Unconventional monetary policy ante litteram: Richard Kahn and the monetary policy debate during the works of the Radcliffe Committee

Unconventional monetary policy ante litteram: Richard Kahn and the monetary policy debate during... Abstract The appointment of a Committee on the Working of the Monetary System in 1957 (Radcliffe Committee) in the UK gave the occasion to debate innovative ideas in the field of monetary policy. Many prominent economists participated in the Committee’s works, which ended with the publication of the Radcliffe Report in 1959. Making an exception to the rule of not mentioning individual contributions, the Radcliffe Report claims to ‘follow Professor Kahn … in insisting upon the structure of interest rates rather than some notion of the “supply of money” as the centrepiece of monetary action’. As we reconstruct in the paper, Kahn based his proposals on a conceptual framework, emphasising the link between inflation, wages and aggregate demand and the importance of combining incomes policy and term-structure control in view of achieving external equilibrium, full employment and price stability. One finding of our research is that Kahn’s recommendations regarding term-structure control—with their relevance to the present day—are best understood in connection with his theory of liquidity preference as presented in Kahn (1954). We also find that the Radcliffe Committee, while espousing Kahn’s policy recommendations, surrounded them with so many caveats as to render them de facto inoperative. As we suggest in our reconstruction, this depends partly on institutional considerations and partly on differences of opinion between Kahn and the Radcliffe Committee regarding monetary policy transmission and the causes of inflation. 1. Introduction In recent years, policymakers have deployed conventional and unconventional measures to combat financial instability, unemployment and risks of deflation. In this context, ‘[c]entral banks through their balance sheet policies and government through their debt management policies have sought to directly influence the long-term interest rate’ (Turner, 2014, p. 19). This has had repercussions on debt refinancing costs, monetary transmission and bank leverage, which a rapidly expanding literature is investigating. Neither attempts to use long-term interest rates as policy variables, however, nor debate about the relation between public debt management and monetary policy are completely novel,1 as we show in this paper, which focuses on the works of the Radcliffe Committee, established in the UK in 1957, and the subsequent publication of its Report, known as the Radcliffe Report (henceforth, RC, 1960A). The Committee on the Working of the Monetary System, chaired by Lord Radcliffe, included eight members. Only two of them, Richard S. Sayers and Alec Cairncross, were professional economists. As Cairncross (1999) recalls, it fell upon himself and Sayers to draft the final version of the Report. Sayers concentrated on the central, theoretical part, while Cairncross took on the introductory chapter, the capital market, external relations and statistics. The Committee met on 59 days to hear oral testimony, which appears in the Minutes of Evidence (henceforth, RC, 1960C). In addition to this, the written evidence fills three volumes (henceforth, RC, 1960B). Many of the ideas and policy recommendations presented in the Radcliffe Report came directly from Richard Sayers,2 who had expressed them in several papers and essays, which shared a general Keynesian orientation. The Radcliffe Report was regarded as one of the most important documents in Keynesian monetary economics (Chick, 1973; Cobham, 1992) and Sayers and Cairncross were generally identified as Keynesians.3 More specifically, the Radcliffe Report combined the widely held Keynesian idea that monetary policy influences aggregate demand through long-term interest rates and investment, with the insistence—which we trace back to Sayers—on the blurred boundary between monetary and financial assets and the influence of interest rates on the supply of liquid funds. The Report’s main recommendation was that monetary policy should determine the entire term structure, including long-term interest rates, with the aim of regulating liquidity in a broad sense and, through that, aggregate demand. Technically, term structure control required the combination of ad hoc open market operations, Bank rate policy and other controls (including the enforcement of cash and liquidity ratios on clearing banks and the regulation of bank advances). Coordination between monetary policy and national debt management were also part of this strategy, which aimed at reconciling full employment, price stability and external equilibrium. The experimental nature of these recommendations, which came at the end of an extensive consultation process involving some among the most prominent economists of the time,4 makes their reconstruction of particular interest and motivates our research. The