Psychology and economic fluctuations: Pigou, Mill and Keynes

[Note for bibliographic reference: Melberg, Hans O. (1998) Psychology and economic fluctuations: Pigou, Mill and Keynes, www.oocities.org/hmelberg/papers/981101.htm]

[Note: This paper is based on secondary sources and it is only a quick draft of my initial thoughts.]

Psychology and economic fluctuations
- Pigou, Mill and Keynes

by Hans O. Melberg

Introduction
According to A. Pigou, "the varying expectations of business men ... and nothing else, constitute the immediate cause and direct causes or antecedents of industrial fluctuations." (Collard 1983, 411). The purpose of this paper is to examine this view in more detail. I shall do so using the writings of three economists who all emphasized the importance of psychology and expectations to explain economic fluctuations: A.C. Pigou, J.S. Mill and J.M. Keynes.

Pigou
Pigou's argument is summarized as follows by David Collard:

"Fluctuations (irregular cycles) are driven by variations in the profit expectations of business people. These, in turn, are set off by 'impulses' which may be 'real', 'psychological' or 'monetary' in nature. Once the initiating impulse has made itself felt, it may be sustained by any or all of the three. Consequent fluctuations in employment will be greater the greater the degree of price or wage rigidity and are (because of externalities) probably larger than socially desirable" (Collard 1996, 913).

Modern business cycle cannot be accused of ignoring real or monetary impulses, but that psychological mechanisms causing expectations to change constitute an independent cause of fluctuations, is less accepted. To justify the inclusion of psychology, Pigou appeals to a number of mechanisms. For instance, Pigou argues that there is a "general tendency to expect processes that we observe in action now to continue in action at all events for some time to come" (Collard 1983, 412, originally TU 235). He also argues that there is an asymmetry in the expectations among borrowers and lenders. Finally, there are psychological interdependencies among business men which may create herd-behaviour. These expectational interdependencies, Pigou argues, are also multiplicative and interact in a way so that small changes can create large fluctuations in aggregate.

Exactly how do these mechanisms create cycles? If people believe that the current is a good guide to the future, many will invest in the industry when profit is high. This may lead to excess capacity, high production and low prices. This, in turn may lead to low investment, low output and high prices and so the circle continues. This is formally called a cob-web cycle and there is empirical evidence for this cycle for instance in the Norwegian ship-building industry.

Psychological interdependencies may also create cycles. One possible story may go as follows. Information is costly. One way of economizing on information, would be to imitate those you believe have reliable information. If a car-mechanic in your street buys a particular car, you may rationally infer that this is a good car even if you do not collect independent evidence on your own. If many people imitate, small changes in behaviour by the "leaders" may turn the market i.e. there is an inherent tendency towards large fluctuations (see Bikhchandani et al. 1998 for more on imitation and herd-behaviour). Of course, the psychological interdependencies need not be of the rational type as described above. One might, for instance, argue that optimism and pessimism is contagious. For instance, Keynes wrote that: "the market will by subject to waves of optimistic and pessimistic sentiment, which are unreasoning and yet in a sense legitimate where no solid basis exist for reasonable calculation" (Koppl 1991, 206. Originally in Keynes 1936, 154).

J. S. Mill
The key distinction in Mill's theory of economic fluctuations, according to Evelyn L. Forget, is between "rash speculators" and "professional gamblers" (Forget 1990). The rash speculators are attracted to a market when the price increases. They focus on price since they, unlike the professional dealers, do not know the underlying forces of demand and supply.

Any shock that leads to price increases is a potential impulse to speculation. There are, however, some structural reasons why the economy peridocally experiences more serious ups and downs. First, J. S. Mill views the opening of a new market as a shock that is particularly likely to create speculation and fluctuation. Second, calm periods by themselves create the conditions for a crisis. In calm periods professional dealers increasingly exploit the available profit opportunities. By so doing they accumulate capital, but the interest rate on capital investment decreases as profit opportunities are exploited. This tempts a number of professional dealers to become speculators (Forget 1990, 634). They have the money and they want to maintain the high flow of profit. The speculation must eventually go wrong (Mill argues) since the professional dealers will stop buying from the speculators (running down inventories instead) and the speculators will panic.

