Frank Knight Essay
Frank Knight (1885-1972) was cynical of positive action as a means of combating social and economic ills. Harking back to the classical liberalism of Smith and Hume, he helped mold the economic views now linked with the economics faculty of the University of Chicago, which he joined in 1927. His distrust of reformers never altered and was never more clearly expressed than in his 1950 presidential address to the American Economic Association in which he remarked, while a man or group asks for power to do good, my impulse is to say, “Oh, yeah, who ever required power for any other reason?” And what have they done when they got it? So, I instinctively want to cancel the last three words, leaving simply “I want power,”—that is easy to believe.’ (American Economic Review, 41 March, 1951).
Frank Knight’s effort to explain the nature and role of economic man is elemental to his defense of neoclassicism. Knight accepted that the determinants of human behavior are multifaceted; all of these aspects cannot be integrated and reflected in the behavioral assumptions made by economists. Knight retains that it is essential to abstract from reality and focus on only those aspects that are pertinent to explaining economic behavior. Thus, the way of thinking of the economist is predicated on the supposition that when it comes to the material aspects of life, individuals make choices which will exploit their gains, both as consumers of goods and services and as producers. In making this concept, the economist is following precisely the same procedure as the usual scientist who also excludes the influence of those variables whose operation is either inappropriate or prejudicial to the conclusion the researcher is seeking to begin.
An ‘economic person’ does not and, certainly, cannot approximate the individual of the real world. But the concept is useful, in Knight’s view, for helping us understand the only economic dimension of human behavior. His basis is that economic activity is intended toward maximizing producer and consumer gains and that it is financial behavior the economist is trying to explain. The consumer is ruler and, in an indecisive world, it is the producer who correctly anticipates what consumers’ desire who will be rewarded with profit. To Knight profit is thus the return for bearing uncertainty. It comes as outstanding, after all contractual compulsions have been met, and only as there is no guarantee that sovereign consumers will in fact purchase what has been produced. Thus, Knight’s contribution to the theory of profit is directly related to his argument concerning consumer sovereignty.
Frank Knight Influence on Chicago University
Frank Knight was hesitant of positive action as a means of combating social and economic ills. Harking back to the classical liberalism of Smith and Hume, he assisted mold the economic views now linked with the economics faculty of the University of Chicago, which he joined in 1927. His mistrust of reformers never altered and was never more coherently expressed than in his 1950 presidential address to the American Economic Association in which he remarked, ‘when a man or group asks for power to do good, my impulse is to say, “Oh, yeah, who ever wanted power for any other reason?” And what have they done when they got it? So, I impulsively want to cancel the last three words, leaving simply “I want power,”—that is easy to believe.’ (‘The role of principles in economics and policies,’ American Economic Review, 41 (March, 1951), p. 29).
The microeconomic propositions of contemporary Chicagoans build chiefly on the work of Carl Menger, as interpreted and transmitted by Frank Knight. Their analyses ensue from the premise that choice is governed by individual discernments of the utility associated with alternative courses of action. Following Menger, Knight maintained that the appropriate cost of any economic decision is the utility of the alternatives sacrificed. No resource has any value other than that imputed to it by the consumer, whose objective it is to maximize the returns yielded by a given supply of resources. Knight credits Menger for establishing the authority of this principle (James Dingwell and Bert Hoselitz, 1950, p. 15).
Modern writers in the Chicago tradition have built on the Menger-Knight viewpoint of the relationship between utility and cost (i.e. the cost of any choice is the utility lost in choosing one alternative rather than another) to look at the behavior of the household in managing its time and income resources. An imposing range of topics traditionally examined by sociologists or psychologists has come within their possibility of analysis. Using the framework provided by economic theory, the new microeconomics has examined such topics as the provision of time to education and training as investment in human capital, the rearing of children, criminal behavior as an different to market behavior, and choice among sex partners (Richard B. McKenzie and Gordon Tullock, 1975). These inquiries represent modern efforts to explore Knight’s classic observation: To live, in the human plane, is to choose.’ (Frank Knight, 1925)
It must be noted that the modern microeconomic understanding of time as a scarce allocatable input reflects a formation different from the Marshallian one that relates time to continuing processes.
Knight and methodological individualism
Was Knight a methodological individualist? The answer evidently depends upon what is meant by practical individualism. It is a notion that should be defined carefully. Gonce (1972:552) identified its Austrian school meaning and stated that: “it signifies that the subjectivity of the self-interested, rational and free individual is to be used to explain all human conduct and social and economic phenomena.” Gonce’s definition of methodological individualism shall be used here. Note that this version of practical individualism requires that the individual is used as the base of the explanation of all socioeconomic phenomena. It is dependable with Lachmann’s (1969:94) slightly looser definition “that we shall not be content with any type of explanation of social phenomena which does not lead us eventually to a human plan.” It is also dependable with one of the most precise definitions, by Jon Elster (1982:453), who saw methodological individualism as: “the principle that all social phenomena (their structure and their change) are in principle reasonable only in terms of individuals—their properties, goals, and beliefs.” With a considerably similar definition of methodological individualism, Gonce (1972:552 n.) made the debatable claim that Knight “continuously…employed this procedure.” To assert support for his view he referred to some parts of Knight’s work. But, in contrast, turning to these passages, we find Knight attempting no such method for explaining all socioeconomic phenomenons.
