Time and Economics | Mario Rizzo

Once upon a time, when I was a student at Fordham University, I wrote a letter to Ludwig von Mises asking him to come to Fordham to speak. I tried to be as respectful as possible and also remind him that we were just poor students, so we couldn't give him any big honorarium. I sent the letter off, and a week or so later I got a phone call. On the other side of the phone was a man who said, 'Mises here.' So I stood up straight. We started talking about what he would speak about, and I tried to explain that the organizations sponsoring the lecture were the Fordham Libertarian Alliance, the Philosophy Club, and the Economics Club. I thought a good topic that would get interest in all the groups was epistemological problems of economics. He said, 'No, I don't want to talk about that.' I asked what he wanted to talk about, and he said he wanted to talk about inflation, or as he said, 'money problems.' I tried to explain, but finally he agreed to a compromise: he would speak on epistemological problems of money. I had no idea what that would be, but I thought it was a good solution. So he came and spoke about inflation. I've been asked to speak about time and economics, and so I'm going to speak about something else. Seriously, I don't think I can speak about time without talking about equilibrium and learning, so I'm going to try to mix all three together. It's also the case that you've been going through a lot today—this is your seventh session—so I'm going to try to keep this as pleasant as possible. There's a statement by Saint Augustine of Hippo that I like: 'What then is time? If no one asks me, I know; if I wish to explain it to one who asks, I do not know.' In some respect, the less you think about time, the more you know about it; the more you think about time, the less you know. If you really want to get into this, I can recommend an article by myself on my Mises Press site, published in 2000, 'Real Time and Relative Indeterminacy in Economic Theory.' But for now, I want to start out by talking about Hayek's project in the late 1930s and early 1940s, which was to marry equilibrium and time. The great Hayekian contribution, and also contributions from people like Erik Lindahl, was to center economic analysis on the idea of the plan and to talk about equilibrium in terms of plans. Once you talk about plans, you're talking about actions over a period of time, and what unites them is that they are integrated and consistent. Hayek chose to look at equilibrium in terms of plans rather than specific actions. There are four elements to this plan analysis. First, the plan from the individual's point of view and whether it is in equilibrium. The idea of individual equilibrium is usually taken for granted in economics, co-extensive with rationality. Hayek suggested it was more or less a tautology. We can have plans at the individual level, and we can think about whether actions are integrated. There is an ex-ante perspective and an ex-post perspective. An individual may have plans based on certain assumptions about the world, and those assumptions may prove true or not. So the individual could have a self-consistent plan ex-ante but not ex-post. From an individual perspective, we can talk about equilibrium in terms of the relationship of actions within a whole complex of actions—the plan. So we have brought in time immediately because we have scheduled actions over a period, and we've allowed for a difference between ex-ante and ex-post. As I said, Hayek thought the idea of ex-ante individual equilibrium was more or less a tautology. In recent years, that has been challenged by behavioral economics, which tries to show that individuals do not obey various consistency requirements in their planned actions over time. For example, individuals plan to stop smoking and then change their minds. I think a lot of that can be interpreted as suggesting that individuals do not automatically have consistent plans. But even today, most economists think of individual equilibrium as pretty much taken for granted. The second column has to do with the social aspects of equilibrium—the interpersonal compatibility of plans. If I think it's going to rain and plan to make umbrellas, do you also plan to buy umbrellas? Is there compatibility of plans ex-ante? But ex-post, it may turn out that the anticipation of rain proves incorrect or correct. If correct, the ex-ante equilibrium is transformed into ex-post; if not, we have ex-post disequilibrium. This is a good diagram expressing Hayek in his 1937 article 'Economics and Knowledge.' But the story goes beyond that. Time is in this model in a very static way—you allow for disappointment, but once there is disappointment, plans are incompatible, and that's that. The real dynamic aspect rests on what happens when there is disequilibrium in the system. We have to distinguish between mutual compatibility of plans and more generally coordination. In 1937, Hayek said that complete compatibility of plans need not be a Pareto optimal situation. One way to interpret this is to think of Kirzner's idea of the entrepreneur as an arbitrager. Imagine two parts of a single market where the same good is traded at different prices because the parties are unaware of each other. In each part, you could have complete compatibility of plans—no one is bumping heads—yet it's not Pareto optimal because there is room for arbitrage. Complete coordination of plans would be a case where not only is there mutual compatibility but also all arbitrage opportunities are eliminated. The reason this distinction is important has to do with learning. In the 1937 article, Hayek emphasized that if there were incompatibility of plans, there would be an immediate incentive to do something different. The parties would know something had to change because they would be directly frustrated. However, in the case of lack of complete coordination, the learning is not as obvious because no one is bumping heads. It's a case of ignorance is bliss. In the case of mutual incompatibility, Hayek says the parties are bound to find out what it takes to produce compatibility. It's easy to understand how they're bound to find out there is incompatibility, but how they're bound to find out what the equilibrium is is a separate question. If we're dealing with parties in direct contact trying to engage in mutually beneficial trade, the chances of finding equilibrium are probably greater. So I interpret 'bound to find out' as a statement of relative ease. So we have criteria for a full Hayekian equilibrium: individual consistency of plans both ex-ante and ex-post, mutual compatibility of plans across individuals both ex-ante and ex-post, and exhaustion of profit or arbitrage opportunities. The status of this equilibrium concept is that no Austrian, and really most other economists, would have looked at these equilibria as a description of reality. It's simply an analytical tool. Hayek himself said that some assumption about a tendency toward equilibrium is necessary for economics. This idea of equilibrium justifies itself by a notion that there is a tendency toward it. The word 'tendency' is messy and vague. Some interpret it as meaning we get arbitrarily close to this position in the real world, and there is some evidence in Hayek's 'Pure Theory of