resonance of the Radcliffe Report, at the time of its publication and in subsequent years, until Monetarism and Rational expectations submerged the Keynesian framework on which it was based, provides further motivation in this respect, especially in connection with the ongoing research on post-Keynesian monetary theory5 and the renewed interest in Keynes and Keynesian economics after the recent crisis.6 In this paper, in particular, we focus on the contribution by Richard Kahn to the works of the Radcliffe Committee. This choice can be justified as follows. Making an exception to the rule of not mentioning individual contributions, the Radcliffe Report claims to ‘follow Professor Kahn … in insisting upon the structure of interest rates rather than some notion of the “supply of money” as the centre-piece of monetary action’ (RC, 1960A, §395). This statement makes the investigation of Kahn’s position, as presented in his Memorandum and oral evidence in front of the Committee, of particular relevance to contextualize the Report’s conclusions. As we show in our paper, while connections between Richard Kahn’s ideas and the Radcliffe Committee are much more tenuous than the explicit reference made in the Report might suggest, Kahn’s contribution is worth investigating per se, as it provides a conceptual framework where the main task of the central bank consists in determining interest rates at different maturity dates with the aim of achieving simultaneously external and internal equilibrium. It is our main contention that Kahn’s policy proposals, with their clear Keynesian orientation,7 are best understood in connection with his theory of liquidity preference, as put forth in Kahn (1954), and in particular with his sophisticated view of the role agents’ expectations play in determining interest rates. By combining Kahn’s evidence to the Radcliffe Committee with his theory of liquidity preference, our approach complements Koutsobinas (2007, p. 56), who builds on Kahn (1954) (and Tobin) to support ‘the adoption by monetary authorities of some form of “bills-bonds” approach that is consistent with the importance of long term interest rates for the effectiveness of monetary policy’.8 While the Radcliffe Committee espoused Kahn’s recommendations regarding the need to make the term structure, rather than money supply, the centrepiece of monetary action, it differed from Kahn on the causes of inflation and monetary policy transmission. More specifically, Kahn regarded inflation as mainly caused by wages running ahead of labour productivity, with aggregate demand influencing price dynamics principally, but not exclusively, through its impact on wage negotiations.9 Based on this framework, Kahn recommended to keep long-term interest rates as low as possible, in order to stimulate investments and productivity, leaving it to incomes and fiscal policy to restrain any concomitant inflationary pressure. By contrast, the Radcliffe Committee envisaged inflation as mainly determined by the direct pressure of aggregate demand on available resources, to be restrained by raising long-term interest rates with the aim of discouraging lenders, instead of borrowers, through the operation of a lock-in effect10 (on this, see Rousseas, 1998, p. 121). Our findings go some way towards explaining how it came about that, in spite of the prominence given to Kahn, the secondary literature about the Radcliffe Report has largely neglected and sometimes misrepresented Kahn’s position. In particular, our research suggests that Kahn’s contribution has remained almost completely hidden behind that of Richard Sayers,11 and that reference to ‘Professor Kahn’ was in part misleading, as it created the impression of a complete harmony between Khan’s ideas and those of the Committee. We show in the paper that on the crucial issue of the term structure of interest rate, the Report attributed to Kahn ideas that were not entirely his own and were at least in part those of Sayers. In particular, while Kahn claimed that monetary authorities could determine long-term interest rates by swaying expectations about them through consistent actions, Sayers did not share this belief entirely. This fact, coupled with official opposition to any impression of tampering with the gilts market, would eventually lead to emasculation of Kahn’s proposals. While early reviews of the Radcliffe Report reflect this attitude,12Tily (2006, 2007 and, in particular, 2009) emphasises Kahn’s insistence, in front of the Radcliffe Committee, on recommending the advantages of low long-term interest rates to promote investment and full employment, in line with Keynes’s own indications. Rousseas (1998, pp. 119–20, 124), instead, presents Kahn’s views in opposition to the conclusions reached by the Radcliffe Report, although as part of larger, and less personal, views of ‘the economists’, and not in some relation to Kahn (1954), as we do in this paper (see also Marcuzzo and Rosselli, 2017, pp. 710–11). On the whole, however, Kahn’s active