Frank H. Knight: The Forgotten Austrian

0:00My title is, of course, somewhat tongue-in-cheek, because Frank Knight is not, of course, a member of the Austrian School. Knight is one of the most important American economists and most important 20th century economists in the world. Knight was the intellectual father or one of the intellectual fathers of the Chicago School and is known primarily among Austrians for a very sort of bitter attack against Hayek and Boehm-Bawerk in the 1930s and 1940s over their theory of capital. Knight is also widely considered one of the founders of the modern theory of perfect competition, which of course has been sort of a bet noir for Austrian economists and other heterodox types for many decades.

0:52So my claim is not that Frank Knight belongs, that Knight's name should go up here. Maybe we could replace Gary Robin's name with Knight's name with some footnotes to explain exactly what is meant. The point that I wish to make is that in many ways, Knight's thought is much closer to that of the Austrians, Mises in particular, than has been generally recognized. Knight's epistemology, while somewhat eclectic, does have some similarities to the praxeological method espoused by Mises. Knight's development of perfect competition theory was really intended more as a kind of analytical foil to explain certain features of real markets in the absence of perfectly competitive conditions, is sort of analogous to the way that Mises and Rothbard have used the concept of the evenly rotating economy and other so-called imaginary constructs.

1:56In particular, Knight's theory of the entrepreneur, Knight's theory of profit, and his explanation of the differences between profit and other forms of business income are central to Austrian notions of entrepreneurship and the firm. And I'll try to convince you that Mises in particular takes a view very similar to Knight's view on the theory of the entrepreneur and the place of profit in a competitive market economy. One of the reasons I was tinkering with my slides a little bit during the previous presentations because I realized that in this room you can't see the bottom of the slides, correct? So except for a few of you in the front, you can only see what's at the top.

2:42Now, I tried to bump some stuff up and I have a few slides where it kind of scrolls up in the middle so those of you who can't see, you'll be able to see a little bit better as we go and anything that you can't see, I can assure you, is brilliant and highly insightful. Interesting person with a diverse background and of course Knight wrote in a number of different fields, not only economics but was an important social thinker in other disciplines, he studied at the University of Tennessee and received his PhD at Cornell in 1915. His major professor was Alvin Johnson. He also worked very closely with Alan Young and with Herbert J. Davenport at Cornell.

3:33I'll mention Davenport again later, but Davenport was one of the leading members of the so-called American Austrian School, including John Bates Clark and Frank A. Fetter, who were adherents of Menger, Devotees of the Menger's Price Theory and the Austrian Method, more or less. In fact, they affected somewhat of a synthesis or they tried to affect a synthesis between Menger and some aspects of British classical and neoclassical economics. They took some of Marshall and some of other leading British economists, incorporated those with much of Menger's Price Theory and that of the other Austrians. Come back to that point later. There's a strong influence on Knight of Davenport's thinking. Knight taught for eight years at the University of Iowa and he left and went to the University of Chicago in 1927 where he remained until his retirement in 1952 and he stayed on as an emeritus professor at Chicago and Knight was one of the huge, almost godlike figures at Chicago.

4:46in Chicago, along with Jacob Viner, extremely important economic theorist, particularly in international issues, who was also at Chicago. Knight and Viner have often been compared to Wieser and Boehm-Bawerk, where Viner, like Boehm-Bawerk, was a very careful and systematic thinker and a clear writer. Knight, like Wieser, was more of an iconoclast and was less clear and not as easy to understand. but in many ways a more original and sort of creative thinker. At Chicago, of course, Knight was the dissertation chair and sort of intellectual guru for the important figures of the later Chicago School that we would all recognize, people like Friedman and Stigler, of course, James Buchanan and so on.

5:39So in many ways, the sort of late 20th century neoclassical Chicagoites, against whom the and the Austrians engaged in many disputes were the intellectual proteges of Frank Knight. Knight's most important book was his first book, Risk, Uncertainty and Profit, published in 1921, a revised version of his doctoral dissertation at Cornell, and one of the most important and one of the most highly cited books in the history of economic thought, known obviously in particular for making the distinction, which I'll discuss a bit later, between risk and uncertainty. Knight wrote a number of other important works on different aspects of economics and its social, moral, and political implications.