Others have suggested that Knight was a practical individualist, but with an absolutely different definition of that term. For example, McKinney (1977:1441) referred to one passage (Knight, 1951:23-24) alone to support the dubious claim that “Knight is a methodological individualist.” In this passage Knight argued that “the organization as a whole has no value in itself or purpose of its own.” Though, this statement is about the alleged location of values and wills in individuals rather than organizations; it is not concerning the explanation of socioeconomic phenomena in terms of individuals alone. The passage does not give substantiation that Knight was a methodological individualist in the sense defined by Gonce or Elster. The claims of Gonce and McKinney to have found proof of methodological individualism in Knight’s works do not stand up to decisive scrutiny.
To play the Devil’s Advocate, we might address another passage by Knight which, perhaps at first sight, may seem to sustain the case that he was a methodological individualist. Consider the following:
“Institutional Economics”…is essentially a continuation or revival of the historical standpoint. But if human and social phenomena can be completely explained in terms of their own history, the result is the same as that of a complete mechanical explanation; there is no such thing as purposive action or as practical relevance…. The view here adopted… treats social institutions as a product of social choice based on social knowledge of patterns between which choice is made, and has meaning only in so far as such choice may be real.
(Knight 1935b:285 n.)
Here, Knight made two (valid) points. First, he warned against some actual tendencies in historical and institutional economics who wished to explicate all socioeconomic phenomenons in terms of historical trends, omitting human judgment and creativity. Second, he stated that institutions should be seen as the result of real social choices. There is not anything here to suggest that institutions have to be explicated entirely in individual terms. Though Knight saw human will at the root of all socioeconomic activity, he did not consider that human choices alone were an enough explanation of all social and economic phenomena. Certainly, Knight (1924:263) devoted much consideration to the fact that “wants are reliant variables…largely caused and shaped by economic activity.”
In fact, his proclivity was against this reductionism thrust of methodological individualism. Knight (1924:247) wrote: “the way to improve our technique is not to attempt to analyze things into their elements, reduce them to measure and determinate functional relations, but to educate and train one’s intuitive powers.” Furthermore, Knight discarded the individualist view of knowledge underlined by the Austrian school. He was much closer to Veblen than to (say) Friedrich Hayek when he wrote that: “all knowledge is itself ‘social’; it is based on intercommunication between individuals, each of whom is both subject and object, both to himself and to all others in the thinking community in which knowledge has its being” (Knight 1956:136). Knight saw the forces of institutionalization and of indefinite individuality as both at play. Like the old institutionalizes he stressed the role of habit, simulation, and custom. For Knight, the forces that help to mould human society
Belong to an intermediate category, between instinct and intelligence. They are a matter of custom, tradition, or institutions. Such laws are transmitted in society, and acquired by the individual, through relatively effortless and even unconscious imitation, and conformity with them by any mature individual at any time is a matter of “habit.”
His emphasis on the role of customs and institutions was repeated elsewhere:
Human society must always be largely of the original institutional character; custom and habit must rule most of what people feel, think, and do. Institutions, I repeat, are more or less explained historically rather than scientifically and are little subject to control. The ideal type is language, about which we can do so little that we hardly think of trying.
But nowhere here does Knight fall into the determinist trap of suggestive of that human behavior is completely determined by custom or institutions. Indeed, all through his writings, he attempted to merge the non-contradictory propositions of individual “free will” with the added notion that individual behavior was molded and constrained by institutions. Such settlement is problematic, but it does not falter on simply logical grounds. No “paradox” is involved. Certainly, the settlement of these two different ideas is part of the continuing agenda of social theory. Knight nowhere evades the dilemma of structure and agency by yielding to the untenable extremes of either cultural determinism or methodological individualism. The fact that his own solution was incomplete, and even defective, is beyond the point. The point here is to set up that attempts to describe Knight as a methodological individualist do not work.
Knight, understood that orthodox theory was valid in detaining only one facet of economic life; a universal and possibly the most significant aspect, but nonetheless only one. Therefore they had no dilemma in accepting the idea that it was essential to supplement orthodox theory with “applied” or “historical” studies:
Such a mass of interrelated data seems to call for a combination of three methods of treatment which must logically be sharply differentiated. The first is economic theory in the recognized sense, a study, largely deductive in character, of the more general aspects of economic cause and effect, those tendencies of a price system which are independent of the specific wants, technology and resources. The second division, or applied economics, should attempt a statistical and inductive study of the actual data at the particular place and time, and of the manner in which general laws are modified by special and accidental circumstances of all sorts. … The third division of economics is the philosophy of history in the economic field, or what some of its votaries have chosen to call “historical” and others “institutional” economics, studying “the cumulative changes of institutions” (Knight, 1924, 264).