interest in anti-inflationary policies during the 1950s and 1960s and his original views on monetary policy as related to his monetary theory have remained largely neglected. It is especially on this latter point that we focus our attention in this paper. Our main conclusion is that, albeit that Kahn’s impact on the Radcliffe doctrine was clearly not decisive, his views on the centrality of managing expectations about long-term interest rates based on his theory of liquidity preference mark a significant contribution to post-Keynesian monetary economics, one which appears particularly relevant today, in connection with the application of unconventional monetary measures. Based on these considerations, the structure of the paper is as follows. Section 2 briefly describes the historical and analytical background and the main conclusions of the Radcliffe Report. Section 3 discusses Richard Kahn’s contribution to the works of the Radcliffe Committee, focusing on his Memorandum and oral testimony. Section 4 elaborates on the linkages between Kahn’s ideas and his theory of interest rates and monetary policy as presented in his 1954 paper on liquidity preference, and explores the institutional dimension of these ideas in connection with the analysis presented by the Bank of England and the Treasury. Section 5 concludes the paper. 2. Radcliffe Report: background and main policy recommendations Between 1932 and 1951, Bank rate, the chief instrument of British monetary policy, had been held at 2%. This contributed to keeping long-term interest rates at a low level, with positive repercussions on debt servicing costs and aggregate demand. This policy, known as ‘cheap money’, ended in 1951 amidst rising inflationary pressures and a negative and deteriorating balance of payments.13 The crisis in the balance of payments, and the concomitant fall in official reserves, led to the reactivation of monetary policy in 1951. This was followed by declining inflation until 1953. Inflation picked up again in 1954 and 1955, and a second balance of payments crisis occurred in 1956. This triggered a second cycle of restrictive monetary policy, which led Bank rate to peak at 7% in September 1957. Initially, monetary restriction seemed to yield the expected results, leading inflation to fall and the balance of payments to improve. However, after declining to 3.1% in 1956, during 1957 inflation picked up again, reaching 4.6% during the 12 months.14 This led to the emergence of doubts about monetary policy and its effectiveness in maintaining price stability. As Cairncross (1996) reconstructs, these doubts were part of a deeper contrast between the Treasury and the Bank of England on how to control inflation. The Treasury preferred direct controls on bank advances and recommended the enforcement of variable cash and liquidity ratios on banks as a means to contain the growth of money and prices. The Bank was sceptical about the effectiveness of these controls and preferred constant cash and liquidity ratios.15 It also favoured the containment of public expenditure in view of restraining inflationary pressures.16 Opinions on monetary policy and the appropriate level of interest rates were mixed and shifting. In this context, the government appointed the Radcliffe Committee, at the suggestion of the Bank of England, with the mandate ‘to inquire into the working of the monetary and credit system, and to make recommendations’. More details on the actual mandate of the Radcliffe Committee can be gleaned from its report. It is clearly stated in §5 that when the ‘Chancellor of the Exchequer announced the decision to appoint the Committee, the problem that was pressing itself most urgently on his attention was that of controlling inflation’. The ‘persistence of inflationary pressure’ after the measures adopted over the last two years had ‘encouraged fears that the inflation was not just a temporary maladjustment … but a fundamental and perhaps inseparable concomitant of the economic circumstances and state of opinion which emerged after the war’ (RC, 1960A, §5). The Radcliffe Committee regards regulation of total demand for goods and services with the aim of achieving full employment, price stability and external equilibrium as the main purpose of monetary policy, together with fiscal policy and other measures (RC, 1960A, §§382–83). Traditionally, monetary policy would be expected to work through the control of money supply, as envisaged by quantity theorists.17 However, the existence of non-bank financial intermediaries, which provide the economy with money substitutes and credit (RC, 1960A, §§523–24), implies that spending is not limited by the amount of money in existence but rather by the ‘overall liquidity position’ or ‘whole liquidity position’ of the economy, defined as ‘the amount of money people think they can get hold of, whether by receipt of incomes (for instance from sales), by disposal of capital assets or by borrowing’ (§390). 