6:26A collection of essays on ethics, called The Ethics of Competition, published in the 1920s. A less well-known but also very important book called The Economic Organization, not about firms as organizations but about sort of development of a kind of circular flow model of the U.S. economy and then Knight published several essays and letters on capital theory critiquing the Austrians, many of which were collected in a book called The Quantity of Capital in the Rate of Interest and Knight also wrote a very interesting methodological article that I think has been underappreciated by Austrians called Fact and Value in Social Science published in 1942. What were some of Knight's major contributions? The distinction, the idea for which Knight is best known among contemporary economists is the distinction between risk and uncertainty. Risk being described by Knight as situations in which actors do not know exactly what will obtain in in Future Periods. But they can describe the set of possible outcomes in a fairly precise manner. In other words, one can enumerate a list of things that could possibly happen tomorrow and assign probabilities to each of those events occurring, such that one can, in the familiar contemporary jargon, use expected value theory to come up with the optimal outcome in the face of these unknown future events. But Knight argues

8:01that in many aspects of life, different ways that human beings act, we're not faced with that kind of situation. We have to act under conditions in which we do not even know the set of possible outcomes that could occur tomorrow, much less have the ability to assign probability weights to them so that we can make expected value calculations. Rather, we have to act with a different kind of understanding and exercise what Knight called judgment, meaning a way of evaluating future conditions, of making conjectures about future outcomes in the absence of known, fixed and stable probability distributions. Actually, the first half of, or almost two-thirds of Knight's book, Risk, Uncertainty and Profit, is not devoted primarily to this issue, but rather to the development of a model, if you like, of perfect competition.

9:02Knight's, the second half of the book focuses on resource allocation under conditions of uncertainty, a condition that of course is absent when perfect competition is present. And Knight explains the role of the entrepreneur exercising judgment under uncertainty in directing the allocation of resources in a market economy. In particular, Knight develops the notion that an entrepreneur's judgment is non-contractable, meaning that one cannot buy and sell judgment on the market, but rather judgment can only be exercised through action, specifically by ownership and control of material resources, which are then combined and recombined in various ways in anticipation of earning profits in the future.

9:52Knight also did some work on investment theory, the sort of standard notion that investments will tend to be made across opportunities, across investment opportunities until the rate of return is, rates of return are equalized at the margin. This notion was elaborated by Knight in his book called The Economic Organization. He also wrote an interesting piece on congestion pricing, long before Ronald Coase and the modern Chicago law and economics approach to dealing with external benefits and external costs. He wrote a critique of Pigou who had argued that there needed to be regulation over publicly used resources like roads to make sure that they weren't overused and Knight pointed out, where's Walter Block? Walter's not here but much to Walter Block's delight that privatizing the roads and allowing the road owners to charge appropriate prices would be an alternative to Pigou's solution for a way of mitigating the congestion problem.

11:01Now, you know, Austrians, many Austrians are familiar with Knight's critique of Hayek and Boehm-Bawerk on capital. Capital. That's not going to be my main focus today. Obviously these differences were important and were very deep. Some of them rested, perhaps it's not quite right to say they rested on misunderstandings, but the key to understanding the differences between the Austrians and Knight and to some degree John Bates Clark, who also took Knight's side in this debate, is to realize that they conceived of capital in very different ways. That for Knight, capital Capital in the economically meaningful sense was not, as the Austrians conceived it, a stock of heterogeneous resources, specific capital goods, rather Knight and Clark conceived of capital as money, as a fund of value that would maintain its value, even in the absence of particular productive activities on the part of agents, and that interest in Knight's The Knight's formulation was not determined at all by time preference, by sort of technical considerations of the marginal productivity of capital.

12:15You know, going back and reading through the literature and this long set of debates between Knight and the Austrians, just to give you an example, this was a recent paper in the history of political economy called The Knight, The Hayek-Knight Capital Controversy, University, has a little bibliography of the relevant articles and personal letters between Hayek and Knight where they hashed this out, and as you can see there was quite a lot. It's a lot to go through. And it's not easy reading because you find that in many cases the authors are using similar terms to mean entirely different things. So clearly, you know, Knight's Austrianism does not extend to certain critical aspects of Capital Theory, of Production Theory, the Marginal Theory of Distribution and so on.

