2013년 1월 8일 화요일

[J.R. Hicks] The Hayek Story

출처: Hicks, J. R. 1967. "The Hayek story", in Critical Essays in Monetary Theory, Oxford, Clarendon Press. pp. 203 ~

* * *


When the definitive history of economic analysis during the 1930s comes to be written, a leading character in the drama (in was quite a drama) will be Professor Hayek. Hayek's economic writingsㅡI am not concerned with his later work in political theory and in sociologyㅡare almost unknown to the modern student; it is hardly remembered that there was a time when the new theories of Hayek were the principal rival of the new theories of Keynes. Which was right, Keynes or Hayek? There are many still living teachers of economics, and practical economists, who have passed through a time when they had to make up their minds on that question; and there are many of them (including the present writer) who took quite a time to make up their minds. How was it that happened? It is, at the least, an interesting historical question; and it may be more. If a theory (a rather complex theory) is to have even a temporary success, there must be something about it that rings a bell. That the immediate impact of the Hayek theory was extremely misleading is not now to be questioned; yet some of the issues to which it drew attention were real issues, issues that economists have found it hard to understand and which perhaps even now have not been completely cleared up.

It will be well to begin with some chronology. Hayek's Prices and Production was published in September 1931, but the lectures on which it was based had been given in London the previous February; Keynes's Treatise on Money was published in December 1930. It may thus be said that the two works came into the world at almost exactly same time; and it was a timeㅡjust when the full dimensions of the World Slump were declaring themselvesㅡat which the need for some new knowledge on the subject of Fluctuations was exceptionally high. Each of the authors had been working on the subject for years, but there had been hardly any contact between them. There is, indeed, in Keynes's book a curious passage[1: TOM, vol. 1, p. 199] in which he seeks support for his own innovations in the work being done, by Hayek and others, in Vienna; but the publications mentioned are all in German, and it is clear that Keynes had not studied them closely. For it was only too obvious, when the books were available for comparison, that Keynes and Hayek were not saying the same thing. The practical conclusions which they reached clashed far too violently.

What was in common between themㅡall that seemed to be in common between themㅡwas the intellectual descent which each claimed from Wicksell; but Wicksell plus Keynes said one thing, Wicksell plus Hayek said quite another. Which was right? It must be insisted, in defence of those who got so puzzled in 1931-33, that no more was yet available to them, on Keynes's side, than what was in the Treatise; stimulating ideas which, on their author's own admission, still needed a lot of sorting out. The pattern of Hayek's thought, at a first impression, looked more coherent. The obstacle which confronted one on his side was his Boehm-Bawerkian model; an analytical framework that had become familiar, even orthodox, in some continental countries, but was unfamiliar in England. Prices and Production was in English, but it was not English economics. It needed further translation before it could be properly assessed.

Several of us made attempts at that translation; the journal of the 1930s are full of them.[2] But what emerged when we tried to put the Hayek theory into our own words, was not Hayek. There was some inner mystery to which we failed to penetrate. We absorbed Hayekian ideas (there are Hayekian influences, not only on Robbins and Robertson, but even on Harrod and Kaldor, for instance, if one looks them) but there was something central that was missing. It is not so much that it was rejected; it slipped through our fingers. I hope, at long last, to be able to show in this paper what it was.
[2] Dennis Robertson's 'Industrial Fluctuation and the Natural Rate of Interest' (Economic Journal, 1934)ㅡsee above, p. 202ㅡis a particularly distinguished example.
Hayek himself endeavoured to show us; but I do not think he succeeded. It is indeed true that by the time he made his later efforts, and had got them into print,[1] his audience had dispersed. Keynes had spoken again, with greater clarity; to the opportunities that had been opened up by the General Theory what Hayek was saying appeared to have little relevance. It was perhaps again to have relevance, later on; but before that could happen, the 'Revolution' would have to proceed much further on its course.
[1] Profits, Interest and Investment(1939); Pure Theory of Capital(1941); 'The Ricardo Effect'(Economica, 1942, reprinted in Individualism and Economic Order, 1949).
Let us go back to the beginning, and try to see what it was that had happened. That means going back to Wicksell.