18 The Committee rejects the view that monetary policy would become more effective by substituting ‘for the traditional control of the supply of money a complex of controls over an indefinitely wide range of financial institutions’ (§394). Direct controls can be useful in case of emergency but, ordinarily, the central bank should control the overall liquidity position by manipulating ‘the entire structure of interest rates’, as distinct from action ‘confined to the short-end of the market’ (§§393 and 394). In relation to this specific issue, the Committee follows ‘Professor Kahn … in insisting upon the structure of interest rates, rather than some notion of the “supply of money” as the centrepiece of monetary action’ (§395). The Committee regards changes in interest rates as affecting aggregate demand in two ways. One is the ‘interest incentive effect’ by which a change in interest rates can induce a change in the incentive to hold inventories and/or purchase fixed capital goods. The other is the ‘general liquidity effect’, which works as follows. ‘A movement of interest rates implies significant changes in the capital values of many assets held by financial institutions; a rise in rates makes some less willing to lend because capital values have fallen, and others because their own interest structure is sticky. A fall in rates, on the other hand, strengthens balance sheets and encourages lenders to seek new business’ (§§393 and 394). In general, and coherently with the view expressed in Sayers (1958, ch. 7), the Committee was sceptical about the relevance of the first effect and more confident of the second (on this, see in particular RC, 1960a, ch. VI).19 Manipulation of the term structure requires monetary authorities to take a view on the appropriate level of short- and long-term interest rates and enforce this view by a combination of ad hoc open market operations, Bank rate policy and other controls. It also requires close coordination between monetary policy and national debt management (§§533–35), effective communication (§499) and wider access to economic statistics (§865), which together contribute to influencing ‘expectations and confidence’, regarded as prime determinants of interest rates. The Committee’s position, as summarised above, seems to envisage the possibility of reactivating monetary policy based on innovative conceptual premises, relying on interest rate control and synergy between monetary action and national debt management within a Keynesian policy aimed at full employment. As to that, the Radcliffe Report incorporated in the official view new and unconventional elements in monetary policy that had begun to circulate over the last decade, especially on the concept of the liquidity position being the new target of monetary policy. It also took into account elements of revision that had been introduced in economic theory since the end of the 1930s, in particular by the Oxford Economists Research Group. In his contribution to Oxford Studies in the Price Mechanism, edited by A. W. S. Andrews and T. Wilson in 1951, Sayers offered a survey of the literature on ‘The Rate of Interest as a Weapon of Monetary Policy’, in which he looked back at ‘[t]he belief that changes in the rate of interest powerfully influence economic activity’. This doctrine—Sayers argued in introducing his survey—had ‘reached its high-water mark in Keynes’s Treatise on Money’ (Sayers, 1951, p. 1). Afterwards came the inquiry by the Oxford Economists Research Group on the relationship between the rate of interest and investment decisions, the results of which were published in the earlier issues of Oxford Economic Papers.20 The ‘Oxford inquiry’ reached the conclusion that changes in the rate of interest had almost no effect on investment decisions. Although based on a very narrow basis of empirical evidence, the inquiry had a large impact on the profession, as Sayers’s survey clearly shows. In this paper (written early in 1949), Sayers remained relatively prudent on this issue, but he was nonetheless ‘astonished at the confidence and simplicity of Keynes’s argument’ (1951, pp. 1–2). In 1952, the Economic Journal published Kahn’s review of Oxford Studies in the Price Mechanism (Kahn, 1952B). While commenting favourably on Sayers’s survey of the literature, Kahn sounded very sceptical of the results of the ‘Oxford inquiry’ (as well as of Hall and Hitch’s full cost article of 1939, which was also reprinted in the book of 1951). By contrast, the fourth edition of Modern Banking (ch. 7) and §§386–87 of the Radcliffe Report bear witness to the fact that Sayers had eventually embraced the new ideas, strongly rejecting the theory that the price at which money could be borrowed was a decisive factor in determining the level of investment in favour of the idea that the decisive factor was the ‘availability’ of credit. On the other hand, the endorsement of a theory of interest rates as conventional phenomena and the parallel rejection of a stable relationship between money supply, liquidity preference and interest rates constitute as many elements of continuity between the theory underlying the Radcliffe Report and Richard Kahn’s ideas on the subject, as he presented them in front of the Radcliffe Committee. 