By “applied economics” Knight referred to the quantitative research of Mitchell at Columbia and the NBER, and “institutional economics” was his description of Commons’s research program at Wisconsin. The role of applied economics was to conduct “a statistical and inductive study of the actual data at the particular place and time, and of the manner in which general laws are modified by special and accidental circumstances of all sorts.” Knight emphasized that “this branch of the science is focus to very narrow limitations” because the data being gathered “lack the stability, classifiability and measurability requisite to scientific treatment” (Knight 1924, 264).
More essential, Knight believed that applied economics could do well only when it relied upon theory as a benchmark. He compared applied economics to applied physics and said that in both fields “the application of principles is impossible without principles to apply. It is no argument against the practical value of pure theory that taken alone it does not yield definite rules for guidance” (1924, 259).
Similarly, Knight limited the function of chronological and institutional studies and doubted their capability to yield “scientific” results in the strict meaning of the concept. Historical and institutional economics, he argued, was “a field for the exercise of informed judgment rather than for reasoning according to the canons of science” (ibid., 264).
Raymond Bye (1924, 273–75), who argued that the theory of value as well as distribution should persist to stand at the center of the field, raised another validation for empirical research: the requirement in descriptive information in fields which lack theory, such as consumption and indecision. He added, however, that it was “difficult to conceive of an ‘institutional economics’ which can ever take the place of current economic science.” “A mere description of the organization and technique of institutions,” he explained, would “furnish much information of value to the business man and statesman, and help to build up an applied economics as distinguished from the broader general or ‘pure’ science,” but of course, it was the “pure science” which was more significant even for practical matters.
The institutionalisms did not recognize the separation between “pure theory” and “applied economics”. Thus, Spengler maintained at a roundtable meeting throughout the AEA annual conference that a virtually good theory “must give due weight to the role of institutions and must allow for more flexibility in economic behavior than is assumed in the classical analysis” (in Homan 1931, 136).
A similar contention is made by Sumner Slichter. In contrast to Knight, Slichter insisted that “the problem of industrial organization and control is … a fit subject for theoretical speculation, and there are advantages in having it investigated by theorists as well as by specialists in other fields.” In order to reach useful solutions, practical problems needed “to be studied not merely as a problem of marketing, banking, railroads, agriculture, or municipal utilities, but as a general problem of economics.” Whereas Knight doubted the capability to reach general conclusions about the development of economic institutions, Slichter believed that “the comparative method should be used and an effort made to arrive at general laws and principles, to build up … a general theory of economic organization and control” (Slichter 1924, 353).
Transactors as Maximizers
…There is a science of economics, a true, and even exact science, which reaches laws as universal as those of mathematics and mechanics. The greatest need for the development of economics as a growing body of thought and practice is an adequate appreciation of the meaning, and the limitations, of this body of accurate premises and rigorously established conclusions. ( 1935b, 135)
Knight’s understanding of economics’ assert to be a science, and of the role the maximizing supposition plays in establishing that claim, comprise one of the distinctive features of his framework. For Knight, economic theory studies the abstract, idealized world of perfect competition, in which balanced individuals with perfect knowledge professionally satisfy their preferences through exchanges in markets that work completely. The theorist’s task as a scientist is to cut off the assumptions about human behavior and the market which make certain that such exchanges are reduced to an entirely mechanical process. Only when those assumptions are well specific can the theorist determine the critical distance between the perfect markets of economic theory and the imperfect markets of actual human practice.
According to Knight, the fundamental assumption for a scientific economics was the maximizing assumption: the idea that the individuals engaging in exchange relations are balanced, utility-maximizing agents. Placed second in the majority of his lists of the central assumptions of economic theory (after a general statement concerning the nature of market society), the maximizing postulation was central to his formulation of neoclassical economics, for three reasons. First, it recognized the nature of human activity in the market as the premeditated, knowledgeable, and calculated adjustment of known means to given ends. Secondly, it illustrious the morally neutral notion of economic prudence from the morally charged nature of social discussion concerning what is “good” in human action. And finally, as he formulated it, the maximizing postulation explicitly connected maximization with the postulation of perfect knowledge — a connection which was fertile theoretical ground for Knight himself (1921b) and many others. All three of these reasons for the centrality of the maximizing postulation to the neoclassical tradition can be seen in his statement of the assumption in Risk, Uncertainty, and Profit:
We assume that the members of society act with complete “rationality.” By this we do not mean that they are to be “as angels, knowing good from evil”; we assume ordinary human motives…; but they are supposed to “know what they want” and to seek it “intelligently.” Their behavior, that is, is all “conduct” as we have previously defined the term [i.e., efficient adaptation of known means to given ends]; all their acts take place in response to real, conscious, and stable and consistent motives, dispositions, or desires; nothing is capricious or experimental, everything deliberate. They are supposed to know absolutely the consequences of their acts when they are performed, and to perform them in the light of the consequences. ([ 1921b ] 1971, 76-77)
The theoretical conclusions Knight draws from the maximizing conjecture need not detain us here, for they are recognizable to anyone schooled in neoclassical economics. There are, though, a couple of exceptional features of his formulation of the maximizing postulation that deserve special mention. The first, is the special relation he draws between the maximizing postulation and the assumption of perfect knowledge. With the exemption of his discussion of risk in Risk, Uncertainty, and Profit, he leans to conflate the two assumptions: people maximize want-satisfaction through actions based upon entire knowledge of the consequences. “In acts looking to the future,” he tells us, “intelligent action requires perfect foreknowledge” ([1935a] 1935b, 283 n). The question we might ask is, why did he think it necessary to conflate the two assumptions?
