2- 货币主义的惩罚--英文原文

THE SCOURGE OF MONETARISM SECOND EDITION PART I The Radcliffe Report and Monetary Policy The Radcliffe Lectures delivered at the University of Warwick 1981 PART II Monetary Policy in the United Kingdom Evidence to the Treasury and Civil Service Committee July 1980 NICHOLAS KALDOR OXFORD UNIVERSITY PRESS 1985 LECTURE II The main fault of the Radcliffe Report was that it interpreted its own terms of reference very narrowly, and having come to conclusions that were so largely negative, it did not attempt to give any outline, or even a hint, of its views of how the very real problems which monetary policy was called upon to deal with could be tackled. The Committee was undoubtedly well aware of these problems -- not least because of the series of excellent memoranda which it received from distinguished British economists 1 -- and these comprised the persistence of inflation and its world-wide character which was so contrary to previous experience; inadequate growth which they ascribed to insufficient long-term investment; inadequate export performance in relation to competitors; and cyclical instability leading at times to excessive pressure of demand. Keynesian principles of economic management were universally accepted by all political parties and by academic economists -- there were no monetarists of those days -- and the question for discussion concerned only how the performance of the economy could be improved in terms of growth as well as the avoidance of inflation. The famous Coalition White Paper on Employment Policy of 1944 which accepted the maintenance of a high and stable level of employment as 'one of the Government's primary aims and responsibilities' was followed in 1956 by a further White Paper issued by a Conservative Government on the Economic Implications of Full Employment which stated that: 'The Government is pledged to foster conditions in which the nation can, if it so wills, realise its full potentialities for growth in terms of production and living standards'. 2 The main instrument for attaining these ends was the regulation of the pressure of demand by fiscal policy supplemented by monetary policy, and the main question at issue was how the various objectives ____________________ 1 I can strongly recommend to present-day students the study of some of the memoranda reprinted in the third volume of the Committee's Memoranda of Evidence. The volume contains over thirty memoranda submitted by economists, a number of whom (particularly A. J. Brown, J. C. R. Dow, R. F. Harrod, R. F. Kahn , I. M. D. Little, R. R. Neild, and R. Ross) gave an excellent presentation of the issues that has lost none of its relevance -- in sharp contrast with present-day writers on the subject, whose reasoning and terminology have been hopelessly befuddled by the outpourings of monetarist writers from across the Atlantic. 2 Cmd. 9725, para. 25. -17- of policy -- growth leading to rising living standards, as well as full employment, the avoidance of inflation, and a satisfactory balance of payments -- could be pursued more successfully through a better orchestration of instruments. The Keynesian model, on the basis of which economic management was conducted, had however one conspicuous weakness -- it assumed money wages as exogenously given. Indeed, everything in the General Theory was measured in terms of 'wage units' -- which meant that, in a closed economy, the rate of change in the price level was made to depend on the rate of increase in money wages relative to the rise in productivity. Fiscal and monetary policy were brought to bear to ensure the full utilization of resources but they were unavailing for dealing with wage-induced inflation. Some economists thought that keeping unemployment steady at a somewhat higher level -- say 2.5 instead of 1.5 per cent -- might suffice to prevent excessive wage increases; others, like myself, thought that full employment requires a permanent incomes policy, and elaborated a scheme to that effect as early as 1950. In fact a policy of restraint on personal incomes, introduced by Sir Stafford Cripps in 1948, successfully held the rise in wages and prices to below 3 per cent for two years in succession, 1949 and 1950, until the outbreak of the Korean war. But with the return of the Conservative Government in 1951, the political atmosphere for a scheme of this kind -- which invariably involved the attainment of some consensus over the distribution of income between wages and profits -- was lacking. So inflation went on -- though in the second half of the 1950s, thanks to falling raw material prices, it did so at a lesser rate than before and almost came to a standstill in 1958-9. But in the 1960s it started up again, showing distinct signs of acceleration towards the end of that decade, probably as a result of the inflationary effects of the Vietnam war. The Committee with rare perspicacity concluded that '. . . the economy of the United Kingdom in the 1960s will, in the relevant ways, be more like that of the 1950s than like that of any earlier period.' 3 Unbeknownst to the Committee a little man (I use this expression purely literally, meaning physical height, and in no other sense) was beavering away in America making correlations on times-series data which showed that fluctuations in the changes of the amount of money in circulation are closely related to changes in the level of aggregate ____________________ 3 Report of the Committee on the Working of the Monetary System, Cmnd. 827 (HMSO, August 19), para. 486. -18- money income -- the money national income. I am referring of course to Milton Friedman, who spent years in demonstrating a strong correlation between M and Y (where M stands for the stock of money and Y the money national income) and who convinced himself that this high correlation in itself constitutes a refutation of the Keynesian model of the economy and an empirical confirmation of the classical theory, the quantity theory of money. The quantity theory of money which dominated both economic and political thought in the nineteenth century (the best proof of which was the Bank Charter Act of 1844) asserted that the value of money varies in inverse proportion to its quantity, the relationship depending on the demand to hold wealth in the form of money expressed as a proportion of income. Suppose people desire to hold a stock of money, or an average daily balance, which is the equivalent of three months' income on expenditure. Both Marshall and Walras asserted that since the actual size of balances are given exogenously by the quantity of gold