13:04However, there are important similarities between Knight and the Austrians that deserve, I think, a second look. Knight's views on uncertainty have been characterized in different ways. There's almost a cottage industry of scholars producing huge amounts of secondary literature on what Knight really meant, what was Knight's epistemology really all about, and that's one indication that Knight was a very provocative writer, but not an especially clear writer, particularly on these kinds of topics. Some of Knight's critics, much to my amusement, have dismissed Knight's writings on methodology and on epistemology as being extended Austrian-style disquisitions on the foundations of human knowledge and conduct and the like.

13:54and the like is supposed to be dismissive, like, oh gosh, there they go again, the Austrians with their blah blah blah about human, you know, the foundations of knowledge and so on. So this is meant by the writer to be a dismissal of Knight. In other words, you don't bother to read Knight's methodological writings. He's as crazy as an Austrian. Actually, Knight's views on probability are extremely interesting. and it turns out that the way many people think of the risk uncertainty distinction is not exactly the way I think Knight meant it. He actually distinguishes among three different types of probability what he calls a priori probability where we can know on purely deductive grounds what is the likelihood of a particular event taking place by which he meant for example a coin flip right so simply knowing the properties of the coin that If we can assume that it's a fairly weighted coin that has two sides and so on, the probability of heads or tails coming up is 0.5.

14:59We don't need to perform experiments to come up with those numbers. We can deduce them in an a priori fashion. But there are other cases where the things we don't know cannot be parameterized this way using a priori criteria. and he distinguished two different cases what he called statistical probability and what he called estimated probability so statistical probability describes cases where we have a fairly well-defined situation in which we can repeat a certain experiment over and over and over we can count how many times a particular outcome arises and we can infer from that what is the The Statistical Probability that any particular outcome will obtain in a series of sort of repeated trials, as opposed to other cases where we do not have enough occurrences of similar events to put them in a category where we can perform this kind of repeated experiment across, we cannot perform repeated trials because we have unique heterogeneous events that cannot be pooled in this fashion.

16:06Now, both a priori probability and what he called statistical probability, according to Knight, are categories of risk. So risk describes both of these two situations in which it is possible for different reasons to come up with a well-specified sort of characterization of the problem. There are X number of outcomes, each of which obtains with probability P and so on, as opposed to the situation of estimated probability, which Knight called uncertainty. Okay, so the risk uncertainty distinction is really a distinction among these three different types of what Knight called probabilities. Now many of you will recognize, particularly in the way I described Knight's approach to statistical probability, that this sounds very much like the so-called frequentist approach of Ludwig von Mises' brother, Richard von Mises.

17:02So Richard von Mises, a very famous natural scientist, very important figure in the history of statistics, professor at Harvard University. Richard von Mises is famous for distinguishing between what he calls class probability and case probability. So, probability for Richard von Mises and the so-called frequentist school is sort of the mathematical limit of what we observe in a series of repeated trials, as I described, whereas what Richard von Mises calls case probability refers to situations where each event is a unique case and therefore we cannot perform these repeated trials and come up with statistical estimates of the likelihood of particular things happening.

17:53happening. In fact, the term case probability is a little bit of a misnomer. What Richard von Mises means is that probability only applies to situations where we have members of a homogeneous class. So case probability is really case non-probability or case something else, okay? The terms case, class and case probability are well known to Austrians because the same Human Terminology was adopted by Ludwig von Mises in his own treatment of these issues in Human Action. And actually, my chapter in the Hoppe Festschrift, which came out in 2009, I think, 2008, 2009, deals specifically with the positions of Richard von Mises, Ludwig von Mises, and Knight, and also Hans Hoppe, who has written an article on this, on the distinction among different So, while Knight did not exactly reject the so-called subjective probability of Keynes in which all probabilities are simply subjective value judgments formed by individuals, he did believe that there were qualitative differences between these kinds of estimates that are formed in the minds of judgmental entrepreneurs and the kinds of probabilities that we can deduce either a priori or through a series of repeated trials.

