[발췌: Lawlor & Horn's] Notes on the Sraffa-Hayek exchange

출처: “Notes on the Sraffa-Hayek exchange,” Review of Political Economy, 4.3(1992), pp 317-340
지은이: Michael Syron Lawlor & Bobbie L. Horn

※ This is a reading note of this reader with some annotations and remarks added in his trying to understand the article. Please reach to the referred journal to see the original.

※ 최초 메모: 2013.1.4. 관련 주제의 다른 글도 이 시기에 근접할 것임. 관련 자료 목록:

• [Reading Hayek's] Reflections on the Pure Theory of Money of Mr. J.M. Keynes
• [Sraffa-Hayek: 1/3][Reading Sraffa's] Dr. Hayek on Money and Capital : Hayek (2nd of 3); Sraffa (3rd of 3) 
• [발췌] Hayek’s Early Cycle Theory Re-examined 
• [Excerpt of R. Skidelsky] Hayek versus Keynes: The Road to Reconciliation
• [J.R. Hicks] The Hayek Story 
• [발췌] Sraffa’s Reception of the German Economics Literature: A Few Examples
• [책] Piero Sraffa: Critical Assessments
• Some readings on Piero Sraffa (and F. A. Hayek)
• [자료] Hayek Translated: Some Words of Caution

※발췌 (Excerpt):

* * * * *

(... ...)

I Introduction

At issue is the meaning and significance of Piero Sraffa's 1932 review of F.A. Hayek's Prices and Production, and the ensuing exchange (Sraffa, 1932; Hayek, 1931; 1932).

The exchange has recently attracted attention from writers with historical inclination.[1] Commentary has exhibited a wide range of interpretation or evaluations. According to which of the authors one reads, the Sraffa-Hayek debate may represent any of the following:

1. A difficult to interpret and ambiguous way-station between Sraffa's early work on the distributional aspects of monetary policy, and his later full-blown critique of orthodox economic theory (Panico, 1988).
2. The analytical basis of Keynes['s] most elaborate ^General Theory^ analysis of the 'essential properties of interest and money' (Deleplace, 1986; Majewsk, 1988; Kregel, 1982; Mongiovi, 1990; Potetio, 1986; Rymes, 1978).
3. The opening shots by Sraffa of a planned counter-revolution against subjectivism which culminated 30 years later with the ^Production of Commodities by Means of Commodities^ (Lachmann, 1986; Caldwell, 1986).
4. An early discussion of the true problems associate with the attempt to integrate money into a Walarsian general equilibrium model, and the forerunner of the modern mathematical treatment of this issue (Desai, 1982; McCloughry, 1982). 

This is an impressive list of ideas all to be contained in a 12 page review, a 13 page reply and a 3 page rejoinder. (...)

(...) Our approach is to read out of the exchange the discussion of the issues actually treated there alone, taking the authors (and the words they wrote) at face value. When our more narrow vantage point is employed, the contours of the exchange take on a clearer meaning, and one that believes the collective view of confusion (and even subterfuge)[3] noted above.

Some important matters at issue will be the following:

• First, to clarify the historical record of this crucial period in the development of economic theory; 
• secondly, to highlight a critique of the traditional distinction between 'natural' and 'money' rate of interest, and the view of the place of money in the economic system implicit in this distinction, that has gone virtually unnoticed in the monetary literature since; 
• thirdly, to show that Sraffa provides a crippling objection to the 'forced saving' doctrine which constituted the theoretical core of Hayek's Prices and Production. In view of some recent calls to 'go back to Hayek' in formulating a theory of the business cycle, this specific objection to Hayek's own formulation is most relevant.
• Finally, we want to end by noting that the exchange sheds new light on some interesting issues in the scholarship on Sraffa. At one level it offers another example of the unmerciful logic and devastating critical style of the man from whom Keynes claimed 'nothing was hid'(Keynes, 1972: 97). But though Sraffa is arguably a perfect example of Buffon's dictum 'le style est l'homme meme', the exchange also offers the substantive yield of some rare glimpses into Sraffa's view on monetary theory.

II Setting the stage

Prices and Production(1931)[4] consists of a series of four lectures Hayek presented at the London School of Economics in 1930. They were intended to outline the "Austrian" approach to the explanation of business cycles originating in the work of Mises and extended by his student Hayek. (... ...)

[4] We will quote from the first edtion of Prices and Production, published in 1931ㅡthe one Sraffa reviewed, of course. The second, revised and enlarged, edition, which appeared in 1935 is now more commonly used. A side-by-side perusal of these two editions reveals no substantive revision beyond the addition of a few extra footnotes. The 'enlargement' consisted of a shot note after essay IV on the practical difficulties of implementing a neutral monetary policy, and a reply to a critique by Hansen and Trout(1933). It would seem that Hayek did not find Sraffa's criticism important enough to change his argument when the opportunity arose; although see Hayek(1941: 35, n. 1) in this context.

Hayek's approach was an almost eclectic blend of the Austrian capital theory of Boehm-Bawerk, the short-run quantity theory effects of the early British monetary theoriests and the relative-price and interest rate theories of Wicksell. Hayek argued that any legitimate monetary theory would turn upon relative price effects (pp.22-28). This derived from his position that the notion of a general price level is theoretically unsound[5] and his corresponding view that the notion of a general value of money was superfluous.

And, indeed, I am of the opinion that, in the near future, monetary theory will not only reject the explanation in terms of a direct relation between money and the price level, but will even throw a overboard the concept of a general price level and substitute for it investigations into the causes of the changes of relative prices and their effects on production (Hayek, 1931: 25-26).

