Behavioral science research; Decision making in economics

The following program is produced by the University of Michigan broadcasting service under a grant of aid from the National Educational Television and Radio Center in cooperation with the National Association of educational broadcasters decision making in economics. A program from the series human behavior social and medical research produced by the University of Michigan Broadcasting Service. The people you will hear today are Professor Jacob Marsh of the Department of Economics at Yale University and Professor Kenneth J erro of Stanford University. And my name is Glenn Phillips Professor arrow was at the time of our discussion on leave of absence from Stanford University and doing research at the Rand Corporation in Santa Monica California in economic decision making. I ask him to describe this and other research being conducted in this field. The following then is part of that interview as economists in general see it. The economic system is a. Very complex. Device for

coordination in. At least an ideal model. Each of the agents in the system. Makes decisions. On the basis. Of the prices of the various commodities. With which he deals. The consumer may make his decisions as to how to split his income derived from wages and possibly from interest or other forms of profit. Among the other commodities. That interest him. Daily commodities. Of. Food clothing. More durable commodities like. Housing. Appliances. Similarly the producer. Has a great many choices to make. We usually assume the producer is making decisions on the criterion of profit maximisation. But this is really a complex. Matter. The profits referred

to are not only present profits but also future ones. He must of course decide. About the amount of labor he wants to hire. The output he aims at. The raw materials which go into it. But he must also decide. The capital goods with which he is going to produce. His output. Factories. Machines. And the like. Decisions about capital goods are investment decisions. And because of the durable nature of the commodities. Included. They must. Be made not only on the basis of the present but also on the basis of expectations about the future. We have then a tremendous host of individual decisions but the very

nature of economy demands the coordination to put in the simplest economic language we expect that supply and demand are equal. There's no point in individuals desiring a great deal of butter if in fact there is not enough to meet their demand. Prices are regarded by economists. As a link. Which. Brings all these decisions into coordination. The demand by the consumer for example. For commodities will be affected by the price ordinarily of price rises demand is choked off. On the other hand as price rises the willingness of producers to supply goods. Also increases. The simple textbook model treats usually have a single market. And points out triumphantly. That eventually a price will be found.

Where the willingness of producers to supply is just matched by the desire of consumers to buy. Rights. My colleagues and in particular I may mention Gerard Dobro of Yale and Leonid which of Minnesota. And I. Have been seeking to. Understand more fully. The operations of the economic system. In a situation many. Connecting markets. But explain what I mean by. The connections among the markets. If the price say it better goes up we have something more happening than merely a decline in the purchase of the butter. We probably also have an increase in the purchases of margarine. We may also have a decline in purchases of bread. As a person likes bread with better

is perhaps going to be less interested in buying the bread. If he cannot buy the butter at the prices demand it. So a change in any one market will spread itself to other markets. Same thing incidentally is true on the supply side. And in many different ways. An increase in the price of butter increases the desirability. Of dairy farming and as a byproduct we have an increased supply of milk. The connections of course ramify endlessly. The improved position of the dairy farmers will reflect itself in the greater greater willingness to hire labor and increase in the hiring of labor may lead to a general increase in consumer demand throughout the economy. Depending on the general situation it may induce investment in milking machines. As an ultimate byproduct.

And as I needn't perhaps trace the consequence of this too far. But this formulation of an interdependent economy in this way. Is due to. The famous French economist lam virus. In the last quarter the 100 cent.. But the detail study the properties of the system. I read a property. Of the Presidents era and it's a particular the last 20 years or so. The problem as posed. Requires first of all a careful formulation of the economic system in precise language. Precise language here. I mean mathematics. It does not seem possible. To use the ordinary resources of the English language to express this vast

array of interdependent quantities. Only a highly symbolic terminology will suffice. One central concept in analyzing a complex system of any kind. Mechanical physical. Biological or economic is that of equilibrium. An equilibrium state is one which tends to perpetuate itself. In the economic context. We would like to know is there any set of prices. And decisions based on them. Such that they will. Tend. To remain the same. That is supposing we can find prices such that the individual's decisions are in the coordinated through the markets in the sense that supply and demand are equal in every market.

Such a constellation of prices will clearly an equilibrium. Set. Once the problem. Has been formulated mathematically. We can raise as a purely mathematical question. The question of the existence of such an equilibrium. This may seem a little surprising. But it is perfectly conceivable. That there may be no set of prices. At which the decisions of many of the many individuals. I really in represent each other. It can be shown. By. Rather advanced mathematical reasoning that under very plausible assumptions.

