How to Do Economics

Discussing the methodology of Austrian economics


The Tom Woods Show, November 22, 2013

WOODS: I want to talk today about the method of Austrian economics. Is it unscientific? What is Mises’ method of doing economics? And is this some kind of oddball method that only a weirdo would use?

These are the sorts of things that come up on the Internet quite a bit, and you get the sense that a lot of people talking about it probably haven’t read Mises’ epistemological works. You get the sense that they’ve learned it from three sentences they saw quoted by some guy one time. So I wanted to get you on here, because you’ve written about this so effectively.

Let’s start from the beginning. What is the nature of the dispute? What is this argument all about when it comes down to what Mises believed, or how Mises believed the economist should pursue his craft? What’s different between Mises and what the sort of man on the street might think is the way an economist should operate?

SANCHEZ: Well, you mentioned “scientific,” and that reminds me of a funny thing that happened a couple of years ago. Tom DiLorenzo went to testify to the House Subcommittee on domestic monetary policy, and Congressman William Lacy Clay accused Austrian economics of lacking what he called “scientific rigor,” and the reason for that was, he said, “It uses deductive reasoning.” So, I think it’s so funny, this McCarthyite-type situation where he’s saying, “Is it true, sir, that you are part of this group of people who use—dum-dum-duuuum—logic?”

WOODS: We can’t have that! So, in other words, it’s a question of deductive versus inductive reasoning, and a question of which type of reasoning is appropriate to a given discipline. And the question that Mises is trying to answer is: which approach makes more sense for economics? The approach we would use in geometry and in legal theory, for example? Or is it better to try to gather data in the scientific method sort of way, and go into economics more or less agnostic, and then see if we can derive general principles from observing empirical data? Is that more or less it?

SANCHEZ: Yes. You mentioned geometry, and that’s a really important comparison. People often will say, well, if you are not subjecting your proposition to these two empirical tests, then that’s a dogma. That’s a religion. I wonder if they would accuse geometry of being a religion. I mean, geometry also doesn’t put its propositions to the test of experiment, and of experience in general. Geometry is what is called an aprioristic discipline. That’s kind of a strange word, it means that the theory of geometry logically deduced is prior to any kind of experience. So any kind of experience with measurements of objects in reality—Euclid’s Elements, all the system of geometry logically deduced, doesn’t depend on the measurements we take of real-world objects. And they can’t be invalidated by such measurements. And that is not a controversial idea. It’s not just Austrian surveyors and Austrian engineers who treat geometry as prior to dealing with real-world objects. So, it’s not to be just laughed out of court just because it’s considered prior to experiment.

WOODS: So, in other words, nobody would say, “Hey, you dogmatic geometer. You’re telling me you believe that the sum of the squares of the two sides of a right triangle equals the square of the hypotenuse and you haven’t measured a single triangle to verify this? What’s the matter with you?” Nobody would act that way. Nobody would be, frankly, stupid enough to act that way. So we all see that an empirical approach to geometry makes no sense and is ridiculous and fails to understand the nature of geometry. But what would make us think that the nature of economics is such that the method would be similar?

SANCHEZ: In the natural sciences, the phenomena that they’re describing are characterized by regularity. But human action is not characterized by strict regularity where, just because you see a phenomenon happen in certain conditions in the past, it necessarily must happen in the future. So, Mises showed how the character of economics is more akin to geometry, especially because just as with geometry, there are certain implications that are bundled up in these basic concepts that everyone introduces into their reasoning.

Wood: All right, let me jump in here. This is a good starting point for the whole analysis. Mises uses the term “a priori” or “synthetic a priori”—that we can have a statement that’s meaningful and yet that we can know prior to all experience. The action axiom is supposed to be an example of the synthetic a priori. So can you explain that?

First of all, what’s the action axiom, and what do we mean by calling it an example of a synthetic a priori statement? And, by the way, to the people listening today, you had no idea how awesome this podcast was going to be, did you? Now you’re learning about synthetic a priori statements. I mean, people are just going to love you now.

SANCHEZ: Well, an a priori statement is something that is prior to experience. And calling it a synthetic a priori statement basically is getting toward the fact that it applies to something in the real world. And that is something that is characteristic of Austrian economics, as opposed to certain other schools of economics, that it’s very realistic, that it does apply directly to the real world.

Now the action axiom basically is that human action exists, or that man acts. And actually a funny thing that not many people are aware of is that Mises himself never used the phrase, or the term, action axiom. He never posited it as a proposition that man acts. He called it the category of action. So he focused on the concept itself, of action, and what can be logically deduced from even thinking in terms of action in any case. And so, what can be unpacked from the concepts of action? Every thinker and social scientist—of all schools, not just the Austrian School—introduces concepts, especially action, that have certain implications that necessarily follow from them.