Hayek's explanation of business cycles centered on the supposed effects of monetary policy on the relative prices of capital goods and consumption goods. Here ensued the centrality of Mises’ forced saving theory and the Wicksellian natural-rate doctrine to his argument.  In Hayek's scheme, [:] 

• (1) the natural rate was the rate which equalized ‘the supply and demand for real capital’. 
• (2) If the bank rate (or actual rate) differed from this, the result was a divergence of ‘voluntary’ saving and investment. It was a lowering of the interest rateㅡconceived as an intertemporal price ratio between current(consumption) production and future(capitalistic) productionㅡthat signalled the investors to invest more than was ‘natural’, and thus led to the creation of ‘forced savings’. 
• (3) Crises and cycles resulted from what Hayek considered the inevitable destruction of capital that followed such a misdirection of production by bank-influenced relative prices. [※ 저리 신용공급에 따른 상대가격 변화→생산 왜곡 →자본 파괴]

Hayek (following Wicksell) focused on the natural rate('equilibrium rate') as the rate which would equalize investment and voluntary saving in a ‘nonmonetary’ barter economy. The only influence of money in this scheme came purely through its role as the medium of exchange. Yet, since saving and investment actually took the form of money transactions, the banking system (including central banks) had the ability to induce actual saving and investment to diverge from their ‘natural’ proportions. This occurred whenever the money rate was set above or below the natural rate. (cf) In what Hayek saw as the normal course of the boom to the too low money rate misdirects production towards a more capital intensive mix (in Austrian parlance a ‘lengthening of the period of production’).

※ (cf): 영문상 다음 셋 중의 하나를 의도한 문장인 듯:
1. In what Hayek saw as the normal course of the boom to the too low money rate production is misdirected towards a more capital intensive mix. [아니면 misdirect를 자동사로 간주하고 주어(production)과 동사(misdirects)가 도치]
2. In what Hayek saw as the normal course of the boom[,][to] the too low money rate misdirects production towards a more capital intensive mix.
3. [In] What Hayek saw as the normal course of the boom to the too low money rate misdirects production towards a more capital intensive mix.

Hayek saw himself as differing with Wicksell, though, over the role which the 'natural rate' could serve as a guide to policy in an expanding economy. As a then contemporary student of these rate theories{=Hayek} put it, there were[:]

... four tests which Wicksell gave for his ‘natural’ rate: viz., (i) stabilization of the general price-level; (ii) equalization of current savings and investment; (iii) the identity of the 'natural' rate with the non-monetary barter rate; and (iv) its identity, again, with the prospective yield on future real capital (Adarkar, 1935: 30-31).

Of these four results of an actual rate being set equal to the natural rate ([cf] for Hayek by a ‘neutral’ money policy of maintaining the money supply constant), Hayek found a contradiction between (i) and the rest. Hayek argued that while setting the natural rate equal to the money rate would give expression to the voluntary decisions captured in (ii) through (iv) it could not at the same time keep the price level constant in an expanding economy:

Nevertheless, it is perfectly clear (a) that, in order that the supply and demand for real capital should be equalized, the banks must not lend more or less than has been deposited with them as savings. ... And this means naturally (a') that ... they must never allow the effective amount of money in circulation to change. At the same time, it is no less clear (b) that, in order that the price level may remain unchanged, the amount of money in circulation must change as the volume of production increases or decreases. (c) The banks could either keep the demand for real capital within the limits set by the supply of savings, or keep the price level steady; but they cannot perform both functions at once (Hayek, 1931: 23-24).

※ [cf] 실제금리를 자연금리에 맞추는 것은 하이에크에게서 통화량을 일정하게 유지하는 ‘중립적’ 통화정책으로 달성된다. 

※1: (a)=(a'). (a) 실물자본의 공급과 수요가 일치하려면 은행은 예치된 저축만큼만 대출해야 한다. (a') 저축된 돈만큼 시중 통화량이 줄어들므로 은행이 정확히 저축액만큼 대출하면 당연히 시중 통화량에는 변화가 없어야 한다. 

※2: (b) 또한 생산량이 증가하거나 감소하는 와중에 물가가 변하지 않으려면 [생산량 변화에 맞춰] 시중 통화량이 변해야 한다. (c) 따라서 은행이 저축액 한도 내에서 실물자본 수요(즉, 돈을 빌려서 실물자본을 구매하려는 투자 수요)를 유지하면ㅡ통화량이 불변이므로ㅡ 물가수준의 불변을 유지할 수 없고, 물가수준의 불변을 유지하려면ㅡ통화량이 생산량 변동에 맞춰 변해야 하므로ㅡ통화량의 불변을 유지할 수 없다.

Thus, Hayek stakes out his position. He views the goal of a ‘neutral’ monetary policy (i.e., one that allows for the full expression of voluntay decisions over saving and investment) as a proper one. Giving full vent to individual decisions requires maintaining the money-rate equal to the natural rate. And since this policy is incompatible with attention to the general price level, Hayek believes we must abandon any such fuzzy aggregative concepts as the price level and focus our attention on relative price effects of a monetary policy:

But it seems obvious as soon as one once begin to think about it that almost any change in the amount of money, whether it does influence the price level or not, must always influence relative prices. And as there can be no doubt that it is relative prices which determine the amount and direction of production, almost any changes in the amount of money must necessarily also influence production (Hayek, 1931: 24).

This is the argument that so fascinated economists in the 1930s. Like many theoretical developments it contains, in equal parts, (1) methodological prescription and (2) positive analysis.

• (1) Hayek insisted that a valid monetary theory must proceed from the ‘method of neutral money’. 
• (2) This involves comparing (a) a hypothetical barter price system in equilibrium, with (b) the disturbances to that system that result from changes in monetary policy.