In Wicksell, the 'Cumulative Process' is a matter of prices. When the 'market rate' of interest is reduced below the natural rate, prices rise. Nothing is said about the movement of quantities (inputs and outputs). On the bearing of his construction on the causation of Trade Cycles, Wicksell is open-minded.[2]
[2] Lectures on Political Economy, vol.2, pp. 209-14.  It is interesting that there is the same  shift, from an interest in prices to an interest in quantities, between Treatise and the General Theory. The ^Treatise^ is a theory of price-levels; the latter is the ^General Theory of Employment^. No wonder that there was this shift in attention, in view of what was happening, contemporaneously, in the real world.
Hayek was asking the question; what happens to quantities in a Wicksellian cumulative process? He took his model very 'pure': much purer than Wicksell himself had been accustomed to take it. Prices(all prices) are perfectly flexible, adjusting instantaneously, or as nearly as matters. Price-expectations are not introduced explicitly, for in 1930 their day had not yet come. There must, however, have been some implicit assumption about expectations. We shall hardly go wrong if we take it to be the simplest possible assumption: the 'static' assumption. The same prices  as rule today (whatever they are) are expected to continue to rule in the future.

If these assumptions had been strictly maintained, what should have been said about the time-path of the Cumulative Process? When the market rate is reduced below the natural rate, what will happen to the ^quantities^ of inputs and outputs? The correct answer, on these assumptions, is very simple: the effect will be nil. Prices will rise uniformly; and that is that. When the Wicksell model is taken strictly (as it was being taken strictly), it is in ^Neutral equilibrium^. The whole ^real^ system, of quantities and of ^relative prices^, is completely determined by the supply and demand equations in the particular markets; in this ^real^ system ^the^ rate of interest is included. There can only be one rate of interest when the markets are in equilibrium; a market rate that is equal to the natural rate.  The 'reduction' of the market rate below the natural rate must therefore be interpreted as a disequilibrium phenomenon; a phenomenon that can only persist while the markets are out of equilibrium. As soon as equilibrium is restored, equality between market rate and natural rate must be restored. Thus there is no room for a prolonged discrepancy between market rate and natural rate if there is instantaneous adjustment of prices. Money prices will simply rise ^uniformly^; and that is that.

The point is, of course, the same as that which is made in Frisch's famous definition of ^dynamics^: consideration of 'the magnitude of certain variables at different points of time' and the introduction of 'equations which embrace at the same time several of these variables belonging to different instants'.[1] If a system has no lags, so that everythig is determined contemporaneously it cannot (endogenously) engender a process.
[1] 'Propagation Problem and Impulse Problems in Dynamic Economic', in ^Essays in Honour of Gustav Cassel^, p. 171.
Hayek's model does engender a process; some kind of lag (or lags) must therefore be implicit in it. Where is the lag to be found? One gets no help from expectational lags, such as were used b Lindahl;[2] by introducing a lag of price-expectations behind current prices, one can throw the Wicksell story into a determinate form, but it does not have much resemblance to the form that Hayek gave it. Nor is it consistent with Hayek's statements to suppose that the lag is attributable to rigid prices (for instance, wages): an interpretation which was congenial to Robertson,[3] and which would fit well enough into some versions of Keynsian theory. A reformulation of the Hayek theory on these lines does indeed have something to be said for it; we shall be returning to it later. But it is not what Hayek said. His lag (for he must have had a lag) is of a different kind.
[3] See his 'Hayekian' paper, above quoted.
It is not the production lag (of outputs behind inputs), the 'period of production' on which he has so much to say. That also is no help. What is needed is a lag in market adjustment, of prices behind the demands and supplies of commodities, or of the demands and supplies behind current prices. If there are no lag in market adjustment, the time-structure of production is irrelevant to the Cumulative Process; for there will not be time, before equilibrium is restored, for the structure of production to be changed. What then was Hayek's lag?