3. Richard Kahn’s contribution to the works of the Radcliffe Committee In the opening section of his Memorandum to the Radcliffe Committee (Kahn, 1960), Richard Kahn reviews the state of the British economy in the 1950s, focusing on the impact of monetary and fiscal policy on economic activity, inflation and the balance of payments. Two aspects of his analysis are of particular interest in the context of the present study: his theory of inflation and his considerations on monetary policy and the supply of money. Kahn regards the level of prices as determined by three interacting factors: money wages, productivity and the level of aggregate demand. The price level raises to the extent that nominal wages—determined through negotiations between trade unions and employers—rise faster than productivity.21 Under normal conditions, demand influences wages directly and the price level indirectly, unless it is so high as to cause severe labour shortages, which drive nominal wages above negotiated levels. It is the rise in wages which is the cause of prices rising. In that sense all price inflations are ‘cost inflations’ (or wage inflation). But the speed at which wages, and therefore prices rise is likely in some degree to depend on demand. In that sense there is nearly always some measure of ‘demand inflation’. (Kahn, 1960, p. 138) Kahn’s position on inflation goes together with his refutation of the quantity theory of money as a causal process, a long-held position, as Marcuzzo (2002) among others recalls. In the Memorandum, Kahn reiterates that no direct link exists between money and the general level of prices. He also stresses that money velocity, far from being constant, adjusts to changes in money supply and subsequent changes in economic activity.22 A lower quantity of money causes interest rates to rise in connection with banks reducing their holdings of securities and their advances to private investors and industry. This leads security prices and private investment to fall, reducing inflationary pressures at the cost of rising unemployment. This side effect, together with the concomitant reduction in aggregate investment, is one of the principal reasons that led Kahn to be wary about monetary policy restriction as a cure against inflation.23 Following on these considerations, Kahn opines that the whole structure of interest rates matters for monetary transmission. Short-term interest rates count mainly for their influence on the international capital movements and the exchange rate while long-term interest rates matter principally for their influence on investment and domestic demand.24 As such, monetary authorities should adjust the former in view of balance of payments considerations and the latter to balance domestic demand and supply, bearing in mind connections between internal and external equilibrium.25 The Memorandum, while not entering into the analysis of the relationship between short- and long-term interest rates, treats their separate, but coordinated, determination as a technical problem, whose solution requires ad hoc operations. Within wide limits it is possible to achieve any desired structure of interest rates by a suitable combination of monetary policy with management of the National debt. The Exchequer, by issuing short-dated securities in the place of long-dated, or vice versa, can secure the desired shift in relative rates of interest against the background of a monetary policy which operated on rates in general. Of course, the interest charge on the National debt will thereby be affected and this consideration will play an important part in determining the objectives of policy. (Kahn, 1960, pp. 148–49) The Minutes of evidence (RC, 1960C) clarify the conceptual framework underlying Kahn’s recommendations. The questions put to Kahn by the members of the Radcliffe Committee revolve around the link between money supply and interest rates and the impact of changes in interest rates on investment and domestic demand, leaving inflation out of the picture. In oral evidence, Kahn reiterates his position about the primacy of interest rates over the supply of money as a monetary policy instrument capable of influencing spending decisions. Money is a stock, it is not a flow. The things which influence people’s conduct are flows, profits, incomes, that sort of thing: things you measure as rates per unit of time; not a stock which is simply a matter of how at any particular moment people decide to dispose of their personal wealth and companies of their liquid assets. (Kahn, RC, 1960C, Q. 10986) If a stock matters for business decisions, that is overall liquidity, rather than bank deposits only. In a developed financial system, companies have many alternative financing sources at their disposal (banks, non-bank financial institutions, bill market, bond markets, Stock