Sraffa clearly saw the dual nature of Hayek's thesis, as his review analyses both Hayek's chosen framework and his use of that framework. Thus both an internal and an external critique of Hayek's argument is presented by Sraffa.[6]

[6] There is a case to be made that this is Sraffa's basic mode of analysisㅡit is something very similar to both Marx's method of dialectical critique, and to the ‘immanent’ criticism carried on by the Swedes. Another instance where Sraffa employes this method is his famous critique of the Marshallian supply curve. In a fascinating unpublished Ph.D thesis(Coates, 1989) John Coates makes the argument that this method of analysis was at the heart of Sraffa's influence on Wittgenstein.

III 'Assuming away the object'

In Sraffa's view, though, it is not possible to consider these two elements in isolation. The basic flaw of the theory originates with the method. The subtlety of the review is the manner in which Sraffa weaves the internal contradictions of the theory into the basis of the external critique. It is useful to begin with this external criticism.

The external critique in Sraffa's review addresses the important issue of the place of money in a theoretical representation of the economy. His complaint is that while Hayek professes to be analysing the influence of money on relative prices, he is in fact concerned with a wholly different question.

... the reader soon realizes that Dr. Hayek completely forgets to deal with the task he set himself, and that he is only concerned with the wholly different problem of proving that only one particular banking policy (that which maintatins constant under all circumstances the quantity of money multiplied by its velocity of circulation) succeeds in giving full affect to the ‘voluntary decisions of individuals’, especially in regard to saving, whilst under any other policy these decisions are ‘distorted’ by the ‘artificial’ interference of banks (Sraffa, 1932: 45)

By Sraffa's account the reason Hayek strays from the path he sets out for himself can be traced to the theoretical method he employs;[:] the technique of ‘neutral money’.

The starting point and the object of Dr. Hayek's inquiry is what he calls ‘neutral money’; that is to say, a kind of money which leaves production and the relative prices of goods, including the rate of interest, ‘undisturbed’, exactly as they would be if there were no money at all. (Sraffa, 1932: 42)

The flaw of this method lies in using a framework which is exactly as if ‘there were no money at all’ as the guide to his inquiry into monetary theory and policy. Sraffa notes that the neutral economy might be useful if it were to be employed by way of comparison with various real monetary economies. It then might be possible to compare the disturbances to the monetary economies with those to the monetary ‘neutral’ economy. ‘This would bring out which are the essential characteristics common to every kind of money, as well as their differences, and thus supply the elements for the merits of alternative policies’(Sraffa 1932: 43).

Instead, Hayek conducts his investigations completely within the context of the neutral money, which by his own admission conceives money only as a medium of exchange. By doing so, Sraffa recognizes that Hayek must ignore those very characteristic that define a monetary economy:

The differences between a monetary and a non-monetary economy can only be found in those characteristic which are set out at the beginning of every textbook on money. That is to say, that money is not only the medium of exchange, but also a store of value and the standard in terms of debts, and all other legal obligations, habits, opinions, conventions, in short all kinds of relations between men, are more or less rigidly fixed (Sraffa, 1932: 43).

Since Hayek's starting point of a hypothetical ‘neutral’ barter economy considers only money which is ‘used purely and simply as a medium of exchange’, his inquiry cannot admit the ‘most obvious’ effects that a monetary policy will have in a real money economy, where ‘when the price of one or more commodities changes, these relations (described above) change in terms of such commodities; while if they had been fixed in commodities, in some specified way, they would have changed differently or not at all’ (Sraffa, 1932: 43-44).

In sum Hayek's method of neutral money ‘amounts to assuming away the very object of the inquiry’.

Having said this, Sraffa as a reviewer is faced with a dilemma. For if the whole method of the work under review undermines the investigations from the start, what more is the reviewer to say about the author's use of this method?

• Given that Hayek's method of analysing the influence of monetary disturbances on saving, accumulation and relative prices emasculates the role of money from the start, then some other factor besides the monetary disturbance must explain his results. 
• As Kaldor reminds us above, Hayek seemed to be proposing that the influence of monetary policy on these matters was, in fact, the whole of the explanation of the great enigma of the 1930s; industrial depressions. [이 문장이 삽입된 문맥상의 이유?]
• But if this conclusion is based on the use of a system that is ‘as if there were no money at all’, an alternative explanation is needed. Finding this explanation is the task of the internal critique of Sraffa's review:

Dr. Hayek invariably finds, when he comes to compare the effects of alternative policies in regulating this emasculated money, (1)that there is an all-important difference in the result, and (2)that it is ‘neutral’ only if it is kept constant in quantity, whilst if the quantity changed, the most disastrous effects follow. The reader is forced to conclude that these alleged differences can only arise, either from an error of reasonng, or from the unwitting introduction in working out the effect of one of the two systems compared, of some irrelevant non-monetary consideration, which produces the difference, attributed to the properties of the system itself. The task of the critic, therefore, is the somewhat monotonous one of discovering, for each step of Dr. Hayek's parallel analysis, which is the error or irrelevancy which causes the difference (Sraffa, 1932: 44-45).

IV. 'Incantations and a little poison'

In terms of Hayek's ‘internal’ argument sketched above, Sraffa addresses himself to the two cardinal points: forced savings and the natural rates of interest. Both involve Hayek's basic conception of the influence of changes in the quantity of money in the economic system.

In the first case Sraffa is interested in clarifying the distinction, crucial to Hayek's theory, between ‘voluntary’ savings and ‘forced’ savings. Sraffa recognizes that the relevance of this distinction rests on the alleged permanence of the capital accumulated in the voluntary case as opposed to the ‘inevitable’ destruction of that accumulated in the forced saving case. In Hayek's argument,[:]

• monetary expansion may increase saving and accumulation for a time, but when the quantity of money ceases to expand any such involuntary financed capital will be dissipated as consumers revert to their former proportions of intertemporal money expenditure. 
• This ‘spoilage’ of capital is what Hayek describes as an economic crisis. 