The panic send the price below the "true" price as dictated by normal supply and demand because both professional dealers and speculators are dependent upon bank credits. In a time of panic banks will deny both credits, although (in Mill's view) the professional dealers have a legitimate claim on credit. (One reason might be informational asymmetry: The banks do not know who the professional dealer is). (Forget 1990, 639)

How convincing is Mill's argument? In the real world, of course, we cannot divide people into only two categories (speculators or professionals). More seriously, I am unsure why calm periods should decrease the possibilities for highly profitable capital investments. The argument seems to assume some kind of limited supply of such investments. An alternative view might be that economic progress only increases these opportunities.

In addition to the view presented above, Mill relies on a number of other mechanisms to explain fluctuations. For instance, in the same way as Pigou (only Mill was first of course), he believes that people base their expectations on the current situation and that this creates fluctuations. Consider the following quotation:

"Every one calculating upon being before-hand with all his competitors, provides himself with as large a stock as he thinks that the market will take off; not reflecting that others, equally with himself, are engaged in adding to the supply; not calculating upon the fall of price which must take place as soon as this increased quantity is brought to the market. The deficiency is soon changed into excess." (Forget 1990, 630. Originally in Mill 1967a, 76.)

Another and related example, is Mill's concept of competitive investment. Overinvestment need not be the result of failure to take account of the view that others may be doing the same thing you are. Instead it might be the result of two different mechanisms. First, a person might invest even when he knows that other people are doing so, because he is sure of doing it better or faster than everybody else. Why should he be sure of this? Mill's answer is that there is a "universal propensity of mankind" ... "to overestimate the chances in their own favour" (Forget 1990, 635, Mill 1967a, 77). In addition the nature of the situation may be such that "... each dealer has an incentive to attempt to satisfy the new market before his competitors" (Forget 1990, 630, my emphasis). In some competitions/situations the reward for being second is not very different from being first. In some situations, however, the winner takes all. Winner takes all situations tend to induce too much investment (even rationally so).

Keynes
Keynes' starting point is that we cannot explain investments using theories of rational choice. The basic problems is simply that the future is uncertain. [This is, incidentially, a position he shared with Pigou who wrote that sometimes "there is no secure basis of past experience to act as a guide" - quoted in Collard 1983, 412. Originally in Pigou's Industrial Fluctuations, p.76.] Keynes then argues that the driving force behind investment is animal spirits. These are defined as "spontaneous urges to action rather than inaction" (Matthews 1984, 209. Originally in Keynes 1937, 161-162). Because of the lack of rational foundations for probability calculations, the investments are volatile.

The stock market is volatile for another reason as well. The is the classic beauty contest argument. Imagine a number of people all trying to guess who the winner of a beauty contest will be. The question is thus not who you think is the most beautiful. Nor is the question what you believe other people think about who is the most beautiful. You are trying to guess what the other people are guessing. Or as Keynes writes: "We devote our intelligence to anticipating what average opinion expects the average opinion to be" (Koppl 1991, 206. Originally Keynes 1936, 154).

R. O. C. Matthews believes Keynes confused the two arguments to some extent (Matthews 1984, 211), but I would question this. I do not think the first applies only to real capital formation and the second only to the stock market. It seems to me that even when deciding on real capital investments the kind of impossible cyclic strategic uncertainty that arises in the beauty contest may be a real problem. (I will not enter the market if I believe a number of others will do so as well, but whether they will depends on whether they think I will enter and so on). And the stock market may be volatile because of radical uncertainty and animal spirits.