The answer will appear from a look at the other unique feature of Knight’s formulation of the maximizing supposition, namely, his belief that, properly formulated, the supposition enables economic exchange to be reduced to an impersonal, mechanical process.
The social organization contracted with in economic theory is best pictured as a number of Crusoes interacting throughout the markets exclusively. To the economic individual, exchange is a feature in production, a mode of using private resources to understand private ends. The “second party” has an indistinct existence, as a detail in the individual’s use of his own resources to assure his own wants. It is the market, the exchange opportunity, which is functionally real, not the other human beings; these are not even means to action. The relation is neither one of collaboration nor one of mutual development, but is entirely non-moral, non-human. ([1935a] 1935b, 282)
Conceptualized in this fashion, economic exchange is just a “mechanical sequence” of want-satisfaction ([1935a ] 1935b, 280). Blessed with perfect foresight, the individuals in the completely competitive economy know what symmetry prices are and can carry on automatically to accomplish their wants and desires within the restrictions of their resources. There is no competition, haggling, cooperation, rivalry, or any other type of human interaction between the two parties to an exchange in the neoclassical model. The market, which gives the context for want-satisfaction to be fulfilled, is what is genuine to each person; the individuals with whom exchange occurs are mere details — they might as well be “vending machines” ( 1960, 73). “The Economic Man,” Knight says, “neither competes nor higgles…he treats other human beings as if they were slot machines” ([ 1939 ] 1947, 80; see also 1923, 1935a, 1930).
Diminution of all exchange relations to mechanical processes brings us full circle and exemplifies the close relation for Knight between the maximizing conjecture and the scientific status of economics. If people can exploit perfectly, exchange is reduced to a perfunctory process and economics can assert to be a scientific study of social organization. “The statement that economics describes the way the economic order works,” Knight tells us, “refers to its working as a mechanism; that is the meaning of being scientific” ( 1960, 72). From this perspective, then, “the first question in regard to scientific economics is this question of how far life is rational, how far its problems reduce to the form of using given means to achieve given ends” ([ 1924 ] 1935b, 105).
Knight’s Criticism of the Maximizing Assumption
Given his participation in the neoclassical tradition and the implication he attached to the maximizing assumption, unsuspecting readers of Knight’s writing may be surprised when they discover passages in which he rejects the maximizing assumption as a convincing guide to either personal action or the understanding of human action
For instance, one finds a criticism of the maximizing postulation in Risk, Uncertainty, and Profit, where Knight suggests that uncertainty is the essential context of many (perhaps most) economic activities, and that when uncertainty is present, the decision required for good behavior departs from the rational computation required for maximization. For him, the necessary differences between certainty, risk, and uncertainty are linked to the limits of human knowledge and, hence, the prospect for maximization. In a certain universe all outcomes are known perfectly, and maximization is definite. In a risky universe the prospect of all outcomes can be known completely, and maximization is also guaranteed. Hence maximization works evenly well in a fully determined, or a fully stochastic, universe. However, maximization is provided sterile by uncertainty, which exists whenever we lack any idea basis — either deterministic or stochastic — upon which to base our knowledge of the consequences of probable actions: “It is a world of change in which we live, and a world of uncertainty. We live only by knowing something about the future; while the problems of life, or of conduct at least, arise from the fact that we know so little” ([ 1921b ] 1971, 199). When neither the outcomes of our actions nor their probabilities are known, wise choices must be characterized by critical judgment (a nonmechanistic notion for Knight), rather than maximization (197 – 232).
The condemnation of the maximizing assumption first advanced in Risk, Uncertainty, and Profit is prolonged upon in Knight’s methodological and philosophical writings, where he argues that exchange behavior is basically experimental — “an exploration in the field of values” ([ 1924 ] 1935b, 105) in which individuals seek, through collaboration with others, to discover what they actually want. In this literature Knight’s criticism focuses on the basic wavering of preferences rather than the predicament of imperfect knowledge. Human wants are unstable, he suggests, for three reasons. First, our actions are often impetuous “a relatively unthinking and undetermined response to stimulus and suggestion” ( 1935b, 50). Also, because our wants are learned and shaped by the economic system itself (through advertising, for instance, but also through education and social relations), the ends we practice are not as independent as the maximizing assumption presumes ( 1923). Third, there is an intrinsic dynamism to human preferences that is linked to the basic drive to improve: “The chief thing which the common-sense individual actually wants is not satisfactions for the wants he has, but more, and better wants” ([1922 ] 1935b, 22). Unfortunately, ruse and deceit in the marketplace often weaken our desire to improve, so that the wants we actually pursue do not lead us to become good people.