What is crucial to this argument is some mechanism for ensuring that the initial (those prior to the monetary expansion) proportions of money income saved and consumed will always be re-established when the expansion comes to an end.

... the use of a larger proportion of the original means of production for the manufacture of intermediate products can only be brought about by (1) a retrenchment of consumption. But now this sacrifice is not made by those who will reap the benefit from the new investments. It is made by (1)' consumers in general who, because of the increased competition from the entrepreneurs who have received the additional money, are forced to forego part of what they used to consume. It comes about not because they want to consume less, but because they get less goods for their money income. There can be no doubt that, (2) if their money receipts should rise again, they would immediately attempt to expand consumption to the usual proportion ... then at once the money stream will be redistributed between consumptive and productive uses according the wishes of the individual concerned, and (3) the artificial distribution, due to the injection of the new money, will, partly at any rate, be reversed. If we assume that the old proportions are adhered to, then the structure of production too will have to return to the old proportion (Hayek, 1931: 53)

There are numerous problematic aspects of this argument that Sraffa notes in passing.[7 각주: The problems concern the following issues:]

1. Hayek's definitions of income, net income and saving.
2. His neglect of the issue recently dubbed 'the time to build' (Kydland and Prescott, 1982). This concerns the fact that it is crucial in determining the effect of forced saving on capital accumulation to know whether the monetary expansion lasts long enough to complete a project and have it start 'earning its keep'. Time to build is particularly important in discussing the 'permanence' and 'destruction' of capital.
3. Hayek's view that it is the barter-like natural state that would give full expression to 'the voluntary decisions of individuals'.
4. A cavalier treatment of the relationship between the velocity of circulation and changes in the supply of money.
5. The most obvious objection, though, when viewd from the post-General Theory vantage point, is that Hayek's argument assumes full employment throughout.

Some were addressed in Hayek's later evolution and adjustments of his theory in the 1930s and 1940s. Though these technical aspects are interesting, and in fact offer further opportunities to contrast the outlooks of Sraffa and Hayek, they are not the central elements of the argument at hand.

For the sake of an internal critique Sraffa provisionally accepts the groundwork established by Hayek and asks a very simple question. What is to ensure that accumulations of capital that results from the forced savings will ultimately be consumed? As Sraffa sees it, there is no reason to differentiate between accumulations that results from forced or voluntary savings. For Sraffa, once income has been redistributed by the inflation and saved in the form of capital assets, there is nothing to distinguish those assets from ‘voluntarily’ accumulated assets.

As a moment's reflection will show, 'there can be no doubt' that nothing of the sort will happen. (a) One class has, for a time, robbed another class of part of their incomes; and has saved the plunder. (b) When the robbery comes to an end{즉, 통화팽창과 물가승승이 멈추면}, it is clear that the victims cannot possibly consume the capital which is now well out of their reach. (b1) If they are wage-earners, who have all the time consumed every penny of their income, they have no wherewithal to expand consumption. And (b2) if they are capitalists who have not shared in the plunder, they may indeed be induced to consume now a part of their capital by the fall in the rate of interest; but not more so than if the rate had been lowered by the 'voluntary savings' of other people (Sraffa, 1932: 48).

For the purposes of Hayek's argument that inflation induced accumulation produces the seeds of the ensuing crisis[,] the crucial element is the ‘inevitable destruction’ of the ‘forced’ accumulation. Sraffa is not persuaded it must be dissipated just because of its source and claims ‘Dr. Hayek fails to prove the contrary’(p. 47). In his reply, Hayek admits the centrality of this argument to his thesis (‘My theory stands or falls on this point’), thus it is worthwhile to consider this further defence and adjustment of it.

Sraffa, in effect, cut to the heart of the ‘great mistery’ (Hicks, 1967: 205) of Prices and Production. It is the forced lengthening and inevitable reversion of the period of production that constitutes the action in Hayek's case for monetary induced cycles. The trouble for commentators, then and since, has been to account for the mechanism of Hayek's 'concertina effect'. Given that change in the quantity of money seem neutralized from the start by Hayek's chosen methodological framework, Sraffa is requred to conduct a 'parallel analysis' of the operation of the system. In the case of forced savings the problem is to explain why the ‘involuntary’ nature of the accumulation requires that it be dissipated when the monetary expansion comes to an end.

The set piece of Hayek's analysis is his critical ‘proportion’ between consumer and producer goods that constitutes a Boehm-Bawerkian view of the quantity of capital as some average period of production.

• It is noteworthy that Hayek defines this proportion in terms of ‘the money stream’ or ‘total expenditure’ directed to the two categories. Thus the influence of monetary expansion on the ‘structure of production’ comes through altering the proportions between these critical money streams. 
• Under conditions of forced savings this change in proportions is brought about by additional credits granted to producers by the banks. When this lending slackens[,] the question is what causes the capital accumulated to be destroyed?

If the crisis is a ‘reversion’ to the previous ‘proportions’ the only way that it can come about is by a redirection of part of the spending stream from producers (who were ‘robbing’ consumers of part of their natural proportion during the inflation) back to consumers. Hayek asserts that this inevitably comes through a rise in money receipts of consumers once the expenditure of the producer's credits work back through the system. This is where Sraffa internally evaluates the argument and connects it to the external critique already presented.

Sraffa's argument is as follows.[:]

If the essential element of the story is the change in proportions of the spending stream, it cannot be monetary expansions themselves that account for the crisis. The reasoning is driven home in the exchange by both Sraffa's and Hayek's comments on an alternative case of monetary expansion to the forced savings scenario.