in existence (they were postulating that gold was the money commodity), and since all the gold that is anywhere must be somewhere, there must be a mechanism to ensure that actual balances are brought into conformity with desired real balances, and that mechanism consists of changes in the value of gold in relation to other commodities. If actual balances exceed desired balances, and people increase their spending so as to get rid of the excess, the value of money will fall, and conversely. There is therefore a particular commodity value of gold which ensures conformity of actual with desired balances. This doctrine is not inconsistent with the proposition that the quantity of gold is only 'given' for the time being, and over a period it can be increased by new production or decreased by nonmonetary uses of gold. In the long run, according to Ricardo and his followers, the value of gold will itself conform to its labour cost of production in the same way as in the case of other commodities. There was, however, always a complication, due to the existence of paper money. Originally this was thought not to make any difference: paper money was preferred as more convenient to handle and to carry about, but it was really no more -- or thought to be no more -- than a cloakroom ticket for the gold deposited with trustworthy persons like goldsmiths, and, later, moneylenders or bankers, who had strong-rooms for safe keeping. However, this simple idea about paper money could no longer be maintained when it was discovered, rather to the distaste of economists -19- like Walras, that the volume of paper money in circulation was a multiple of the amount of gold deposited in the vaults of banks. But after further thought it was concluded that this does not make any difference either provided that paper money maintains its parity with gold through convertibility, and provided that there are firm rules or conventions, whether legal or prudential, to ensure convertibility. Still later this doctrine was extended when bank deposits became a further circulating medium, and then it was still further extended to the case of inconvertible paper currencies, provided that the quantity of paper money in circulation was determined quite exogenously by the flat of the monetary authorities. After all, it was not the intrinsic value of gold but its limited quantity (its 'scarcity') which determined its value. This was the dominant doctrine in which all Anglo-Saxon economists of the present century were brought up -- whether Irving Fisher in America, or Keynes in Cambridge, who lectured on it for many years before and after World War I without questioning the basic principles. Its implications were that inflation is always the result of the 'overissue' of banknotes (to use Ricardo's original expression), whether caused by the financial needs of Governments (as in times of war) or the greed of banks who (with the connivance of the Central Bank) extended too much bank credit. Now Keynes's intellectual development, spread over several decades, consisted of a long struggle to escape from this theory; he succeeded in doing so in stages -- which meant that he never abandoned it altogether4. The first stage was the realization that labour is different from commodities: the labour market is different from commodity markets, in that an excess supply will not cause a reduction in wages, nor does an excess demand necessarily lead to a rise in wages, at least not immediately. Hence his opposition to the return to the Gold Standard at pre-war parity: the domestic price level is tied to the level of wages which are not adjusted downwards so as to keep supply and demand in equilibrium. The second stage came with the realization or recognition that effective demand for commodities in the aggregate is not determined by monetary factors but by autonomous demand financed by loan expenditures and the multiplier which depended on the propensity ____________________ 4 He himself complained, in an oft-quoted passage, of the difficulties of escaping from habitual modes of thought and expression, 'which ramify into every corner of our minds' ( The General Theory of Employment, Interest and Money ( London, 1936), Preface). Keynes was a pupil of Marshall and bore traces of it to the end of his life. -20- to save out of income. This meant that investment and savings, which are always brought into equality ex post do so through the adjustment of incomes and not, as the traditional theory had it, through movements in the rate of interest in the market for loans. This left the rate of interest 'in the air', as Keynes himself put it (because it could no longer be held that the rate of interest is the 'price' which equates savings with investment), until he thought of the idea of liquidity preference -- that people's demand for money will be the greater the lower the rate of interest -- which provided the mechanism through which monetary variables accommodate themselves to the 'real factors', the underlying relationships which generate the equilibrium level of effective demand. Unfortunately, the way he presented this solution was a modification of the quantity theory of money, not its abandonment. For he wrote: 5 M = L(Y, r) or M = k (r) Y where L(Y, r) represents the demand for money as a function of both the level of income Y and the rate of interest, r, while k(r) represents the demand for money expressed as a proportion of income, and (according to Keynes) is an inverse function of the rate of interest. Or, to put the same thing in Fisher's terminology: D≡Y=MV(r) (instead of Y=MV) This implies that all the adjustments of monetary to real factors are through changes in the velocity of circulation -- since the quantity of money, M, is still shown as an independent variable, determined by the monetary authorities. It was perhaps this form of presentation which led the Radcliffe Committee to the rather extreme-sounding statements about the variability of the velocity of circulation quoted in the first of these lectures. And it led young Milton Friedman into believing that the empirical validity of the Keynesian theory depended on the absence of any correlation between M and Y. Clearly if V adjusts to variations of Y, M and Y could not be closely related. Much to his surprise, he found the opposite -- a strong correlation between M and Y. He worked and worked and re-worked the historical series on money and income on all the data he could get hold of, and then extended it in time, and though he encountered some difficulties he was able to eliminate these ____________________ 