Knight’s understanding of the instability of preferences and its moral implications are summarized in “The Ethics of Competition,” his moral critique of the market system:
[Even] the freest individual, the unencumbered male in the prime of life, is in no real sense an ultimate unit or social datum. He is in large measure a product of the economic system, which is a fundamental part of the cultural environment that has formed his desires and needs, given him whatever marketable productive capacities he has, and which largely controls his opportunities. Social organization through free contract implies that the contracting units know what they want and are guided by their desires, that is, that they are “perfectly rational,” which would be equivalent to saying that they are accurate mechanisms of desire-satisfaction. In fact, human activity is largely impulsive, a relatively unthinking and undetermined response to stimulus and suggestion. Moreover, there is truth in the allegation that unregulated competition places a premium on deceit and corruption….It is plainly contrary to fact to treat the individual as a datum, and it must be conceded that the lines along which a competitive economic order tends to form character are often far from being ethically ideal. ([ 1923 ] 1935b, 49-50)
One way of shortening Knight’s criticisms of the maximizing supposition is to say that they focus our attention on the troubles linked with the impersonal nature of a market operating under the maximizing assumption. This is not surprising, where we saw that the aloof and mechanistic nature of exchange was inner to his understanding of the role of the supposition in neoclassical theory. When people know what they desire and how to get it, their exchanges become impersonal and the procedure by which they exchange becomes perfunctory. The discipline which studies the mechanical progression of exchange, therefore, can seek to the scientific status given to classical mechanics. Knight’s criticisms remind economists, though, that attaining the status of a science has a cost, as reference to the mechanical nature of the price system obscures as much as it divulges. And one of the things it obscures is that mechanical maximizes are neither human, nor do they interrelate with each other as humans: “The view of human behavior as a mechanical sequence… is impossible to human beings….This is one of the main differences between the economic man and the real human being” ([ 1935a ] 1935b, 280, 282).
Because human action, for Knight, is eventually a discussion about the kind of people we desire to become, the impersonal nature of market exchange destabilizes our ability to become good people. Thus the answer to the “first” question of economic science — how far is life rational? -is “not very far; the scientific view of life is an inadequate and partial view,” and this “sets a first and most sweeping limitation to the idea of economics as a science” ( 1935b, 105).
Knight on neoclassicism and institutionalism as complements
What was Knight’s attitude to the prospect of a combination between neoclassical economics and institutionalism? Generally, he assented to this idea. Knight did not believe that neoclassical theory must be entirely replaced. But it had to be supplemented by studies of history and institutions. The role of “institutional economics” was therefore seen as a complement to some features of neoclassical economics, including ethically directed utility theory and Marshallian-type price theory.
Does this limited acceptance of neoclassical theory prohibit Knight as an institutionalist? If so, then we must also indict others, including leading institutionalists such as J.R. Commons, W.C. Mitchell, J.M. Clark, P.H. Douglas, and A.F. Burns. They all saw institutionalism as compatible with aspects of Marshallian price theory. This passage from Commons is particularly revealing:
Sometimes anything additional to or critical of the classical or hedonic economics is deemed to be institutional…. [However] institutional economics…cannot separate itself from the marvellous discoveries and insight of the classical and psychological economists…. Institutional economics is not divorced from the classical and psychological schools of economists.
Although Commons did not use the word “neoclassical” here, he evidently was referring in part to this tradition while he used terms such as “hedonic economics” and “classical and psychological schools of economists.”
Accordingly, when trying to decide whether person X was or was not an institutionalism, it is hard to disqualify X simply on the grounds that X accepted elements of neoclassical economics in their theory. If such an excessively restrictive criterion is raised, then we are left with very few institutionalists at all. That is one reason why this decisive factor must be rejected. What is significant is whether X accepted neoclassical thinking as adequate for economic theory. Those economists who consider that it is adequate can be described as neoclassical economists. Those who think that it is inadequate may, subject to further conditions, be described as institutionalists, even if they recognize parts of neoclassical economics as valid. Hence to point out that X accepted elements of neoclassical economics in their theory does not inevitably disqualify X from being an institutionalist.
Consequently, there is an asymmetry between the two definitions. An institutionalism does not finish being so if she accepts key elements of neoclassical theory. However, a neoclassical economist ceases to be such, once she accepts the center proposition of institutionalism: that economic theory should take account of changing preferences and purposes and not take them as given.