• As a point of explication, Sraffa notes in passing that by Hayek's reasoning ‘if the banks increased the circulation but apportioned the additional money between consumers' and producers' credits so as not to disturb the initial "proportions", nothing would happen’ (Sraffa, 1932: 48).
• In his reply, Hayek confirms this view and insinuates that Sraffa got it from the German edition of Prices and Production, where it had been stated that a monetary expansion could only avoid misdirecting production ‘if it were possible to inject the additional quantities of money, ... into the economic system in such a way that no change in the proportion between the demand for consumers' goods and the demand for producers' goods would be brought about’ (Hayek, 1932: 244-45).

The key point for Sraffa is that something other than monetary effects are in fact responsible for Hayek's conclusions.

• If a monetary expansion that balances increased credit to producers and consumers in the ‘natural proportion’ does not induce crises, then a monetary expansion is not the sine qua non of the cycle. 
• Sraffa's parallel analysis concludes that, instead, it is the ‘supposed power of the banks’ to misdirect these credits that does all the work in Hayek's theory.

What has happened is simply that, since money has been thoroughly ‘neutralised’ from the start, whether its quantity rises, falls, or is kept steady, makes not the slightest difference; at the same time an extraneous element, in the shape of the supposed power of the banks to settle the way in which money is spent, has crept into the argument and has done all the work. As Voltaire says, you can kill a flock of sheep by incantations, plus a little poison (Sraffa, 1932: 49).

V 'Essential confusion'

Having shown that the consequences Hayek claims for forced savings cannot be explained within the context of his theory, Sraffa is led to consider the second cardinal point, the distinction between natural and money rates of interest. In Hayek's argument, forced savings results from the monetary policy of lowering the money rate below the natural rate. As Sraffa note, Hayek's use of this distinction is 'mainly given by way of criticism and development of the theory of Wicksell'. Quoting from Prices and Production, Sraffa indentifies the main theme of Hayek's theory of the relation of money to the rate of interest:

In a money economy, the actual or money rate of interest may differ from the equilibrium or natural rate, because the demand for and the supply of capital do not meet in their natural form but in the form of money, the quantity of which available for capital purposes may be arbitrarily changed by the banks (Hayek, 1931: 20-21; quoted in Sraffa, 1932: 49)

In keeping with his emphasis on the ‘method of neutral money’, Hayek sets up the natural rate as a benchmark by which to measure the neutrality of monetary policy.

• (1) A policy is neutral if it sets the money rate equal to the natural rate. 
• (2) His contention is that such a neutral monetary policy, while it would leave the natural system undisturbed(‘as if there were no money at all’), could not be counted on to simultaneously stabilize an aggregate price level if the economy were expanding. 
• (3) Thus, the appropriate guide to policy becomes the maintenance of a stable quantity of money and not the maintenance of a stable price level.

Sraffa's critique of this part of Hayek's argument is that it represents a confused conception of the relationship between money and prices (or alternatively between a barter and a monetary economy:

An essential confusion, which appears clearly from this statement, is the belief that the divergence of rates is a characteristic of a money economy: and the confusion is implied in the very terminology adopted, which identifies the 'actual' with the 'money' rate, and the 'equilibrium' with the 'natural' rate. (a) If money did not exist, and loans were made in terms of all sorts of commodities, there would be a single rate which satisfies the conditions of equilibrium, (b) but there might be at any one moment as many 'natural' rates of interest as there are commodities, though they would not be equilibrium rates. The 'arbitrary' action of the banks is by no means a necessary condition for the divergence; if loans were made in wheat and farmers (or for that matter the weather) 'arbitrarily changed' the quantity of wheat produced, the actual rate of interest on loans in terms of wheat could diverge from the rate on other commodities and there would be no single equilibrium rate (Sraffa, 1932: 45)

In order to further illustrate the 'essential confusion' involved here, Sraffa again 'internalizes' the argument. Taking at face value Hayek's argument that the 'natural rate' is indeed the rate that would obtain in a barter state, Sraffa asks a not so obvious question. What would loans and interest rates look like if money did not exist, and how would they be different from rates of interet in a money economy? His answer: if we are really were in a barter state, the only meaning that loans, savings, or investment could have would be defined in physical terms, money being nonexistent by definition. Thus, 'natural' rates of interest would be rates defined in 'real' or physical terms. The ratio between the amount of a physical commodity today to the amounts it trades for at some future date would be the physical analogue to rates of interest. 'In order to realize this, we need not stretch our imagination and think of an organized loan market amongst savages bartering deer for beavers. Loans are currently made in the present world in terms of everyday commodity for which there is a forward market' (Sraffa, 1932: 49-50).

Armed with this concret conception.[8] Sraffa conducts a 'parallel analysis' of Hayek's scheme. In such a barter world, equilibrium means that 'the spot and forward price coincide, ... and all the "natural" or commodity rates are equal to one another, and to the money rate'. But, if the supply and demand get out of long-period equilibrium for any reason, the spot and forward prices diverge, and the 'natural' rate of interest on that commodity diverges from the 'natural' rates on other commodities. In other words, 'natural' rates are not necessarily 'equilibrium' rates if by equilibrium it is meant prices equal cost of production[9] and by 'natural' we mean barter-like (money-less) intertemporal loans. Consequently, Hayek's attempt to equate the equilibrium rate with the putative natural rate represents an 'essential confusion'.

This is most easily grasped in terms of Hayek's own preferred situation of an expanding economy. If, following Hayek's language, this means an increase in output of the producer goods industries, the only way to attract the new resources to this production is for the expected rate of return in the production of these goods to increase. Recall that Hayek explicitly wanted to address himself to an investigation of the effect of monetary influences on relative prices. He also claimed to have rejected a reliance on 'vague' concepts of averages, such as the price level. He was thus led to found his argument for a neutral monetary policy on the cyclical influence of money on relative investment flows in an accumulating economy. His policy prescription was to eliminate 'forced saving' by holding the quantity of money (multiplied by velocity) constant. Theoretical rational here was that the quantity of money must not change if the money rate is to reflect the natural rate.