Exchange), which monetary authorities must control simultaneously in order to regulate corporate spending decisions. This requires coordination between Bank rate policy, open market operations, minimum cash and liquidity ratios and direct regulation of bank advances,26 capital issues and investment allowances (on this, see also Kahn, 1960, pp. 150–52). In this context, any residual significance of money supply control per se rests on its relationship with interest rates and investment, which Kahn regards as substantial although slow in coming into effect and unpredictable in its results (Kahn, 1960, p. 158). Moreover, Kahn considers it impossible to quantify a priori the level of money supply required to achieve the desired term structure of interest rates, as the latter depends substantially on expectations of investors and their reaction to monetary decisions: If the market likes to help by altering its expectations, because it thinks that this is the way the cat is jumping, then the authorities will not have to do much. If the market proves very obstinate and does not believe that they are going to succeed in doing what they are trying to do, then they will have to do much more. (Kahn, RC, 1960C, Q. 10988) Based on these considerations, which reflect an approach heedful of liquidity preference and national debt management constraints, it is natural for Kahn to indicate the term structure of interest rates rather than the quantity of money as the centrepiece of monetary actions. As seen above, Kahn sees no obstacle, within very broad limits, to the possibility that monetary authorities may achieve their objectives, provided they are ready to operate both in money and capital markets, combining open market operations à outrance (Kahn, 1952A, p. 148) with different issuing techniques. This requires firm ideas about the appropriate structure of interest rates and consistent actions, carried out by skilled technicians, aimed at altering the relative balance between bills and bonds in order to achieve the desired term structure. Kahn insists on the fact that as long as markets perceive that the government is determined, ‘It is very difficult to conceive of a situation in which it would not be possible to arrive at any particular price, provided that, as I did say, one is considering this within reasonable limits. … [O]nce the market realizes that the authorities are serious they will dash in and help the authorities’ (Kahn, RC, 1960C, Q. 10988, 10993). Naturally, this requires that the authorities are ready to accept any variation in the amount of money and/or bills and bonds which may be required in order to achieve the desired term structure. Kahn’s recommendations to the Radcliffe Committee bear close resemblance to Keynes’s own proposals on national debt management as presented in the General Theory and again in 1945 to National Debt Enquiry (NDE).27 As Tily (2006) and Peden (2004, pp. 330–31) reconstruct, in the documents prepared for the NDE Keynes regarded it possible for monetary authorities to set short- and long-term interest rates at any desired level, provided they were ready to combine appropriate issuing techniques for gilts (tap issuance) with an elastic supply of Treasury bills. Keynes also favoured issuance of a wider range of bonds to meet different savings requirements.28 The NDE first draft report (dated 15 May 1945) quoted extensively from p. 203 of Keynes’s General Theory to explain the conventional element of the long-term rate of interest and the need to make it as clear as possible that the authorities are strongly committed to their target. It also emphasized that the tap issuance policy fitted in well with a government that was, in most of the cases, a net borrower. Kahn’s views are consistent with these indications, even if differences exist between the two approaches. First, the NDE report insists on the need to pose no limits to the possibility for the Exchequer to issue Treasury bills, something which occurred on an unprecedented scale in the 1940s. It also draws attention to the inflationary impact of this form of financing, which increases the liquid assets of banks and their ability to lend, and on how to limit it (e.g. through Treasury Deposits Receipts, TDR).29 Kahn seems not to be worried about this possibility, coherently with his refutation of any direct causal link from aggregate liquidity to inflation. Second, Kahn appears to share the confidence shown in the NDE report as to the possibility of modifying prevailing market conventions and long-term interest rates with them. As it seems, he was also more confident than Sayers (1958, pp. 172–76), who did not deny that changing the long-term expectations of the public was possible, but warned that it was also very difficult to do so. With the hindsight provided by Hugh Dalton’s recent experiment of ‘ultra-cheap money’, that Sayers (1958, p. 175) took as an example of the potentially disastrous consequences in case of failure, this came not unexpected.30 The determination of interest rates at different maturity dates raises a series of interesting analytical problems concerning the impact of monetary policy on investors’ expectations and portfolio allocation decisions, in the presence of uncertainty and imperfect substitutability between money, bills and bonds. Exploring how Kahn addressed these problems in the context of his own theory of liquidity preference, as presented in 1954, contributes to understanding the theoretical underpinnings of his recommendations to the Radcliffe Committee. 