That change of point of view is sufficient to take away the “neoclassical” nametag and install the “institutionalise one instead. To understand Knight’s own position here it is significant to appreciate his captivation in the work of Weber and the other German historical school economists. Like the German historical school, Knight tried to grapple with the problem of chronological specificity in economic analysis. In his Risk, Uncertainty and Profit, Knight (1921a:9) tried to explicate some economic principles that are specific to “free enterprise” or “the competitive system.” The study of this system, according to Knight, must proceed “as a first approximation” from “a perfectly competitive system, in which the countless degrees and kinds of divergences are eliminated by abstraction.” He went on to argue a selection of mainstream economists who have followed such an approach.
In subsequent works. Knight tried to answer the more basic question, which had been addressed by historical school thinkers such as Weber and Werner Sombart. True, the system of “free enterprise” might have common features from nation to nation, but a more general methodology is necessary to define the system of “free enterprise” and to demarcate it from other systems. This more general “economic” discourse may itself contain general concepts, principles or laws. Therefore, which features, assumptions or laws are general to all economic systems, and which are specific to, say, modern capitalism?
One of Knight’s most absolute attempts to answer these questions is found in his contribution to the 1924 “institutionalise Tugwell volume. For Knight, the discovery of universal economic principles or laws assured a place for the abstractions of neoclassical economic theory. But these do not take us far enough:
The problem of life is to utilize resources “economically,” to make them go as far as possible in the production of desired results. The general theory of economics is therefore simply the rationale of life.—In so far as it has any rationale! The first question in regard to scientific economics is this question of how far life is rational, how far its problems reduce to the form of using given means to achieve given ends. Now this, we shall contend, is not very far.
Yet, Knight held that universal principles must not be disregarded. He alleged that there were universals such as “the general laws of choice,” “general laws of production and consumption.” In other words, for Knight, all individuals all through history, in their economic activities, make choices, and the same “laws” allegedly govern these choices. Although Knight does not convolute much on the nature of these allegedly universal laws, he gave an example: “in the large the conditions of supply and demand determine the prices of goods” (Knight 1924:259). He wrote further:
Institutions may determine the alternatives of choice and fix the limits of freedom of choice, but the general laws of choice among competing motives or goods are not institutional…there are general laws of production and consumption which hold good whatever specific things are thought of as wealth and whatever productive factors and processes in use…. The laws of economics are never themselves institutional, though they may relate to institutional situations. Some, as we have observed, are as universal as rational behavior, the presence of alternatives of choice between quantitatively variable ends, or between different means of arriving at ends…. A large part of the extant body of economic theory would be as valid in a socialistic society as it is in one organized through exchange between individuals.
This argument is questionable, partially because it regards choice as an activity which itself is not framed and formed by institutions. Several may dispute Knight’s assertion here that the neoclassical “laws of economics are never themselves institutional.” It may be the case that “rational behavior” is worldwide in an empty sense, simply because no possible form of behavior can be noticeably inconsistent with its axioms (Boland 1981). Notably, in other passages, Knight (1933: xii-xiii) himself accepted that calculative rationality was an in history specific phenomenon, thus seeming to refute its universality. We may go further: if “supply and demand” are to “establish the price of goods”, they may require a prior non-universal institutional framework of exchange and a degree of price flexibility so that these forces can move their objects. Knight’s “general laws” may not be so general after all.
But the arguable character of this argument is beyond the point. The main point here is that Knight was trying to distinguish universal from non-universal laws. His claim was that the universal “laws of economics” provided the area for neoclassical theory, and they had to be supplemented by the traditionally specific study of institutions. Knight clearly pointed out that not all laws are universal, and that diverse laws had different domains of applicability:
Other laws relate to behavior in exchange relations, and of course have no practical significance where such relations are not established. Still others cover behavior in situations created by even more special institutional arrangements, as for example the differences in business conduct created by the custom of selling goods subject to cash discount or by the existence of a branch banking system as contrasted with independent banks. An intelligent conception of the meaning of science requires a clear grasp of the meaning of classification and sub classification, of laws of all degrees of generality. Each law is universal in the field to which it applies, though it may not give a complete description of the cases which it fits. Quite commonly a law has the form “insofar as the situation is of such a character, such things will happen.”
Knight’s notion here that an “intelligent commencement of the meaning of science requires a clear grasp of the meaning of categorization and sub classification, of laws of all degrees of generality” shows the particular inspiration of Weber and the German historicists. Knight gave this an added, institutionalist twist. Institutions were seen to “determine the alternatives of choice and fix the limits of free will of choice.” This, for Knight, was the place for institutional economics. Knight thus insisted that the “general theory of economics” is valid but it does not get us “very far.” His writings are estimably consistent on this point.
“The principles of the established economics are partial statements, but sound as far as they go, and they go about as far as general principles can be carried…. General theory is a first step, but never a very long step toward the solution of practical problems” (Knight 1921b: 145).