Now Sraffa has shown that, if one really wants to define natural rates in barter-like terms, it is a necessary relative price effect that these rate will naturally diverge in an economy with accumulating going forward. The basis of Sraffa's argument is that any new accumulation is directed to different employments by divergencies of market prices from the 'natural' price, a standard price theory argument. ‘It will be noticed that, under free competition, this divergence of rates is as essential to the effecting of the transition [to a more capitalistic economy] as is the divergence of prices from costs of production; it is, in fact, another aspect of the same thing’ (Sraffa, 1932: 51).

The consequence of Sraffa's internal argument is that a theoretical investigation that utilizes the concept of a natural rate of interest to reflect barter state of saving and investment, must deal with that fact that ‘there may be as many "natural" rates of interest as there are commodities’. Parallel argument results in a contradiction for Hayek's system:

But in times of expansion of production, due to additions to savings, there is no such thing as an equilibrium (or unique natural) rate of interest, so that the money rate can neither be equal to, nor lower than it: the ‘natural’ rate of interest on producers' goods, the demand for which has relatively increased, is higher than the ‘natural’ rate on consumers' goods, the demand for which has relatively fallen (Sraffa, 1932: 51).

Hayek must either give up a unique natural rate as an argument for a constant money policy, abandon the whole Wicksellian conception, or fall back on one of the hated aggregates he began by disavowing to deliver himself from this quandary. This becomes particularly clear when, in reply to Sraffa, Hayek adverts to the possibility (or necessity) of a multitude of natural rates.[10] Sraffa's rejoinder seems to seal the case:

Dr. Hayek now acknowledges the multiplicity of the ‘natural’ rates, but he has nothing more to say on this specific point that they ‘all would be equilibrium rates’. The only meaning (if it be a meaning) I can attach to this is that his maxim of policy now requires that the money rate should be equal to all these divergent rates (Sraffa, 1932b: 251)

In summary, Sraffa's internal case against Hayek's use of the money-natural rate framework rests on the proposition that a literally defined 'natural'(commodity, barter) rate of interest does not exhibit the characteristics that Hayek requires of it. That is,[:]

• in Hayek's chosen context of an accumulating economy (outside the stationary, or 'balanced growth' long-period state), there is no unique barter rate with which to identify the money rate, as required for neutrality. 
• Divergence of rates and disequilibrium is not restricted to monetary economies, but also characterizes such real economies as those with which Hayek's argument contrasts money systems.

[9] If there is any doubt that Hayek (at this date at least) accepts the idea that 'equilibrium' implies the long period equality of prices and cost of production, and not some subjectivist short period concept or the more elaborate 'intertemporal equilibrium' construct(see note 10), his reply to Sraffa should dispel it. In the reply Hayek recasts his forced saving argument and makes use of this very concept. For example, consider following:

... as incomes rise in consequence of the preceding credit expansion and the mass of consumers, who under our assumption spend all their income on consumption goods, increase their expenditure accordingly, while the money available for investment in capital goods does not increase any longer, the vaule of some capital goods produced under the inducement of a relatively stronger demand for such goods will fall below their ^cost of production^ (Hayek, 1932: 244, emphasis added).

[10] An interesting issue that arises from this aspect of the exchange is the relationship between the analysis of different 'natural' rates and the latter day conception of 'intertemporal equilibrium'. Some commentators have identified Hayek's work in this period as the entry point of a new 'method' of economic theory (Garegnani, 1976; Milgate, 1979), whereby the traditional long-period concerns of the classical and neoclassical tradtions are jettisoned or an atemporal equilibrium construct where there is no tendency for rates of return to equalize.
  We cannot agree with the identification of Hayek's cycle theory with this approach. Our objection is a simple one. If indeed Hayek was working with this conception, it is very difficult to see how to conceive any of the various 'concertina effects' that were the set piece of all versions of his cycle theory. Forced saving, for example, is difficult to interpret in the setting of intertemporal equilibrium. Essentially what Hayek wanted to describe was a very complicated dynamic story about changes in the structure of production which 'cycled' around the 'natural', long-period, equilibrium structure of relative prices and production. Forced lengthening and inevitable reversion of his forced saving case is one of these dynamic stories. Yet in the intertemporal setting, as has been said many times, 'everything happens at once'. In fact, avoidance of these messy dynamic issues is the very appeal of the method of intertemporal equilibrium to finical general equilibrium theorist. But this modern approach, driven by the need for mathematical tractability, is conceptually quite distinct from Hayek's 'method of neutral money'.
   On the other hand, (... ...)
  Also, in his reply to Sraffa, perhap without much forethought under the pressure of the critique, he seems at points to appeal to this viewpoint in such claims as: '... there might, at any moment, be as many "natural" rates of interest as there are commodities, ^all^ of which would be ^equilibrium^rates; and which would all be the combined result of the factors affecting the present and future supply of the individual commodities...' (Hayek, 1932: 245).
  Yet as we note in the text and as Sraffa pointed out in his reply, this notion of multiple natural rates renders Hayek's neutral monetary policyㅡof a market rate equal to ^the^ real (equilibrium) rateㅡincoherent.

IV Curtain call: the method of neutral money once again

By conducting a parallel analysis from within Hayek's framework, Sraffa exposes the 'gaps' in the argument referred to by Kaldor.

