to examine the applicability of the subsidiary assumptions made by the theorist concerning such things as the institutional setting in which action takes place and the perceptions of real-world actors. As an example of a praxeological law whose validity has been wrongly questioned Mises considers Gresham's law, which states that a legally overvalued currency will continue to circulate in payments, while an undervalued legal tender will not ("bad money drives out good"). The phenomena described by Gresham's law could fail to appear if agents were ignorant of their ability to pay in money valued lower by the market, or ignorant of the discrepancy between market and legal exchange values, or desirous for some reason of paying their creditors more than legally necessary. But the failure of the phenomena in question to appear in such cases would not in any way compromise the strict logical validity of the law.61
The universal non-appearance of the phenomena described by a praxeological law, because of the universal absence of the contingent conditions it assumes, would of course render that law uninteresting (though not invalid). The praxeologist must therefore refer to historical or empirical or institutional facts, at least in the broad sense of the facts of everyday experience, if he wishes to avoid irrelevance in developing laws that depend on more than the axiom of action. Nonetheless, Mises argues, this fact "does not alter the purely aprioristic character of praxeology. It merely circumscribes the field that the individual praxeologists customarily choose for their work."61
It is worth noting that Mises speaks of two sorts of auxiliary assumptions used in the construction of praxeological laws. One is the class of assumptions regarding environmental or empirical circumstances (e.g., we assume the presence or absence of fractional-reserve banking in developing business-cycle theory) that we have just discussed, whose correspondence to reality is of great importance for historical research. A second class of subsidiary assumptions is not contingent or "falsifiable" in this sense, but consists rather of special analytical assumptions or "imaginary constructions," such as the assumption that market equilibrium prevails before and after a change in the data. The value of this sort of assumption does not at all depend on its realism: equilibrium constructs are indispensable for praxeology and hence for our understanding of real-world events even though equilibrium conditions may not (or could not) ever prevail in historical fact.63
Mises thus insisted on a strict logical separation of theory from history. This has occasionally been misinterpreted as a denigration of historical or empirical research.64 Far from having such an intent, Mises declares that "history is not a useless pastime but a study of the utmost practical importance."65 Its scope is "the study of all the data of experience concerning human action. "66 Any economist must engage in historical research before he can claim that certain praxeological laws apply to or explain concrete historical episodes.67 Empirical research in economics is not made less important by the fact that its task cannot include "testing" or "falsifying" economic theories in the same way that laboratory experiments test natural-scientific theories.
Elaboration of the differences between social science and natural science is a theme present in Austrian methodology from the beginning. For Menger, Mises, and Hayek the fundamental difference is one of subjectivism versus objectivism. The natural scientists, standing as it were outside of their objects of study, must analyze empirical phenomena by breaking them down into hypothetical (unempirical) constituents. But for the social scientists the situation is reversed; here the researchers stand within the objects of their study, namely social and economic structures. The ultimate elements of the phenomena to be analyzed, human activities in pursuit of chosen goals, are known, and must be built up by theory into models of structures which cannot as a whole be directly observed.68 Menger explains:
- The ultimate elements to which the exact theoretical interpretation of natural phenomena must be reduced are "atoms" and "forces." Neither is of empirical nature. We cannot imagine "atoms" at all, and natural forces only by a representation, and by these we really understand merely unknown causes of real motion. From this there arise ultimately quite extraordinary difficulties for the exact interpretation of natural phenomena. It is otherwise in the exact social sciences. Here the human individuals and their efforts, the final elements of our analysis, are of empirical nature, and thus the exact theoretical social sciences have a great advantage over the exact natural sciences.69
In arguing that the "final elements" (really, the starting-point) of economic investigations are individuals and their purposes, Menger advances the doctrine of "methodological individualism" common to Austrian theory. This opposes the doctrine of "methodological holism, which thinks it legitimate for theory to operate exclusively at the level of social groups or economic aggregates, devoid of any link to individual behavior. We shall return to this question shortly.
Menger's more basic argument in this passage, that the proper approach of social science to its subject matter is different from the approach of natural science, is strongly seconded by Mises.70 Following Mises we may call this position "methodological dualism," in contrast with the "methodological monism" preached by behaviorists and positivists who see no basic reason to approach human behavior and social phenomena differently from the way natural scientists approach molecular behavior and physical phenomena.
Mises' well-known strictures against the use of mathematics in economics deserve mention here, as they are related to his methodological dualism. On the one hand, praxeology is like mathematics (and logic) in being an axiomatic or deductive system. On the other hand, as we have already noted, praxeology cannot be pursued as though it were a branch of applied mathematics because its starting point (the fact of human goal-seeking), unlike the axioms of Newtonian physics or other mathematical systems, is not arbitrary. This difference makes the mathematical methods of physics inappropriate for economics. Here Mises restates and extends Menger's argument:
- In physics we are faced With changes occurring in various sense phenomena....... We know nothing about the ultimate forces activating these changes....... What we know from observation is the regular concatenation of various observable entities and attributes. It is this mutual interdependence of data that the physicist describes in differential equations.
In praxeology the first fact we know is that men are purposively intent on bringing about some changes. . . . [T]he economist knows what activates the market process. It is only thanks to this knowledge that he is in a position to distinguish market phenomena from other phenomena and to describe the market process.