19:16Trials. But sort of the key to what Knight had in mind here, when Knight distinguishes between risk and uncertainty in the way that we're all familiar with, really what he means by uncertainty is cases in which the actor, the entrepreneur, cannot articulate his or her beliefs about the future in a way that those beliefs can be communicated to other actors. When the categories of knowledge themselves are unknown, they cannot form the basis of interpersonal Agreement, and Market Exchange, according to an interesting article by Dick Langlois. That's not 1930, no, that's obviously a time, but I think that's 1993. They're not that old. Langlois is not that old.

20:01In other words, the reason that entrepreneurial judgment is non-contractable is because the entrepreneur cannot articulate and, effectively, his assessment of future conditions in a way to sort of sell that judgment to someone else. The only way the entrepreneur can put his judgments into practice is by owning and controlling resources, combining them with other resources, and so on, by being a business owner, a business operator, etc. So, 90 and uncertainty is really about our ability to communicate information that is highly difficult to articulate. This, of course, will strike many of you as familiar to Hayek's distinction between explicit and tacit knowledge, which is often interpreted by modern scholars in terms of precisely the same dimension, the ability to communicate information from one person to another.

20:58What Hayek means by tacit knowledge is knowledge that cannot be articulated in a way that it could be, for example, transmitted to the central planner, who would then make resource allocation decisions. But rather this knowledge can only be utilized by actors in the process of action in the marketplace. You know, Knight's methodological views, as I mentioned before, are somewhat inconsistent, but he has some very interesting statements that make him sound much like a Misesian praxeologist, though perhaps not a consistent one. For example, one of my favorite quotations from Knight about the importance of a priori theorizing is the following statement, the first fact to be recorded is that economic like reality exists or is there. This fact cannot be proved or argued or tested. If anyone denies that men have interests or that we have a considerable amount of knowledge about them, we meaning the economist using deductive reasoning, economics and its entire works will simply be to such a person what the world of color is to the blind man. Okay, so if

22:12If you cannot accept the fact that man acts and that we know certain things and can infer certain things about action and choice and preference and exchange from the primordial fact of the existence of economic reality, then, you know, economics is going to be to you what color is to a blind person. Okay, night continues, but there would still be one difference. A man who is physically, ocularly blind may still be rated of normal intelligence and in his right mind. And he also says, speaking of the kind of positivism that we associate with his Chicago successors, the positivism of Friedman's 1953 essay, Knight says this positivism is, quote, the emotional pronouncement of value judgments condemning emotion and value judgments, which seems to me a symptom of a defective sense of humor.

23:06Now, I've written on previous occasions about Knight's approach to the entrepreneur, and in fact, part of the reason for giving this talk is to make a plug for my next book, which will be coming out later this year, co-authored with Nikolai Foss called Entrepreneurial Judgment and the Firm, published by Cambridge University Press, which elaborates Knight's theory of judgment, his approach to the entrepreneur, and its implication for market competition and so on. And just a few points to be made are that for Knight, contrary to the way Knight has sometimes been interpreted, the model of perfect competition is not meant as a kind of ideal that we should strive for or that public policy should try to achieve, but rather, as I mentioned before, a kind of analytical foil against which real world outcomes can be understood, analyzed and so on. Entrepreneurship is best conceived as this kind of judgmental Decision Making by Resource Owners under Conditions of Uncertainty.

24:07So as I mentioned before, judgment cannot be hired on the market, according to Knight, but is exercised through the ownership and control of resources. Knight says in Risk, Uncertainty and Profit the following, the only risk, quote unquote, which leads to a profit is a unique uncertainty resulting from an exercise of ultimate responsibility which in its very nature cannot be insured nor capitalized nor salaried. Profit arises out of the inherent absolute unpredictability of things, out of the sheer brute fact that the results of human activity cannot be anticipated and then only in so far as even a probability calculation in regard to them is impossible and meaningless.