Consider what happens in the Hayek model. The initial effect of the expansion of creditㅡin 'pure' terms, the reduction of the market rate below the natural rate of interestㅡis that the money value of Investment rises, implying a rise in the money price of producers' goods. We are beginning (it is insisted)[1] from a situation in which there is full employment of labour. Labour is a producers' good; wages are flexible; so money wage must go up. But what happens about the spending of those wages? Everyone else, at that point in sequence, would say that here at least there must be an instantaneous, or nearly instantaneous, reaction. The higher wage must be followed, nearly at onceㅡperhaps in the next Robertsonian 'week'ㅡby a rise in the demand for consumption goods.[2] Then, in view of the full employment assumption, the prices of consumption goods must also rise. And there can be no equilibrium of supply and demand, over the whole system, until they have risen enough to withdraw the incentive for the rise in real investment. So (along that channel) we come back to the nil effect, on inputs and outputs, with which we began.
[1] ^Prices and Production^, pp. 34-35.
[2] There is no reason, under these assumptions, why there should be any fall in consumption out of profits.
But Hayek will not allow that. (1) In spite of the rise in wages the demand for consumption goods does not rise; so the prices of consumption goods do not, at this stage, rise. (2) This is how he is able to maintain that there is a rise in the prices of producers' goods, relatively to the prices of consumers' goods, lasting right through the 'boom': the rise on which so much of his argument depends. There has to be a lag of consumption behind wages, which must be large in comparison with any other lag that is admitted into the system. This is the lag (really, one would think, a most extraordinary lag) which gives Hayek model its peculiar slant.

Obviously this lag is not acceptable. Yet that is not the end of the Hayek theory. For if one can once get over this step, there is much to be learned from the next stage in the argument, where the effect of this change in relative prices on the 'structure of production' is most carefully worked out. Granted the initial change in the producer-price/consumer price ration, and ganted that ^it can be maintained^, the effect on the production process will be of the kind that Hayek describes. The queer thing about the Hayek theory is not the 'lengthening' and 'shortening' of the 'period of production' which attracted so much attention. It is not the answer that Hayek gives to his question; it is the lack of justification, within the model set out, for the question being asked at all.


It will accordingly be well, before proceeding further, to inquire into the possibility of 'mending' this first step in the Hayek argument: of finding a question, a different question, to which the answer that is given may be a sensible answer.

It is tempting, of course, to try a different lag. Suppose that one keeps the rest of Hayek's assumptions, but instead of the consumption lag, which is so implausible, one introduces a wage-lag: a lag of money wages behind the balance of supply and demand in the market for labour. If there is a lag of this sort (such as Keynes had already implied in the ^Treatise^) there can be a rise in producers' good prices relatively to consumers' good prices, such as Hayek requires. And there is room for the phenomenon to which Hayek attributed such importance: that the prices of some producers' goods will rise more than others ^as a result of the fall in the rate of interest^. (It must be the fall in the rate of interest which is responsible, for there is nothing else.) Along this route one can incorporate some part of Hayek's work on production structure; and this was an easy way by which a part (though much less than the whole) of Hayek's teaching could be absorbed. But the rigidity (even temporary rigidity) of ^money^ wages, on which it depends,[1] is a very un-Hayekian concept; if one pursues this line of thought, one is led toward a theory which is more like that of Keynes, or perhaps of Robertson, than of Hayek.
[1] If money wages are rigid, and there is a credit expansion, the prices of investment goods can still rise. In the production of those investment goods there will be abnormal profits (windfall profits as Keynes had called them). It i quite natural to take it that these windfall profit (or most of them) will be saved; so that a failure of consumption to expand, at this stage, becomes on this assumption quite intelligible. Consumption out of wages fails to expand for one reason; consumption out of profit for another.
There is another alternative: at the expense of a more drastic reinterpretation, it enables us to follow in Hayek's footsteps much more exactly.