Capital' that he thought that way. But I think a fairer evaluation is that we can understand processes or disequilibrium states by thinking about them as adjustments that, in the absence of disturbing factors, tend to re-establish equilibrium. In these earlier articles, one difference in how Austrians have looked at this is that some, including Hayek at this period, want to throw everything that disturbs equilibrium into a category called exogenous factors—shocks, demand shocks, knowledge—all outside the system. Then once the system is disturbed, there are factors within the system that constitute the tendency to re-establish equilibrium. So adjusting change is always an equilibrating factor, but it's that way by construction because you throw everything that doesn't tend to re-establish equilibrium into exogenous factors. So to say there is a tendency toward equilibrium is analytically correct but should not be interpreted as settling the matter empirically. Professor Lachmann used to say there are equilibrating forces and disequilibrating forces. He didn't appreciate that the argument about tendency is kind of definitional, but he was right to raise the question of what this says about the real world and whether new equilibria are quickly re-established. So we have to think about the process of adjustment from a more realistic point of view. This is where the two conceptions of time come in: real time and Newtonian or static time. The key difference is whether the passage of time necessarily involves changes in knowledge. In a static conception, the passage of time does not necessarily involve any change in knowledge. You can have time changing and nothing much happening, as in a plan equilibrium. On the other hand, if you think of time as inextricably linked with changes in knowledge, your plan is likely to be much more flexible because you know something unexpected might happen. So the character of planning under a real conception of time would be very different—plans would be much vaguer. That leads to a transformation of the concept of equilibrium. If agents operate in real time, to say their plans mesh is not to say they mesh in any precise way, but that they may mesh allowing for imprecision. In some sense, there will be an optimal amount of discoordination. Now let me move ahead to Hayek in the late 1960s and 1970s. You may be familiar with his article 'Competition as a Discovery Procedure.' Since competition is the essence of the market process, you could think of this as saying the market is discovery. I bring this up because if we have plan incompatibility, agents have to discover what will render their plans compatible. It's not true that they're bound to find out. In the case of the elimination of arbitrage opportunities, there's no being bound to find out that there are opportunities somewhere. So the idea of deterministic learning is outside the picture. We have to think about the market as a discovery process. In that article, Hayek makes much of the idea that competition is desirable precisely because it produces results that are unpredictable. He wanted to contrast his view with the standard view, which has certain conditions that mean we are in perfect competition. In that view, competition is like a computer that churns out the right results and enforces them. On the other hand, if you take a Hayekian or Schumpeterian view, competition is really about discovering new ways of doing things. Competition tends to produce surprises. Now go back to the situation of plan incompatibility or lack of complete coordination. I'm an entrepreneur looking for arbitrage opportunities. If they were known to everybody, there wouldn't be any. So if I act on these opportunities, I'm engaging in actions that many other people could not have predicted because they don't know of the opportunities. So my attempt to discover opportunities creates a situation where I engage in unpredictable actions. If we focus on a particular market with two parts, an entrepreneur discovers the arbitrage opportunity, and you can tell a story about how prices harmonize. But my activity in buying and selling is not something the agents could predict. If you multiply this throughout many markets, entrepreneurs engage regularly in activity that cannot be precisely predicted. The upshot led Hayek to doubt the usefulness of the traditional equilibrium concept. He said that a better concept than equilibrium is order, like a stream in the water—paying some attention to the banks but coming outside. Order is much less precise than our normal idea of equilibrium. It has to do with certain parameters we can rely on, but there's going to be a lot of unpredictability. This is very much like time and changes in knowledge going together. If time is static, everything would stay within the banks and move precisely as predicted. In real time, knowledge is changing, people are doing all sorts of things that make the world much less precise because you can't predict what knowledge they'll acquire. As Popper said, future knowledge cannot be acquired before its time. So competition as discovery creates a world where a certain amount of unpredictability is built in. Earlier, I said there was an analytical device that puts all things that disturb equilibrium into the category of exogenous factors. But now we see that the very essence of adjustment—entrepreneurship—is the discovery of opportunities, and entrepreneurial discovery is essentially an unpredictable phenomenon. So the factor responsible for adjusting to disequilibria also produces unpredictability and discoordination. The two things are mutually connected in a real-world sense. I'm not denying that analytically you can separate them, but in reality, the entrepreneur's discovery may undo others' plans. So we are back to Hayek's notion of order. He said in 'Competition as a Discovery Procedure' that a high degree of correspondence of expectations is brought about by the systematic disappointment of some kinds of expectations. Competition produces novel, unpredictable things; some expectations are disappointed. But the necessary process for there being as much coordination as possible is this process of competition. Without it, we would have almost no coordination. But the process of discovery is not simple; there's a lot of sloppiness, incompatibility, and disappointment generated by competition itself. This figures into the ethics aspect—the mirage of social justice—namely that it's impossible to protect people's expectations of market value without paralyzing the discovery process. Paradoxically, if we try to guarantee expectations, we disable the process responsible for coordination. So Hayek evolved in his thought, which can be seen by contrasting the 1937 piece with 'Competition as a Discovery Procedure.' All of this involves a natural combination of ideas about time, planning, coordination, and competition. My excuse for wandering off from purely talking about time is that it's not really possible to talk about time and make sense of it without these other concepts. As Marshall said, time is the center of the chief difficulties in economics. All progress hinges on coming to grips with the importance of changes in knowledge that accompany the passage of real time. Thank you. Sorry about my voice; I think it was something in the food.