A more serious criticism, however, is that Keynes does not really fill the concept of animal spirits with more precision. The mentioned Matthews, for instance, interprets animal spirits as a motivational variable (act!), as opposed to a cognitive (dealing with how beliefs are formed). Koppl, on the other hand, seems to put more emphasis on the cognitive aspect. Moreover, we should not automatically assume that a lack of rational basis for calculating future returns automatically leads to a volatile market. Radical uncertainty may lead to a reliance on convention that might produce stability. And, situations with known probabilities need not be stable if these probabilities change all the time so the "rational" solution also changes rapidly. Hence, the association between rationality and stability and radical uncertainty and volatility should be discussed in more detail.

Lessons
It is not surprising that the authors of the works used in this paper tend to argue that "Neglect of the psychological forces that I have been discussion is a lacuna in conventional economic theory" (Matthews 1984, 228; Collard 1996, 923). Is this really true? Are the old theories of economic fluctuation better than the new ones? In what sense are they better or worse?

First of all, what are the differences between these theories and modern theories of fluctuations? Collard sums it up in this way in his abstract

"... Pigou gave much more emphasis than does modern theory to: multiplicity of causes; psychology; amplitude; interaction among impulses; non-perfect competition; labour market institutions; externalities. Thus, his work may have some lessons for modern economists in terms of the scope and method of business cycle theory." (Collard 1983, 912)

The general questions I believe, is this: Should we approach the cycle within the framework of (general) equilibrium theory using only quantifiable variables or should we focus on mechanisms the way the authors above do?

Charles Plosser clearly believes it is wrong to focus on mechanisms in this way. He writes:

"... in order to understand business cycles, it is important and necessary to understand the characteristics of a perfectly working dynamic economic system." (Plosser 1989, 52)

"... it is logically impossible to attribute an important portion of fluctuations to market failure without an understanding of the sorts of fluctuations that would be observed in the absence of the hypothesized market failures" (Plosser 1989, 53)

In contrast to this, we have the rather ad hoc mechanisms of Mill and Pigou, and Keynes unspecified animal spirits. As Collard admits the theories described by these authors (at least the first two) and the interaction inherent in these: "lends itself to complex simulation rather than to elegant modeling" (Collard 1996, 923).

In sum, I believe my short investigation into old and psychological theories of economic fluctuations raises the question of how we should approach the issue: Should we use equilibrium models or simply try to find mechanisms we believe are important? This, in turn, raises the question of what we mean when we say that we "explain" and how we should go about finding the most reliable explanation in a situation when we are limited by many practical and theoretical problems. It is to these questions I turn next week. Until then I simply state that sometimes we are better off not striving for perfection when we operate in a situation of inherent limitations to our understanding (i.e. we may apply the theory of the second best).

References
Bikhchandani, Sushil and David Hirschleifer and Ivo welch (1998): Learning from the behaviour of others: Conformity, fads, and informational cascades, Journal of Economic Perspectives, 12 (no. 3), 151-170.
Collard, David A. (1983): Pigou on expectations and the cycle, The Economic Journal, 93, 411-414.
Collard, David A. (1996): Pigou and modern business cycle theory, The Economic Journal, 106, 912-924.
Forget, Evelyn L. (1990): John Stuart Mill's business cycle, History of Political Economy, 22 (no. 4), 629-642.
Koppl, Roger (1991): Retrospectives: Animal Spirits, Journal of Economic Perspectives, 5 (no. 3), 203-210.
Matthews, R. C. O. (1984): Animal spirits, Proceedings of the British Academy, 70, 209-229.
Plosser, Charles I. (1989): Understanding Real Business Cycles, Journal of Economic Perspectives, v. 3 (no. 3), 51-77.

Original Sources
Pigou's arguments are mainly from his book Industrial Fluctuations (Macmillan, London, 1927). Mill's arguments are originally from his Paper currency and commercial distress (1826) which is printed in The collected works of J.S. Mill edited by J.M. Robson (4:71-123, Toronto). Keynes' theory, can be found, of course, in his General Theory from 1936.

[Note for bibliographic reference: Melberg, Hans O. (1998) Psychology and economic fluctuations: Pigou, Mill and Keynes, www.oocities.org/hmelberg/papers/981101.htm]