A lot of people challenge critics of Austrian economics and critics of the free market by starting off with, “Okay, well, do you think man acts?” And then the conversation gets sidetracked in all these sort of meta-philosophical objections and ruminations on the rationality of man. But really you don’t even have to put that proposition to them. You can just point out that your opponent, in his own discussion of human society, is himself positing action. If he is positing action, then he has to accept the logical implications that necessarily come from action.

So, for example, there are certain things without which the idea of action would be incoherent. For example, time. Try to think of an action that doesn’t involve time. The human mind cannot even fathom such a concept, and so time is a logical implication of action. Also, the notion of imperfect conditions in light of the actor’s judgment—that if a person didn’t think that conditions would be imperfect without his intervention, he wouldn’t act. The very notion of a person expecting perfect conditions, with or without his intervention, who then acts anyway—you can’t even imagine why such a person would act. So these are some basic logical and necessary implications of action, and any social scientist who even discusses in terms of action, to be even logically coherent, has to accept these implications.

WOODS: Let me mention some other implications of the action axiom, which says that man acts, or, in other words, that people have goals, and use means to pursue those goals. There are very clear economic implications of this statement that everybody can understand. For example, costs exist. Every time I act, I’m implicitly setting aside other things that I might have done instead. So if I sit on a park bench and eat a ham sandwich, I am setting aside flying a plane at the same time. I am choosing and setting aside. In choosing and setting aside, in turn, I’m demonstrating that I prefer one thing over another, and this gives rise to the idea that there are value scales in my mind. I have a video online in which I start with the action axiom and I end up showing people how supply and demand curves are derived, where the law of marginal utility comes from, and it all just comes from explaining the implications of the seemingly uninteresting statement that human beings act.

SANCHEZ: Exactly. Mises argued that all means, or you could also say “goods,” are necessarily scarce. And by scarce he means that the quantity available of the good is outstripped by the goals that a person has in mind for it. So the very concept of using something that is scarce necessarily implies, as you said, the notion of pursuing some ends and leaving other ends unpursued. So action with regard to scarce means necessarily involves choice, pursuing some ends and setting aside other ends. And a further implication is that when a quantity of a good is lost to the actor, then the actor will sacrifice certain ends, certain goals. Now, the goals that are sacrificed are, by definition, valued less that all the goals that he does not sacrifice. And from this reasoning, a thinker can deduce the law of marginal utility. And this law is bound up in the very notions of action and means and scarcity. And you can derive—and as you said, you derived it in your lecture, and that’s what Austrian economists are basically doing.

WOODS: Now is this some freakishly odd thing that only Austrians do? Can we find in the history of economic thought—I realize I’m stacking the question—that this was the mainstream approach to doing economics until quite recently?

SANCHEZ: Yes. Mises argued that throughout the history of economic thought, basically the message that the economists largely pursued—even though in their writings on methodology where they’re talking about methodology, they don’t necessarily endorse it—but in practice the kind of reasoning that they were doing is theoretical reasoning like we’re discussing here. And there are examples of some economists prior to the Austrian school that actually were pretty much explicit praxeologists. So praxeology is this idea of just theorizing from the basic concept of action. Murray Rothbard covers some of these examples in his Austrian Perspective on the History of Economic Thought treatise, which is a fascinating read.

You might also think that a lot of this is just very vague and very general, so, okay, if you just think about general concepts of action, then you come up with very, very general conclusions, but how does that help us in studying the complex world, where it’s not just that people are acting, but that people are acting in certain ways? But that’s also part of praxeology for Mises, because what we need to do then is think about certain modes of action. And so we need to restrict our notion of action by certain assumptions. And we think of different ways in which people can act.

So, for example, people can think of interpersonal exchange. That’s not a property of all human action, but you can imagine a Robinson Crusoe just sitting on this island, and he has no one to exchange with. But you can posit certain assumptions about what you’re theorizing about, and you can say, okay, let’s say that there’s another person, let’s say that Friday is on the island with him. And they can interpersonally exchange. Just thinking about that restricted kind of action necessarily has certain logical implications. So, for example, one logical implication is that both Crusoe and Friday mutually benefit or expect to mutually benefit from the exchange, because if they didn’t expect to benefit, why would they perform the exchange? By definition, they expect to benefit. So that’s one thing that you can know, just from the very concept of exchange, that must be true. And then, as you said you do in your lecture, you could take the concept of exchange, apply the law of marginal utility to that, and then, for example, derive the law of demand. And you can use that kind of reasoning to construct price theory and then use the reasoning of price theory to construct a profit-and-loss theory. And so this is the way Austrian economists operate.