4. Kahn’s theory of interest rates and liquidity preference One of the key ideas Kahn presents in his 1954 Notes on Liquidity Preference is that the long-term interest rate depends on expectations about the long-term interest rate itself rather than about the short-term interest rate, as theorised by John Hicks and others since the 1930s.31 When a bill is bought rather than a bond the only relevant expectation determining the decision is what the bond rate will be when the bill matures. That expectation is certainly related to the expectation of what the bill rate will itself be at that same date. ... All this is, however, a very different thing from saying that the bond rate depends on expectations about the future of the bill rate itself, rather than of banking and monetary policy generally. Professor Hicks tries to reconcile the two different views, but only by appealing to the necessity for consistency between the two sets of expectations, which brings us back to an imaginary world of expectations held with unanimous and complete conviction. (Kahn, 1954, p. 78) Differently from Hicks and his school, Kahn considers knowledge about the future level of interest rates as highly uncertain, subject to doubt and disagreement, with different persons formulating different expectations about the future course of interests, which they hold with different degrees of conviction and different sensitivity to risk. This provides the basis of Kahn’s analysis of the precautionary demand for money, which he conducts focusing on the case of an investor choosing between money and long-term securities (fixed-rate irredeemable bonds for simplicity), assuming Treasury bills away. Kahn is aware that this assumption involves a serious departure from reality but, if one agrees with him that decisions to go long rather than short depend primarily on expectations about what the long-term interest rate will be when the bill matures, and only secondarily on expectations about the future bill rate, nothing fundamental is lost by focusing on money and bonds only. As Kahn adds, ‘[o]f course the short-term interest rate will still remain lurking in the picture in a certain sense, in the guise of expectations about the behaviour of the long-term rate itself, but these are subjective and personal to each individual concerned and are obscured by risk and uncertainty’ (on this, see Kahn, 1954, p. 79).32 This investor bases her decision on two elements: the current long-term interest rate i, and the rate at which she expects i to rise (or fall) over the relevant investment horizon. On the assumption that the coupon paid to bond holders is equal to 1 euro, the current price of the bond p is equal to the reciprocal of the long-term interest rate (p = i– 1). Defining v as the expected price of the bond at maturity of the first coupon, the investor will be indifferent between bonds and money, whose return is equal to zero, when 1+v–p=0. (1) Dividing both sides by p and isolating i, equation (1) becomes i=v(1/v–1/p), (2) where v (1/v – 1/p) = g represents the expected rate of variation in the long-term rate of interest. As long as v is unique and held with complete certainty, the investor’s speculative demand for money will be equal to zero if g < i and to her entire portfolio if g > i. In this case, the precautionary motive is inoperative and only marginal investors, whose expectations are consistent with equation (2), will hold both money and bonds. On the assumption of heterogeneous agents—each of them holding unique expectations with absolute certainty—the aggregation of individual positions yields the market demand for securities and its obverse, the speculative demand for money. Speculative demand for money, which varies inversely with i, together with that part of money supply which agents hold for investment purposes, determines the equilibrium level of the long-term interest rate.33 If individuals hold expectations with less than perfect confidence, however, uncertainty enters into the analysis. In this context, precautionary demand for money becomes relevant and all investors hold money and securities simultaneously, in variable proportions depending on their aversion to income and capital risk34 and on their lack of conviction about their own expectations. In this case, speculative and precautionary demand for money become almost indistinguishable from one another, introducing an element of indeterminacy in the relationship between money and interest rates. As