Suppose that one had ^not^ started with a 'credit expansion' but had begun with genuine saving, a genuine increase in the propensity to save. The effect of this ^in equilibrium^ (all I thin would now agree) would be a fall in what Wicksell called the natural rate of interest. (Others may prefer other names, but we need not quarrel about that!) Suppose that the market rate of interest is reduced to match, so that demand-supply equilibrium can in fact be maintained. What happens to money prices is quite indeterminate, for (with flexible prices) the equilibrium remains a ^neutral equilibrium^ in Wicksell's sense; what happens to relative prices , and to quantities, should, however, be determinate. Investment increases (in value terms) relatively to consumption; but since there is, and was, full employment of labour, it is only by 'capital deepening', by 'lengthening of the period of production', that a real shift from consumption to investment can occur. All this just as in Hayek; the fall in the rate of interest is matched by a rise in real wages; whether we like to make the rate of interest (or profit), or the rate of real wages, the king-pin of the argument, is a matter of taste. Whichever we choose, it comes to the same thing.

If we adopt this latter interpretation, we can follow Hayek very closely. We can agree that he was doing somethingㅡif it was not what he thought he was doing. The Hayek theory is not a theory of the credit cycle, the ^Konjunktur^, which need not work in the way that he describes, nor is it, in fact, at all likely to do so. It is an analysisㅡa very interesting analysisㅡof the adjustment o an economy to changes in the rate of genuine saving. In that direction it does make a real contribution. But it is a contribution which, when it was made, was out of due time. It does not belong to the theory of fluctuations, which was the centre of economists' attention in 1930; it is a fore-runner of the growth theory of more recent years. In that application we can still make something of it.


Let us try to follow the further stages of the argument, from this point of view. Let there be an increase in the propensity to save; in order to simplify the argument (but it is not an essential simplification) suppose that it takes the form of a diminished propensity to consume ^out of profits^. Let there be a corresponding reduction in the market rate of interest, which (again for simplification) may be supposed to take place so speedily that the price-level of consumption goods can remain unchanged. The fall in the rate of interest will raise the marginal productivity of labour; real wages will therefore rise. But the rise in labour productivity will be greater in those lines of production which are more capital-intensive than in those that are less; in order to restore equality, there must be a shift of labour into more capital-intensive production.That is to say, there will be shift of labour from the production of consumption goods to that of investment goods; but especially into those forms of investment goods production which contribute to processes that, taken as a whole, are more capital-intensive. Let us just say, in order to visualise what is happening, that labour is shifted into the construction of new steel-work and new power-stations. That is just for the picture; the shift can, of course, be expressed in much more complicated forms.

It is an essential (and surely valid) part of Hayek's argument that this shift need not have an immediate effect in diminishing the supply of consumption goods. The current supply is largely a matter of work that has been done in the past; and that has not been changed. If the prices of consumption good are to be kept from rising (as we are supposing) labour must not be taken away from the lat stages of consumption goods production; but there will still be opportunities for shift, by taking labour from earlier stages. Thus what happens, at this point, is that the goods which would have been bought by the receivers of profit (or the equivalents of such goods, for there is no reason why there should not be some redistribution of actual articles) are transferred to provide the increased real wages of labour. And (so far) that is all.

The point must nevertheless come (so Hayek maintains) when the withdrawal of labour from consumption goods production must diminish the supply of consumption goods. A further increase in the saving propensity will then be necessary if the 'boom' (as he would call it) is to be maintained. If this increase is not forthcoming, there will be a 'crisis'.

When the story is interpreted in this manner, though it may not have much relevance to old-style trade crisis (which are surely to be interpreted as disequilibrium phenomena, complicated by price- and wage-rigidities and disequilibrium rates of interest, not here allowed for), it does seem possible that it may have relevance in other ways. (...)


댓글 쓰기