WOODS: Now what about this kind of objection: “This sounds fine and everything, but the real problem with you Austrians is that you’re so dogmatic in your beliefs because of your so-called praxeology and your deductive reasoning, that you won’t even admit statistics into your analysis. I mean, really, what kind of thinker doesn’t even use statistics?” Now, is it true that Austrians just belligerently refuse to use statistics? What exactly is the relationship between Austrians and statistics?

SANCHEZ: Well, they do use statistics for economic history. They don’t use statistics to support their theory, but they use theory in combination with statistics to understand what actually happened in the past. And so this gets at the comparison between geometry and economics. Let’s say a geometry teacher assigns you the task of determining the measurements of an archeological ruin. And this part of this ruined building happens to have the characteristics of a right triangle on one of its faces. And now let’s say that you are able to directly measure only certain parts of the ruin. And the geometry teacher assigns you the task of using the theorems of geometry to derive the other measurements. Now let’s say that you do that, and you find out that the measurements don’t jibe with your understanding of the Pythagorean theorem. And you return to the teacher to denounce the orthodoxy and dogma of Pythagoras, and to proclaim your heterodox rival theorem about right triangles that is based on your measurement of this ruin.

Now that is akin a situation in which an Austrian economics teacher asked you to gather statistics to help you in economic history, but then when you gather your statistics, you decide that you think that this doesn’t jibe with Austrian economic theory, and so you go back to your teacher and say, I am denouncing the orthodoxy and dogma of Mises and Rothbard, and then you try to have your own alternative theory just based on your statistics.

Now if you think about the way that every geometry teacher would respond to that student, it gives you an idea of the way that an Austrian economist would respond to a similar student. There are a few fundamental errors the students may be making. Maybe the geometry student doesn’t actually know the Pythagorean theorem. Maybe he tried to derive it in a certain way, and he messed up. He didn’t reason correctly. And so he thinks the Pythagorean theorem is that a² + b² = c³, for example. So that’s why his measurements don’t jibe with what he thinks of as the Pythagorean theorem, because he doesn’t actually know the Pythagorean theorem. He derived it incorrectly. So, let’s say that he actually measures the artifact accurately, and from his measurements, he finds out that in that one case, a² + b² did equal c², well that gives him a clue that his reasoning was faulty. But those measurements themselves do not substitute for correcting the reasoning itself. You can’t just say I know now that a² + b² = c² because I did these measurements. What you have to do is use that as a hint and go back to your desk and your pencil and paper and then derive the correct Pythagorean theorem using validlyreasoned proofs.

Now the parallel to that is an Austrian economist who doesn’t know the correct theory, and in trying to derive his own theory he reasoned incorrectly, and so that’s the problem, and that’s why the statistics don’t jibe with his theory, because his theory is wrong and he needs to correct his theory through re-reasoning it. He can’t just rely on statistics.

Now another possibility is that it’s not a right triangle at all, so maybe the Pythagorean theorem doesn’t even apply. Maybe it’s even irrelevant to the situation that he’s considering. And so that would be equivalent to applying an irrelevant theory to a certain economic set of statistics. So maybe if you, for example, try to apply the theory of economic calculation to a barter society. It doesn’t even apply because economic calculation applies only to a market society, for example.

So, basically what that means is that every logical proof has premises and a conclusion, but the conclusion is only certain if the premises are given, and so his problem with the fact that he’s not even dealing with a right triangle, is that the premises involved in the Pythagorean theorem aren’t even given in this situation. So, of course, the conclusion of the Pythagorean theorem doesn’t apply to this particular situation. And, finally, another possibility, of course, is that the student just mis-measured. So the student just had human error in trying to take the measurements of the artifact, and that’s similar to just bad economic statistics.

WOODS: Well, Danny Sanchez, we’re just about out of time, so I appreciate your guiding us through what may seem tricky to the beginner, but, if you read Mises’s stuff on this, it’s not actually impossibly difficult to manage. On the TomWoodsRadio.com site next to this particular program, we’re going to make sure to link to your articles, specifically “Mises on Mind and Method,” which will help people understand these sorts of things and then sort of inoculate themselves against some of the ill-informed attacks that you might encounter online from time to time.

So, Danny, thank you so much for being here.

SANCHEZ: Thank you, and just to mention that in that article, the ideas that I reference in this interview, I link them to particular quotes in Mises’s works, where he makes these points.

Get a free subscription to the Tom Woods Show at TomWoodsRadio.com.