In some passages Knight did make it understandable that the approach of neoclassical economics is itself dependent on institutionalism insights for any sort of rational account of “rational economic man.” However, he treats institutions mainly as constraints, rather than them being constitutive of choice and behavior. Yet, for Knight, without institutions, economic man would exist in a vacuum devoid of a history or a future. Hence he saw the association between neoclassical and institutional economics as one of “complementarily” (Knight 1952:46).
As Knight (1924:262) himself put it: “deductive theory and ‘institutional’ economics” are both relevant: “at one extreme we might have a discussion limited to the abstract theory of markets…at the other extreme we should have the philosophy of history…and that is what institutional economics practically comes to. It should go without saying that all are useful and necessary.” Knight (1924:265-66) went on to explain the key role of institutional economics:
The study of such long-time changes would seem to be the most conspicuous task of institutional economics… No one would belittle the importance of studying these historic movements in the general structure of social standards and relations… But neither, we think, can anyone contend that such a study should displace the other branches of economics which either are fairly independent of institutions or take them as they are at a given time and place and use them in explaining the immediate facts of economic life.
For Knight (1924:229), it was essential for the economist to examine the role of institutions and to also get on “an exploration in the field of values.” Institutionalists would agree. In a letter to his friend Talcott Parsons, dated 1 May, 1936, Knight wrote: “I came to Chicago expecting this ‘institutionalism’ to be my main field of work” (Knight, 1936). He complained that other hypothetical controversies and teaching had got in the way of this research project. As at Chicago, Knight taught that institutional economics and Marshallian neoclassical economics had complementary roles. In a letter dated 16 February, 1937, to his friend Ayres, Knight reported that he was giving a course on “Economics from an Institutional Perspective” at Chicago. In fact, Knight had started giving this course in the summer quarter of 1932. Knight’s Reading List for Economics 305, Winter 1937, says: “The task of institutionalism [is] that of accounting historically for the factors treated as data in rationalistic, price-theory economics.” He then lists the topics “individualism and utilitarianism, wants, technology, resources, organization, economic institutions as embodied in law” (Samuels 1977:503). In principle, there was nothing in Knight’s qualified acceptance of some neoclassical tenets to debar him from institutionalism. He himself made the place of institutional economics explicit and extensive.
The evaluation of Knight by James Buchanan (1968:426) is not far from the mark. According to Buchanan, Knight was “that rare theorist who is also an institutionalist, an institutionalist who is not a data collector.” Was Knight an institutionalist? He should have the final word on this question. Knight (1957:43) declared: “I am in fact as ‘institutionalist’ as anyone, in a positive sense.” In a letter to Ayres dated 13 July, 1969, Knight wrote “I’ve always considered myself an institutionalist…as far as possible” (Samuels 1977:519). His institutionalism was nonconformist, but it was an institutionalism nevertheless.
Frank Knight, one of the most famous neoclassicists during that period, concurred, for instance, that the coupled treatment of goals and means by economists was inevitable and, more than this, desired:
[It] is in fact necessary and proper for the question of objectives to occupy a large if not the main part of social discussion. For social science cannot, like the natural sciences, be restricted to the problem of means for achieving objectives taken for granted. Not the winning of power, but its use is, and must be, the leading question (1928, 247).
Knight thought that the effort to separate values and facts stemmed from a feeling of “pronounced irritation” on the part of a large amount of social scientists due to the uncertainty relating to the appropriate method of social science. This feeling caused a substantial number of younger men to “seek break out in the methods of the natural sciences,” and the positive-normative dichotomy was part of this effort. Knight considered this attempt hopeless. He argued instead “that the situation from which escape is sought is not as bad as it seems” (ibid.). Social science was diverse from natural science and one must not aspire to the same level of certainty and instant applicability as in the natural sciences. Economics must consist of a discourse on human problems, and that is why the separation of values from science was entirely misguided.
The common thread linking these varied inquiries is that they have build on the work of Frank Knight as an economic theorist, as a methodologist, and as a social philosopher. These ‘three hats’ of Frank Knight are barely separable from each other, and they are equally blended into the work of most of his followers. From the point of view of economic principle, the starting point of Knight’s analysis, which is obviously reflected in the present Chicago tradition, is his commitment to the principle that sovereign consumers are competent of maximizing behavior. Given freedom, each individual uses the available means to attain his or her own ends and each contract reflects a choice among alternatives. Expressed in the language of the economist, the choices of the autonomous individual are thought of as administrated by the universal principle of opportunity cost. Thus, the aim of the economist is to arrive at a body of scientific (i.e. value-free) truths predicated on individual freedom to choose amongst alternatives.
I personally believe Most of Knight’s teaching as a theorist and as a methodologist reflect his assurance as a philosopher to the dictum that ‘to live on the human plane is to choose.’ The essence of freedom is possibility, as different from coercion, which implies denial of possibility. For Knight, economic freedom is the essential freedom, because he saw it as fundamental all other forms of freedom—religious, political, and intellectual (Frank Knight (April, 1966), pp. 163-77). The perfect market, which is the personification of complete freedom, is identified as ideal in the sense that human capability for maximizing behavior is most totally realized under these conditions. Efficient resource allocation thus became inextricably interwoven with the perfect market in Knight’s thinking and teaching.