• (a) Monetary expansion cannot be claimed as the source of forced saving reversals and a deus ex machina has to be brought in the form of the supposed powers of the banks'. 
• (b1) No unique natural rate can be identified in an expanding economy. Thus utilizing the natural rate as the theoretical maxim of monetary policy becomes incoherent even in a conceptual sense. Furthermore, (b2) the literal meaning of barter rates contradicts Hayek's presumption that the natural system is inherently stableㅡexhibiting a well-defined, unambiguous, unique natural rateㅡwhile monetary influences are inherently destablizing, and thus the primary force disturbing the natural equilibrium.
• (b3) Instead, Sraffa shows the virtual necessity of divergent (disequilibrium) rates in an expanding barter system.

Yet this entire critique was conducted within the purview of Hayek's own theoretical framework, and, as noted above, Sraffa consider it a dead end from the outset. 'But from the beginning it is clear that a methodical criticism could not leave a brick standing in the logical structure built up by Dr. Hayek.' It is, therefore, not surprising that the internal critique would eventually bring back into full view the inherent contradictions of conducting monetary theory from the standpoint of a theoretical system which proceeds as if there were no money at all.

If interest rates defined in barter terms necessarily exhibit divergencies in a growing economy, the natural rate loses any meaning as a policy guide. Playing on Hayek's own theme of immanent criticism of Wicksell, Sraffa claims that his{=Sraffa's} point about barter rates does not necessarily indict Wicksell:

This, however, {though it meets, I think, Dr. Hayek's criticism}, is not in itself a criticism of Wicksell. For there is a 'natural' rate of interest which, if adopted as bank-rate, will stabilize a price-level (i.e. the price of a composite commodity): it is an average of the 'natural' rates of the commodities entering into the price-level, weighted in the same way as they are in the price-level itself. What can be objected to Wicksell is that such a price-level is not unique, and for any composite arbitrarily selected there is a corresponding rate that will equalize the purchasing power, in terms of that composite commodity, of the money saved and of the additional money borrowed for investment (Sraffa, 1932: 51).

For Sraffa, this offers a solution to the dilemma of the ‘method of neutral money’. For any aggregate 'natural rate' selectd, it would be possible to identify a monetary policy that would be 'neutral' in the sense that saving and investment would be equal in money terms and in 'natural' terms. In other words, a money rate set equal to the composite natural rate would result in savings and investment retaining equal value in the sense that the money value of the savings and its value in terms of the selected composite commodity would remain relatively stable (i.e., the 'purchasing power' of money saved and invested would be stable in terms of that composite commodity.) Of course this would also mean that monetary policy only could be neutral in the sense of corresponding to a 'particular' non-monetary economy, out of an infinite possible number. Thus Wicksell has an escape from Sraffa's internal critique, but, this 'way of escape is not open to Dr. Hayek, for he had emphatically repudiated the use of averages' (Sraffa, 1932b: 251).

It is at this point that the immanent critique of the natural rate doctrine transcends the framework of neutral money. Having shown a logically consistent version of the aims of the method of neutral money, Sraffa steps back to consider the result:

Each of these monetary policies will give the same results in regard to saving and borrowing as a particular non-monetary economyㅡthat is to say, an economy in which the selected composite commodity is used as the standard of deferred payments. It appears, therefore, that these non-monetary economies retain the essential feature of money, the sigleness of the standard; and we are not much wiser when we have been shown that a monetary policy is 'neutral' in the sense of being equivalent to a non-monetary economy which differs from it almost only by name (Sraffa, 1932: 51).

Thus the technique of aggregated real rates can define a neutrality of a restricted type. Money can be neutral in the sense of stabilizing the puchasing power of a composite commodity which serves as the numeraire standard of an analogue 'barter' system (like Hayek's 'natural' system). Likewise, a monetary policy can be devised under which saving and investment in terms of either money or the composite commodity are balanced.

But this, again, brings up the original question of what a nonmonetary economy would look like. Sraffa's view, revealed above, is that the essential feature of money is the fact that it is the single standard for all transactions. Consequently, a true nonmonetary economy is one 'in which different transactions are fixed in terms of different standards'.[11] For this kind of economy, neutrality of the 'average-real-rate' variety will not do, since 'there are no monetary policies which can exactly reproduce their results' (Sraffa, 1932: 51).

By a rather circuitous route this point redefines the conclusion of the external critique presented at the beginning of the review. Sraffa had started by saying that if the method of neutral money is to be useful to monetary investigations it seems reasonable that the comparison it utilizes be made between a monetary economy and a nonmonetary economy.

This method of approach might have something to recommend it, provided it were constantly kept in mind that a state of things in which money is 'neutral' is identical with a state in which there is no money at all: as Dr. Hayek once says, if we 'eliminate all monetary influences on production ... we may treat money as non-existent' (p. 109). Thus the parallel analysis into 'neutral money' and various kinds of real money would resolve itself into a comparison between the conditions of a specified non-monetary economy and those of various monetary systems (Sraffa, 1932: 42).

He had further noted that comparison of various cases of disturbance to the equilibrium of these systems could identify 'the essential characteristics common to every kind of money, as well as their differences, thus supplying the elements for an estimate of the merits of alternative policy' (Sraffa, 1932). His complaint was that Hayek had lost sight of this task and dealt instead with the wholly different task of proving the desirability of one particular monetary policy.

Now we have seen Sraffa's presentation of one method of logically constructing a neutrality result utilizing the money-barter distinction of Hayek (and Wicksell). It is notable that the comparison offered in that case, following Hayek, ran strictly in terms of the purchasing power of savings and investment.

• Money was considered only in its role as the numeraire good. 
• A ‘neutral’ result was then defined as a monetary policy which left savings and investment ‘undisturbed’ in terms of the purchasing power of the selected commodity.