24:52In other words, profit arises out of the sheer brute fact that we cannot predict the consequences of our actions with certainty. You've got to love anybody who writes about sheer, brute facts in the context of economic analysis. Now, this is more or less the same formulation that you get in the proto-Austrian account of Richard Cantillon or Cantillon, if you prefer. It's implicit and partly explicit in Menger's own approach to the entrepreneur in his 1871 Principles. And it's very close to the way Mises describes the entrepreneur in Human Action and other works. Just a couple of quotes from Mises, maybe now you can see them in the back. Mises says, the term entrepreneur, as used by catallactic theory, meaning the theory of markets, means acting man exclusively seen from the aspect of the uncertainty inherent in every human action.

25:50So the entrepreneur is the agent acting under conditions of uncertainty, nightian uncertainty, if you like. Elsewhere in human action, Mises writes, the real entrepreneur is a speculator, a man eager to utilize his opinion about the future structure of the market for business operations promising profits. This specific, anticipative understanding of the conditions of the uncertain future defies any rules in systematization. So think about what Mises is saying. The real entrepreneur is an economic agent acting under conditions of uncertainty in pursuit of profit. And these uncertainties defy rules in systematization.

26:37In other words, they cannot be written as a mathematical, you know, expected value decision problem with known probabilities and so on. How am I doing on time, Mr. Moderator? Okay, where's your sign that says... No warning? Okay, I do get five minutes, okay. Well, I'm pretty much done anyway. Yeah, Joe's just enthralled, he's lost track of time. So, again, to go back to my slightly cheeky title of Frank Knight as sort of a forgotten Austrian, really what I have in mind and what I want to challenge you to think about is, you know, sort of a mild reassessment of the relationship between Austrian and neoclassical economics.

27:27And by neoclassical, I don't mean the Samuelson, de Bru, Krugman version or whatever is sort of the currently, whatever you consider to be the currently fashionable flavor of highly formal and technical mathematical analysis, but rather what we might describe more broadly as the theoretical tradition that came from Walras, William Stanley Jevons, their successors in Europe and so on. In other words, the main theoretical school of thought that was sort of a friend to, though sometimes in opposition to, the Austrian School. So there are aspects of old neoclassical economics, if you like, from which Austrians can learn a lot and which have a lot more in common with the writings of Menger and his followers than many of us seem to appreciate.

28:21Sometimes people pull up these quotes from writers like Mises and Hayek implying that there isn't much of a difference between the Austrians and the neoclassicals. Sorry, Mises says in one of his essays that's in Epistemological Problems of Economics, he says that the three marginalist schools, those of Menger, Jevons and Volra, quote, differ only in their mode of expressing the same fundamental idea and are divided more by their terminology and by peculiarities of presentation than by the substance of their teachings. Now that's kind of a startling statement to most of us in the room, right? Mises thinks that the only difference between Austrian economists and Volrasian economists is language and style rather than substance.

29:10Well, I don't think that's what Mises meant. What he did mean is that, look, the thinkers and writers with whom Mises had the biggest beef were not the neoclassical economists of the day, but rather the anti-theoretical Economists of the German Historical School, the Old American Institutionalist School, and so on. Mises is, the context here is that Mises is arguing in defense of theoretical economics as opposed to a theoretical historicist economics, which Mises condemned in very strong terms. And what he meant is the other schools of theoretical economics are very close to our own Austrian school in opposition to other kinds of social thought, right?

29:58So even today, if you were in Di Lorenzo's talk yesterday and other talks about where Austrians have critiqued behavioral economics and some kinds of experimental economics and so on, when I read debates between behavioralists and sort of traditional Chicago price theory neoclassical economists, I typically find myself on the side of the ladder thinking, Hayek says something similar in his entry that was in the International Encyclopedia of the Social Sciences. Hayek says the Austrian school, quote, can hardly any longer be seen as a separate school in the sense of representing particular doctrines. A school has its greatest success when it ceases as such to exist because its leading ideas have become a part of the general understanding of the social sciences. Hayek says the Austrian school, quote, can hardly any longer be seen as a separate school in the sense of representing particular doctrines.