5 Ibid., p. 189. -21- by the postulate of a variable time-lag of a nature that changes in M always preceded changes in Y by anything between six months and two years. However, as even people with an elementary knowledge of statistics know, leads and lags can only be established at the turning points of statistical series. Since in the time-series he examined (mainly data on money and income in the US) money and income were always rising (at least up to the Great Depression), sometimes faster and sometimes more slowly, it could not really be established whether either of the statistical series had a lead or a lag in relation to the other. Moreover, Friedman's emphatic reassertion of the quantity theory of money -- based on a stable demand function for money or a stable velocity: the two come to the same thing -- was crucially dependent on the quantity of money being really exogenous, determined by the flat of the monetary authorities quite independently of the demand for it. Or rather, in a dynamic context, it depended on the postulate that variations in the rate of increase in the money stock are the cause of variations in the rate of growth of money incomes occurring subsequently. When I first heard of Friedman's empirical findings, in the early 1950s, I received the news with some incredulity, until it suddenly dawned on me that Friedman's results must be read in reverse; the causation must run from Y to M, and not from M to Y. And the longer I thought about it the more convinced I became that a theory of the value of money based on a commodity-money economy it is not applicable to a credit-money economy. In the one case money has an independent supply function, based on production cost, while in the other case new money comes into existence in consequence of, or as an aspect of, the extension of bank credit. If, as a result, more money comes into existence than the public, at the given or expected level of incomes or expenditures, wishes to hold, the excess will be automatically extinguished -- either through debt repayment or its conversion into interest-bearing assets -- in a way in which gold could not be made to disappear from existence merely because particular persons find that they have too much of it. They can pass it on to others, but if they have less, others will have more. I find it difficult to get this point across except with the aid of a simple diagram. Keynes's theory of liquidity preference, as shown in that equation, has been diagrammatically represented in the following way: -22- The supply of money is given as M, which is therefore a vertical line. The demand for money is the greater the lower the rate of interest, for any given level of Y. At a given Y there is a particular rate of interest, r*, where the demand curve intersects the supply line -- the rate of interest is therefore the 'price' which equates the demand and the supply of money (as a stock) -- not the demand and supply for loans (which is a flow). The higher Y is, the higher is the demand for money -- so with an increase in Y the curve is shifted to the right. A series of curves can be drawn for different levels of income and their point of intersection with M, the given money stock, can be represented as a single LM curve invented by Hicks, who represented Keynesian equilibrium at the point of intersection of an IS curve (the latter showing points of equality of savings and investment at different levels of income) with the LM curve. (This diagram became very popular, particularly in America, giving rise to endless complications and false conclusions, mainly because the two curves were not in pari materia -- the one related to stocks and the other to flows.) -23- Now, in the case of credit money the proper representation should be a horizontal 'supply curve' of money not a vertical one. Monetary policy is represented not by a given quantity of money stock but by a given rate of interest; and the amount of money in existence will be demand-determined. Demand will vary with incomes as before, and it is possible that the rate of interest of the Central Bank (the old Bank Rate, now the MLR) will be varied upwards or downwards as a means of restricting credit or making credit easier, but this does not alter the fact that at any time, or at all times, the money stock will be determined by demand, and the rate of interest determined by the Central Bank. This means that from the point of view of that equation r is not a dependent but an independent variable; to determine r we need a further equation which in its simplest form could be written as: But of course from the point of view of any single country it could be a far more complicated equation, with the Bank Rate or Minimum Lending Rate being a function of the level, or of the rate of change in Central Bank reserves, of gold or foreign currencies, and/or the rate of interest in other financial centres, etc. To the extent that Friedman is right in his empirical conclusion that the demand for money, or rather the velocity of circulation, is not sensitive to changes in the rate of interest, the money stock becomes mainly dependent on demand, governed by the level of incomes; and monetary policy based on interest rate changes is totally ineffective as a regulator of the money supply. Any effect is thus an indirect one via the influence of changes in the interest rate on investment and hence on the level of incomes generated by a Keynesian multiplier-accelerator process. It -24- is an influence on the demand for money exerted through changes in the level of production and incomes, and not a direct effect on the desire to hold money. Whilst monetarists continually emphasize that the Central Bank can or should directly determine the quantity of money, or at least the 'base stock' of money, consisting of banknotes and bankers' reserves (or balances) with the Central Bank, in fact they can do no such thing, as a recent experience with the Federal Reserve or the Bank of England shows. (Even Friedman admitted this recently when he said that incompetent Central Banks like the Bank of England are not capable of regulating the money supply.) They cannot prevent either a depletion or an accumulation of 'high powered money' (or reserve money) except by a policy of en- or discouragement -- by raising or lowering the rate at which they are prepared to create reserves by discounting (or re-discounting) Treasury Bills and bonds. But the Central Bank cannot close the 'discount window' without endangering the solvency of the banking system; they must maintain their function as a 'lender of last