Now, the mathematical economist does not contribute anything to the elucidation of the market process . . . .71
Mises did not deny that mathematical techniques could be used to describe equilibrium conditions.72 But he argued that description of equilibrium conditions was not the ultimate or even main task of economic theory, which aimed at an understanding of market processes. Mathematical economics cannot yield the sort of causal-genetic explanations that Mises sought:
- . . . its equations and formulas are limited to the description of states of equilibrium and nonacting. It cannot assert anything with regard to the formulation of such states and their transformation into other states as long as it remains in the realm of mathematical procedures. . . . The problems of process analysis, i.e., the only economic problems that matter, defy any mathematical approach.73
Mises' principal indictment of mathematical economics was thus that its typical use, in equilibrium theory, is largely beside the point and not worth all the attention devoted to it. But he added that in other contexts, such as the use of mechanical differential equations to portray the process by which markets reach equilibrium, mathematical modeling is apt to be superficial, misleading, and distortive.74
- 46. Alan R. Sweezy, "The Interpretation of Subjective Value Theory in the Writings of the Austrian Economists," Review of Economic Studies 1 (June 1934).
- 47. "The Interpretation of Subjective Value Theory in the Writings of the Austrian Economists."
- 48. Strigl, in Die ökonomischen Kategorien und die Organisation der Wirtschaft (1923) distinguishes the categories from the data of economic science. The categories are derived from the fundamental fact of choice in the face of scarcity, and their validity is as general. From these categories all the laws of pure economics, which reveal the forms or qualitative relationships present in economic events, can be deduced. Statistical or other historical facts enter only as the "particularizing content" or reference of the laws. Sweezy, op. cit. On the realist-formalist split see also Sweezy, "The Austrian School and the Interpretation of Subjective Value Theory."
- 49. Mises, Human Action , p. 488. For Mises' critique of Menger's and Böhm-Bawerk's insufficient subjectivism see Epistemological Problems of Economics , ch. 5.
- 50. Mises, Human Action , p. 18.
- 51. Mises, Epistemological Problems of Economics, pp. 93-94.
- 52. See Murray N. Rothbard, "In Defense of 'Extreme A Priorism,'" Southern Economic Journal 23 (Jan. 1957): 314-320.
- 53. See the review by Lachmann, "The Science of Human Action," in Capital, Expectations, and the Market Process, esp. p. 95.
- 54. For an outline of the branches of praxeology, see Murray N. Rothbard, "Praxeology: Reply to Mr. Schuller," American Economic Review 41 (Dec. 1951): 943-946.
- 55. Mises, Human Action , p. 124. For a critical discussion of this and other claims see Robert Nozick, "On Austrian Methodology," Synthese 36 (Dec. 1980): 353-392.
- 56. Mises, Human Action , p. 39.
- 57. Lachmann, "The Science of Human Action," p. 416.
- 58. Ludwig von Mises, The Ultimate Foundation of Economic Science: An Essay on Method,(Kansas City: Sheed, Andrews and McMeel, 1978), pp. 4-6.
- 59. Mises, Human Action , p. 30. On history also see Mises, Theory and History (Washington, D.C.: Ludwig von Mises Institute, 1985)
- 60. Mises, The Ultimate Foundation of Economic Science, p. 41.
- 61. a. b. Mises, Epistemological Problems of Economics, pp. 87-88. For another example, Mises acknowledged that the Austrian theory of the trade cycle depended on a particular empirical assumption concerning the expectations of entrepreneurs, namely that they misinterpret what are merely the results of an easy money policy. This assumption need not hold true in every historical case: "It may be that businessmen will in the future react to credit expansion in another manner than they did in the past." Mises, " 'Elastic Expectations' and the Austrian Theory of the Trade Cycle," Economica 10 (N.S.) (Aug. 1943): 251.
- 63. Contrast The Ultimate Foundation of Economic Science, p, 41 with p. 44. Admittedly Mises is less than perfectly explicit on this distinction.
- 64. Ben B. Seligman, Main Currents in Modern Economics (Chicago: Quadrangle Paperbacks, 1971), v. 2, p. 330, protests that Mises "blithely cast out historical investigation of economic phenomena and reduced them to mere forms of economic history, not especially useful for the art of discursive reasoning. . . . The truth of the matter would seem to be that historical approaches are valuable to economists as well as other social scientists."
- 65. Mises, Theory and History, p. 291.
- 66. Mises, Human Action , p. 47.
- 67. This task of historical verification is undertaken, for example, by Murray N. Rothbard, America's Great Depression (Kansas City: Sheed and Ward, 1975), chs. 4-5, to show that credit expansion really did take place prior to 1929.
- 68. See Hayek, "The Place of Menger's Grundsätze," p. 8.
- 69. Menger, Problems of Economics and Sociology, p. 142n.
- 70. Mises, The Ultimate Foundation of Economic Science, pp. 23, The Ultimate Foundation , pp. 36, The Ultimate Foundation , pp. 40, The Ultimate Foundation, pp. 115.
- 71. Mises, Human Action , p. 355.
- 72. Human Action , P. 354.
- 73. Human Action , p. 356, emphasis added. For the argument that this statement has not yet been falsified by a counter-example see Lawrence H. White, "Mises, Hayek, Hahn, and the Market Process," in Israel M. Kirzner, ed., Method, Process, and Austrian Economics: Essays in Honor of Ludwig von Mises (Lexington, MA: Lexington Books, 1982), p. 108.
- 74. Mises, Human Action , pp. 354, 356-357. Hayek, we might note, takes a rather less negative view of mathematical economics, in particular viewing Walrasian general equilibrium systems as useful explanations of economic patterns: Hayek, "The Theory of Complex Phenomena," in Studies in Philosophy, Politics, and Economics (New York: Simon and Schuster, 1969), p. 35.