30:46A school has its greatest success when it ceases as such to exist because its leading ideas have become a part of the general dominant teaching. The Vienna School has, to a great extent, come to enjoy such a success. Now, I think that's an overstatement by Hayek. I'm sure you would probably agree. But you can sort of see the point that Hayek is making, that many of the key insights from the Austrian version of marginalism, Austrian insights on, you know, the theory of distribution and so on. I mean, many of those have to some extent been incorporated into the mainstream quote unquote. If we define the mainstream broadly enough that we don't just mean MIT style micro theory or, you know, certainly not Keynesian macro or whatever. And again, remember that there was an important school of American economists and also some British economists like Wickstede, right, who were Followers of Menger and worked very hard to develop aspects of Austrian price theory and to incorporate it with some institutionalist and some Marshallian views.

31:55And remember that these so-called American Austrians or Anglo-American Austrians were very influential on Mises and on Rothbard. and Rothbard. Actually, Joe Salerno has pointed this out, that there's huge influence of J.B. Clark on Mises, that Mises says that in Boehm-Bawerk's seminar, they read a lot of Clark and Clark's development of what became known as comparative static analysis was very influential on Mises in Mises' development of his own sort of set of analytical tools. And of course, Rothbard's Man Economy and State, if you like, can be interpreted as sort A blend of Viennese-Austrianism and Wickstede, Fetter, Clark, Davenport, and so on, that Man Economy and State is littered with insights that come out of these so-called Anglo-American Austrians, insights that were not really appreciated or not incorporated into the works of Mises, Hayek, and the previous generation of Austrians.

32:57So to conclude, I published an article in 2008 that was called The Mundane Economics of the Austrian School, which has been extremely influential in my household, but not much beyond that, where I try to sort of make the case that some parts of contemporary Austrian economics may have lost their way to an extent by focusing on some of the more exotic aspects of Austrianism, rather than the more plain vanilla mundane aspects of value and exchange and price and so on. Those that were developed in particular by scholars like Fetter and Wickstede and Davenport and Clark, Rothbard, et cetera. And, you know, in the article I contrast the view that I'm espousing with the view of some other modern Austrians, like Karen Vaughan, for example, whose 1994 book on the modern Austrian school has been very She's very influential in sort of interpreting the Austrian, modern Austrianism, but what Karen Vaughan says about Austrian price theory is mostly derogatory, right?

34:06She says Menger's price theory is that of a quote half-formed neoclassical economist. She likes Menger's second book on institutions and evolution and so on, but thinks that Menger's principles is really not all that radical or important. She says, Karen Vons is of Human Action, that it combined quote, some fundamental Mengerian insights with the apparatus of neoclassical price theory to the detriment of both. Of Man Economy and State, she says, it quote, must have seemed to a typical reader to be more or less familiar economics presented almost exclusively in words with a few controversial definitions and some strange discontinuous graphs.

34:51So, from the point of view of an Austrian who is less sympathetic to the price theory of Menger, Mises, and Rothbard, that price theory appears to be kind of similar to neoclassical price theory in many ways, with important differences, of course, and my response to that is to agree with it, that I think there are very important similarities between what we might call good neoclassical economics and the price theory of the Austrians. and the differences to be sure, but similarities that deserve to be explored. Joe has pointed this out in his introduction to the new, the second edition, second scholars edition of Man, Economy and State, something that many of us have noticed for a long time. If you just look through the footnotes of Man, Economy and State, they're shot through with references to the contemporary literature in the American Economic Review and the Journal of Political Economy and so on.

35:44You know, up through about 1950 or the mid 1950s. When Rothbard was writing Man Economy and State in the late 50s, he was learning a lot and gaining a lot of insight from much of what was in the mainstream neoclassical literature of his day, criticizing it to be sure and trying to work through its inconsistencies and contrast it with what he found in Menger and Mises and others. But he was still gaining a lot of insight from that. But by the 1960s, 70s, and 80s, at least from personal correspondence, Rothbard had lost interest in that aspect of the mainstream tradition, thinking that it had gone too far from its own sort of correct roots, if you like. So in other words, to make a long story short, if we place Knight in this same context, being an important early 20th century economist who drew on the so-called Anglo-American Austrians as Rothbard did, who was part of that intellectual milieu, and who had a lot of ideas on economic

36:47Theory, that are congenial to Austrians, I think the Austrian School can learn quite a lot from taking night, reconsidering night, and thinking a little bit more about his contributions.