The thirty years that Knight spent at Chicago, from 1927 to 1957, were dedicated to articulating the utility principle and exalting economic laissez-faire, although he was not blind to the necessity for general legislation to prevent intolerable divergences from free market conditions.’
Buchanan, James M. (1968) “Knight, Frank H.,” International Encyclopaedia of the Social Sciences, Vol. 8:424-28.
Commons, John R. (1931) “Institutional Economics,” American Economic Review 21, 4:648-57. Reprinted in Warren J. Samuels (ed.) (1988) Institutional Economics, Vol. 1, Aldershot: Edward Elgar.
Elster, Jon (1982) “Marxism, Functionalism and Game Theory,” Theory and Society 11, 4:453-82. Reprinted in John E. Roemer (ed.) (1986) Analytical Marxism, Cambridge: Cambridge University Press.
Frank Knight, ‘Abstract economics as absolute ethics,’ Ethics, 76 (April, 1966), pp. 163-77
Frank Knight, ‘Economic psychology and the value problem’ (1925), reprinted in The Ethics of Competition and Other Essays (New York: Augustus Kelley, 1951), p. 88.
Gonce, R.A. (1972) “Frank H. Knight on Social Control and the Scope and Method of Economics,” Southern Economic Journal 38, 4:547-58.
Homan, Paul T. 1931. “Economic Theory—Institutionalism: What It Is and What It Hopes to Become.” American Economic Review 21 (March): 134–41.
James Dingwell and Bert Hoselitz (New York: The Free Press, 1950), p. 15.
Knight Frank H. (1921b) “Discussion: Traditional Economic Theory,” American Economic Review 11, Supplement: 143-46
Knight Frank H. (1933) “Preface to the Re-Issue,” in Risk, Uncertainty and Profit, 2nd edn., London: London School of Economics, xi-xxxvi.
Knight Frank H. (1947) Freedom and Reform: Essays in Economic and Social Philosophy, New York: Harper and Brothers. Reprinted 1982 with a Foreword by J.M. Buchanan. Indianapolis: Liberty Press.
Knight Frank H. (1951) The Economic Organization, New York: Sentry Press. First published privately in 1933. Reprinted 1967, New York: Augustus Kelley.
Knight Frank H. (1952) “Institutionalism and Empiricism in Economics,” American Economic Review (Papers and Proceedings) 42 (May):45-55.
Knight Frank H. (1956) On the History and Method of Economics, Chicago, IL: University of Chicago Press.
Knight Frank H. (1957) “A New Look at Institutionalism: Discussion,” American Economic Review (Supplement) 47, 2:13-27.
Knight Frank H. 1916. A Theory of Business Profit. Ph.D. dissertation, Cornell University. ——. 1917.
Knight Frank H. 1921 a. Cost. of Production and Price over Long and Short Periods. Reprinted in Knight 1935b, 186-216.
Knight Frank H. 1921b. Risk, Uncertainty, and Profit, Reprinted Chicago: University of Chicago Press, 1971.
Knight Frank H. 1923. The Ethics of Competition. Reprinted in Knight 1935b, 41-75.
Knight Frank H. 1924. The Limitations of Scientific Method in Economics. Reprinted in Knight 1935b, 105-47.
Knight Frank H. 1960. Intelligence and Democratic Action, Cambridge: Harvard University Press. LeRoy Stephen F., and Larry D. Singell Jr. 1987. Knight on Risk and Uncertainty. Journal of Political Economy 95 (April): 394-406.
Knight Frank H.. 1935a. Economic Theory and Nationalism. In Knight 1935b, 277-359.
Knight Frank H.. 1935b. The Ethics of Competition and Other Essays. New York: Harper.
Lachmann, Ludwig M. (1969) “Methodological Individualism and the Market Economy,” in Erich W. Streissler (ed.) Roads to Freedom: Essays in Honour of Friedrich A.von Hayek, London: Routledge and Kegan Paul, 89-103.
McKinney, John (1977) “Frank Knight on Uncertainty and Rational Action,” Southern Economic Journal 43, 4:1438-52.
Richard B. McKenzie and Gordon Tullock, The New World of Economics (Homewood, Ill: Richard D. Irwin, 1975).
Samuels, Warren J. (1977) “The Knight-Ayres Correspondence: The Grounds of Knowledge and Social Action,” Journal of Economic Issues 11, 3:485-525.
Slichter, Sumner H. 1924. “The Organization and Control of Economic Activity.” In Rexford Guy Tugwell, ed., The Trend of Economics, 301–56. New York: Alfred A. Knopf.
‘The role of principles in economics and policies,’ American Economic Review, 41 (March, 1951), p. 29.
‘The role of principles in economics and policies,’ American Economic Review, 41 (March, 1951), p. 29.