There exists an equilibrium prices for the whole economy. The existence of Christ has a peculiar meaning here. I mean is that an omniscient computer knowing everything about the decision making structure every individual could find such prices. You talk of equilibrium also decision. In. Human decision. The introduction of variables upsetting the equilibrium. To me would pose a certainly great problem. Just jotting down here some variables as they rapidly occurred to me there would be nature which would produce variances in the crop output. Then inventions which I think you yourself referred to earlier. Such things as strikes and also then the greatest of all variables probably the human decision factor how the individual himself decides to purchase or not to purchase. Here are the variables presented and I'm sure the list could be made

very extensive. But how do these variables tend to if they do counter act each other and thereby keep the equilibrium. He raised a problem which is usually referred to technically as the question of the stability of equilibrium. Stability is a concept with robbery wide ramifications. And it goes back in its name probably to the Greeks. Classic sort of thing is the. In fact where the word equilibrium originated is in an ordinary scale balance. When there are equal weights. Or equilibrium in Latin. For the two scales. Now the stability would mean in that context the following Supposing somebody shook the scale that'll

put his finger in one of the pans and then took it off again with the scale come back. Into its equilibrium position or would it go all the way down. Actually it might do either depending on the exact construction of the scale and example which took up much of the interest of 18th century mathematicians. And for which I am indebted to LEO which is that of shipbuilding. The stability of ships is not an uninteresting matter for a mariner. And. If you think about the construction particularly a sailing ship you can see how it might arise. Supposing a very tall and very heavy masts. The ship rides upon the water as long as the rider is still master upright and there's no problem but a sea with waves comes along these kind of Spawn to the disturbances of the economic system that you mentioned. The mast now tilts over somewhat.

If it is very heavy we may find that the center of gravity the ship has moved out from over the area in which the ship is resting on the water. At this point the ship has become unstable and will continue to turn over until it is lying on its side. Not a state of equilibrium indeed but perhaps not the place in the state that one would like to consider it. In effect you suggested there are ripples endless ripples which disturb the economic system from an equilibrium might achieve. A classical view of the economists particularly in the Anglo-American tradition has. Tended to assume that the economy is really a self regulating mechanism a stable mechanism. If we consider the single market and I mentioned before this seems reasonable enough.

Supposing the price is too low that is below the equilibrium price by which we mean the price at which supply and demand are equal. If the price is lower this probably means the producers are willing to supply less than they would at equilibrium while purchases want more than they would equilibrium. Then there will be an excess of demand over supply. There are people wanting the good. I wanted more of the good and the producers are willing to supply. I would merely common sense would suggest in those conditions that prices will rise and the price will rise in this market until it comes to equilibrium. At this point everybody is satisfied and there's no incentive for anybody to want to change the price. But this simple description overlooks the complex interdependent economy that I mentioned before. If we're talking about a better market run the price of butter is

changing. We are having side effects on the margarine market. The milk market the bread market and many others. These in turn will eventually through the chain of interdependence come back and affect the milk market and the better market and the margarine market. Is it so clear then that we will still have stability. This is a very difficult question. And I can only say that we are made in the study of mathematical models. Some tentative first steps. Suppose we consider a static economy. That is we ignore all the questions if you know the future that I raised before. Goods are not doable they break down immediately before an individual will only

worry about the prices that now exist when making decisions to not worry about his expectations. Remember this is an ideal economy not the real one. In this economy it is possible to show that under conditions which are at least vaguely plausible and less representative of real conditions but not exactly so. That the system will be stable. That. If you start with a set of prices which is not the equilibrium. They will gradually influence each other. Through the effects of supply and demand. In such a way that they will eventually come close to the equilibrium price. And if the process is allowed to go on long enough will eventually reach that the equilibrium price that each market separately. There may be a good deal of oscillation in the process but the end will be clear.

However I must make clear that there is a great deal more. That as not yet been covered and you may ask indeed right we should concern ourselves with a model economy such unreality. I can only answer that such approaches are the characteristics of all science. Science is always based at least to begin with on an unrealistic abstraction from the complex matrix of reality. Would be its. It is interesting to read works in the field like cosmetology. The. Distribution of matter in the universe and look at the models which are used there. They frequently start off with some assumption such as that matter is uniformly distributed throughout the universe. One has only to look around

when you see this is a false assumption. Yet it has been a very useful one because analysis would be impossible without simplification. There is a second reason for studying simple models of complicated realistic models are simply difficult to analyze even though we may have to analyze them eventually. We need practice. We have to build up from small steps to before we are ready to face the full realities of the world. So. We theorists the mathematical bent start with our deliberately unrealistic models hoping that we are catching some of the spirit of the real world and by a process of successively increasing complications.

Arrive at a better understanding. What complications are we likely to encounter as we go from the simple static model where the future doesn't matter to a. Real world of durable goods. Where as I remarked at the beginning the interview the expectations the future the hopes the aspirations the fears of the business man and to a certain extent of the consumer really matter. Well it's not hard to see that small fluctuations may breed larger consequences in that case. Psychologically and indeed rationally an individual will tend to base his expectations the future of good measure on the most recent past.

Supposing there has been for any one of these many accidental reasons a small downturn this will lower the individual's expectations for the future. He may suddenly decide that he has. As much capital goods on hand as he really will need. Given his now lower level of aspiration or lower state of expectations he will cease to buy then more machines will see stuff more buildings constructed. Now we find a certain amount of unemployment in the construction industry. And machine goods industry. These industries in turn becoming pessimistic lever their investment requirements. They also lay off workers. Lower purchasing power. The fear is by other industries now a still worse conditions. And we have the beginning of a cumulative chain of depression.