But having shown this limited sense in which we can finally tie down ‘neutrality’ so defined, Sraffa's objection is that the difference between the two systems is not really relevant to the difference between a monetary and a nonmonetary system.

• It is not relevant because the ‘essential’ element of the monetary economy, the singleness of the standard, is common to both. Consequently, neutrality of this sort does not really say much about the qualities unique to a monetary system. 
• And as for the specific question of savings and investment disequilibrium, the method of neutral money is irrelevant. ‘With or without money, if investment and savings have not been planned to match, an increase of saving must prove to a large extent “abortive”’ (Sraffa, 1932: 52). 
• But barter savings (which one presumes are goods produced, but not consumed) and money savings (which, quoting Robertson, Sraffa notes may be an ‘inducement’ to investment, but can never be the ‘source’), are not in any sense analogous as the method of money neutrality would have it. This distinction, of course, is part and parcel of the differences between a barter system and a monetary one.

What then is Sraffa's view of what would constitute a 'more truly' nonmonetary economy and so provide a more valid benchmark to utilize in the method of money neutrality? At the end of the review as at the beginning the fundamental differences between a monetary and a nonmonetary economy are constituted by ‘those characteristics which are set forth at the beginning of every textbook on money’. A monetary economy is unique in the pervasiveness with which the 'single standard' is used to fix the terms of ‘debts, and other legal obligations, habits, opinions, conventions, in short all kinds of relations between men’. By contrast, a nonmonetary economy would be one in which for these same relations between men ‘different transactions are fixed in terms of different standards’.

Somewhat surprisingly, little interest has been expressed in the role which money would play in, and the manner it should be incorporated into, Sraffa's own theory of relative prices (Sraffa, 1960). Sraffa offers only a single cryptic, but tantalizing, remark:

The rate of profits, as a ratio, has a significance which is independent of any prices, and can well be 'given' before the prices are fixed. it is accordingly susceptible of being determined from outside the system of production, in particular by the level of the money rates of interest (Sraffa, 1960: 33).

Having reviewed Sraffa's critical monetary work as presented in his exchange with Hayek, it is interesting to ask how this might fit into his larger theoretical outlook as we know it. In the brief passage just quoted, one is immediately struck by the fact that it implicitly contains many of the issues treated in the Sraffa-Hayek exchange. But in Production of commodities by means of commodities, Sraffa treats these issues within an entirely different methodological framework from that of Hayek.

First it is significant that Sraffa's theory of relative prices ^separates^ the theory of prices proper from the determination of the rate of profit (or distribution generally). This is in stark contrast to Hayek's virtual conflation of prices, production and interest. Secondly, Sraffa views the determination of his prices as a theoretical investigation 'concerned exclusively with such properties of an economic system as do not depend on changes in the scale of production or the proportion of "factors"' (Sraffa, 1960: v). Again this is diametrically opposed to Hayek' preferred method of analysing an accumulating economy and his deep interest in 'proportions of "factors"'.

These fundamental differences in outlook explain why the role of money will be radically different in Sraffa's sysem. If monetary policy can set the rate of interest and thus ultimately determine the rate of profits ('from outside'), as Sraffa suggests, then the relative price effects which follow will be due primarily to ^distributional^ concerns. Interestingly, this confirms Sraffa's view, expressed in the exchange, that the 'forced savings' consequent upon a change in monetary policy can be viewed as one class robbing another class of part of their income. Further, Sraffa's view that 'the rate of profits, as a ratio, has a significance which is independent of any prices' suggests obvious objections (now of a purely 'external' variety) to Hayek's central thesis that changes in the rate of interest are only stable if they conform to the implicit factor proportionalities of the natural price system.

Finally, it is interesting to note that in Sraffa's method, the role of money in fixing the rate of profits fulfils the theoretical necessity of closing the system. Thus, unlike what Sraffa viewed as the irrelevance of money to Hayek' theoretical system, it could be an essential element of his own theory of relative prices. There is a deep irony in this juxtaposition. Sraffa's theory of prices, so seemingly 'objective' and 'technological', requires money for completeness. Hayek's system, which was explicitly designed to illustrate monetary effects, is incapable of admitting money by the very nature of its characterization of money!

Concluding with this theme of Sraffa and money, It seems appropriate to point out that the combination of these fragmented comments can offer a glimpse (by interpolation) of what Sraffa might have viewed as the proper treatment of money. But applying Sraffa's strictures on Hayek to Sraffa's own later work on the foundations of value theory (Sraffa, 1960) suggests that this latter work must be conceived very narrowlyㅡas its subtitle suggest, as merely a ‘prelude to a critique’, not even a critique itself, surely not a constructive theory of a real monetary economy. Sraffa savaged Hayek for ignoring the social role of money as an institution for fixing ‘relations between men’. Yet it is the hallmark (and some would say the supreme fault) of his own price theory that it runs almost exclusively in terms of relations between ^things^. In the spirit of this objective theoey of relative prices one might be tempted to integrate money into his system by designating one of the commodities the money commodity (Hodgson, 1981). By this method what is essential about the money commodity is its own technical coefficient of production. But if our reading of Sraffa is correct, this approach is inconsistent with his outlook on the role of monetary factors. On monetary factors Sraffa seems much closer to Keynes.[12] Surely money may be a good, but its unique features relate to those issues of ‘habits, opinion and convention’ which are the stuff of relations between people, not things.


VII Reference

• Adarkar, B.P. 1935: ^The theory of monetary policy^. London: P.S. King & Son.
• Caldwell, B. 1986: Hayek's transformation. ^History of Political Economy^ 20, 513-41
• Coates, J. 1989: ^Ordinary language economics: Keynes and the Cambridge philosophers^. Ph.D. dissertation, University of Cambridge.
• ...