Yeah, my name is... Well, first of all, Mises said he liked the article very much. He makes a distinction between the pure logic of choice and the learning that has to take place because there isn't a single entity to whom all knowledge is given at the outset. One interpretation is that Hayek was getting at that Mises, in his emphasis on praxeology and deducing theorems from purposeful behavior, was really just developing the logic of choice. But if he wanted to explain how there can be compatibility of plans among agents with separate minds, there must be attention to the process of learning and communication of knowledge. That is empirical—how people learn cannot be deduced from the notion of action. That's correct in the sense that you cannot deduce useful ideas about learning from the idea of purposeful behavior. The question is whether Mises ever thought otherwise, but analytically it is correct to say that deductions from the notion of action produce tautologies, and these tautologies don't involve how people learn. So they are separate domains. Some of that difference between Hayek and Mises still persists today with people who think they can derive a lot more a priori than they can.

Well, that's a big issue. Hayek never spelt out in any detail the evolutionary process he had in mind except to say it involved group selection. It's hard to agree that this process produces results consistent with what he said. In the context of the common law, it's supposed to produce legal rules that embody the experience of generations and should have a presumption of being valuable. But none of that is clearly worked out. I think it's more suggestive than definitive. One thing missing is the recognition that the legal process in the early days of the common law was competitive—different courts competing with each other—and that's not the same as 19th or 20th century common law. That's why you can get a situation where common law produces things inimical to a market order. What changed is that the production of legal precedent is much less competitive today and more affected by statutory law. So these are considerations one has to raise. I agree more with Menger on this than Hayek. Menger says there's a lot of value in the common law, but every generation has a responsibility to rationally evaluate the law. That's probably contrary to Hayek, but I agree with Menger: a certain deference, but don't go too far. Feel free to criticize. The basic bottom line is Hayek never specified the mechanism very well.

I blog on macro topics but I don't speak on them, so I have no comment. It's because I don't know enough.

Well, if you think about a plan as a mental object, you have certain predictions, and the mental picture is very Newtonian and static. When you implement a plan, the relevant concept of time is more real time because you see the unpredictable come about. The way they are related is that as a planner, I know that the best plan I can make won't be good enough because things I can't predict will happen. So the natural thing is to make a plan with a certain degree of flexibility. Good plans allow you to change when something unexpected happens; they don't commit you rigidly. There's a related idea in an article by Vernon Smith about profit maximization versus survival. If you plan to maximize profit, there's a chance you'll fall so far below positive profits that the firm goes out of business. So better maybe not to literally maximize profits but to satisfice or modify maximization to ensure you're unlikely to fall below survival. Over the long run, you're not maximizing, but at least you're in the game. You know something unexpected will happen, so you don't want to be knocked out. You may plan not to do the optimal thing because the optimal thing doesn't produce survival over the long run. Flexibility is another example. Firms can try to predict output, but predictions won't be exactly correct, so you may want a plant size or structure that is flexible, allowing you to increase or decrease output without high costs, even if it's not optimal for any particular output. This is fairly standard, but if you think about it from the point of view of uncertainty about the future, it affects the nature of plans. So the real time aspect interacts with the Newtonian planning time aspect.

Well, I think what Hayek wanted to talk about was that a tendency toward equilibrium is necessary to give some direction to what happens outside of equilibrium. We can't take for granted that plans will always be compatible. That's a critical aspect of his analysis. What was wrong with the standard analysis at the time was that it simply took the existence of certain data and made it a purely computational issue. But that couldn't be right because the data is given to no one in particular; people have bits of it. So what's going to ensure an equilibrium? There are processes that occur outside of equilibrium as a response to disappointments, but where do these processes lead us? Without the notion of a tendency toward equilibrium, we would have no way of thinking about where the processes lead us, and we would have a kind of 'anything goes' method of analysis. Hayek thought that if we didn't know what happens outside of equilibrium, economics as we understand it couldn't exist.

Okay. And the other... our time is running out. We end at nine. Who's the keeper here? Okay. Any other questions? Okay, thank you.