Professor Jacob Marshak at Yale University has conducted similar research by using game theory a layman's view of this theory might be a theory that pits one opponent against another or one group against another with each selecting a choice which would gain the most benefit for himself. I asked Professor Marsh on how he employs this theory in his research. You know realizing the games they will assume that each player tries to achieve the best for himself. That is either the definition of what he can to be here. The assumption that people who think rather complicated problems in such a way as to do the best for themselves and that is the base of the game theory applies. It's called going behavior but it has been applied to things like like air war tactics for example the duel between two airplanes has been analyzed from point if your theory of

games also major problems in strategy or even foreign policy have been submitted to such an analysis. It is in economics and I was just in this general sense that. The problem is to find the best possible decision. To illustrate the use of the theory Professor Marcia Clark suggested that he experiment by using one of the tests on me. This was the result of that brief experiment. Well here is an experiment which you are performing trying to essentially to measure what we call the executive capacity. Well it's one of many possible experiments you will put a man here a situation in which a course in business but which is simplified so that one knows what is the actually best solution and then the person is asked to provide his own solution. Suppose you have two or you have to sell.

Stock. In 5 days and you know only that any of these days a price can be anything between 20 and 30. Now let's game let's play it again would you like to play it. So you will be in though and then who. Who is. Who is selling a stock. Who knows the stock will be any day. Anything with equal probability. Anything between twenty and thirty dollars the last day on the fifth day will have to accept a new prize. And now I shall give you actual reading off of this book which is a table of millions of random numbers from the Rand Corporation with the help of an electron you can read that really simplify things instead of using being all we can just using these booklets go honest to goodness Roulette produced numbers

ranging from from 20 to 30. And you have five days now to sell this stock and let's see how you will behave. So you're on the first day the price happens to be 23. You know I don't think. Well second day price happens to be 25. You know 30 it price happens to be 27. I think I'd accept. Yes I think you are. You demonstrate here is the thing which psychologists school lowering levels of US British and all of the very first day were sitting high and presumably wouldn't accept anything I don't know. Perhaps not even 30 but not even 29 let's say. But then you got more and more modest.

Now maybe maybe creating a different way now you will finish your fourth and fifth days what would have been offered perhaps. Oh well 30 or 27. Years old the only force. Just 22 and this is one of the worst 27000 words offered both in the third and fourth. I say yes because it was pure event. OK so you don't regret the decision you have made. You don't know how this little experience best we can tell me a little more about your strategy in this way. I will ask you on the first day what is the limit that you would put on the price. And I shall tell you what the price was I think the first limit would have the thirty thousand thirteen. OK and there you get 20 too often so you don't buy. You don't sell you know you know in second day. On the second day I think I'd still ask for 30000. Twenty six.

OK. I don't accept. The third day. Do you think I would offer. I would ask for perhaps twenty seven thousand. Twenty five. So you don't accept. On the fourth day the fourth day. Twenty six thousand. Twenty eight. Well so it's good. So he does. You know the point is that there exists a perfectly rational mathematical answer to what you should have accepted. I think it's rather easy to realize that of course on the last day you have no choice but on the on the preceding day. You could say to yourself If I accept. Anything below 25

I don't fare so well because the next day I show on the Everest get 25 anybody anything between 20 and 30. So therefore in the force your limit should be 25 and by similar considerations one finds that only the third day your limit should be twenty six and a half and on the preceding day should be somewhat higher. 27 I think and a half and so on. And there are there are there is a different medical solution. Now of course in practical life one doesn't have time to proceed to section here magical solution and so you know of course well if it's too complicated. But perhaps people have people who have experience in judgment or ability for this particular kind of life and they really constant decision making under a certain team developing programming strategies in advance of time. They may have developed here closer to 30 years that is closer to the ideal one than other people. So here we are you may perhaps find a measure

of executive capacity. I mean not only is this particular experiment but Riess all sorts of other games that business executives or candidates for business executives might play to demonstrate that they have sufficient judgment and intuition to solve really complicated problems in a way that approaches its ideal solution. Our thanks to Professor Jacob Marshak of Yale University and Professor Kenneth Arrow of Stanford University for their participation on today's program on decision making in economics. Next week you will hear the Honorable David L. bass alone who is circuit judge of the United States Court of Appeals in Washington D.C. as he discusses some historic aspects of the law and mental health. On the next program from the series human behavior social and medical research consultant for today's program with Professor George good Tona of the University of Michigan. We extend our special

thanks to the Mental Health Research Institute of the University of Michigan. Plen Phillips speaking asking that you join us next week and thanking you for being with us at this time. This program has been produced by the University of Michigan broadcasting service under a grant in aid from the National Educational Television and Radio Center in cooperation with the National Association of educational broadcasters. This is the NEA E.B. Radio Network.