Dardi (1994, p. 102) observes, this indeterminacy makes it impossible to know in advance the measure of the variation in money supply, required to bring about a certain variation in interest rates, just as argued by Kahn in his evidence to the Radcliffe Committee. Kahn does not exclude that, even assuming uncertainty, changes in money supply may cause the long-term interest rate to move in the opposite direction, but the effect is unpredictable, due to heterogeneity in investors’ reactions. Observing a change in money supply, some investors will not modify the composition of their portfolios at all, either because they have no clue about how that change may affect interest rates or because precaution overrides the speculative motive completely. Other investors will revise their expectations but will not re-allocate their portfolios as long as they are not sufficiently confident about their views. Others still will react very quickly, causing brusque oscillations in interest rates if their number prevails in the market. The coexistence of different groups of investors, ranging from ‘bankers’ to ‘widows and orphans’,35 shifting in terms of size and composition and depending on current circumstances, magnifies the indeterminacy of the relationship between liquidity preference and interest rates (on this aspect in particular, see Kahn, 1954, pp. 81–90), weakening the case in favour of a quantitative approach to monetary policy, as supported by prominent opinion in front of the Radcliffe Committee (see footnote 17). To the extent that liquidity preference is, a priori, undetermined, targeting money supply would provoke permanent oscillations in interest rates, with negative repercussions on investors’ expectations and the economy at large. If, on the contrary, monetary authorities put interest rates at the centre of their strategy, this will reduce uncertainty and stabilize expectations, enabling them to achieve the desired term structure. Doing so, evidently, turns money supply into an endogenous variable—a concomitant factor, as Kahn put it to the Radcliffe Committee—best left to itself,36 which ceases to be a causal influence on economic activity and has no obvious connection with price stability. Kahn regards expectations of price stability as essential for effective policy action, which combines incomes policy, aimed at maintaining nominal wages in line with labour productivity, with adequate aggregate demand management: The really important thought which underlies advocacy of price stability relates to the harm which is done to the economic system once it becomes generally recognized as a certainty that the price level will go on progressively rising without interruption and without limit. It is continuous and complete confidence in that expectation which does the harm. (Kahn, 1960, pp. 144–45) Monetary restriction, while technically possible, is inferior to incomes policy in restraining inflation as it reduces investment, productivity and productive capacity and may increase unemployment in the absence of efficient coordination between trade unions and manufacturers associations (e.g. §§53–54). Moreover, the apparent political neutrality of monetary actions, if compared to fiscal policy, builds an upward bias into long-term interest rates, which may damage the economy’s growth prospects in the medium to long run. These positions, coupled with his theory of liquidity preference, lead Kahn to refute the idea of controlling money supply in view of achieving price stability. In the same vein, he observes that it is not the overall budget deficit that matters for the control of inflation but the instruments used to finance it, their maturity and the size of their respective markets, as they impact on the shape of the term structure. Kahn’s recommendations gave rise to not only substantial analytical questions but also institutional and practical problems. In oral evidence, Kahn expressed his confidence in the ability of the Treasury and the Bank of England to achieve, within very wide limits, any desired structure of interest rates. This would require cooperation between the two institutions and deliberate if gradual actions to influence market expectations in the desired direction. However, neither cooperation nor deliberate action could be taken for granted. Indeed, official reluctance to countenance sharp fluctuations in government bond prices, combined with an institutional set-up designed to guarantee full coverage of the Exchequer financing needs, posed a formidable obstacle to any concrete attempt to vary long-term interest rates in a deliberate fashion. The Bank of England, for one, opposed the idea that monetary policy should influence directly the level of long-term interest rates, on substantial rather than technical grounds. In the Bank’s opinion, systematic intervention in the market for government bonds, different from simply following the market, would risk undermining the confidence in the government’s credit, particularly in view of the non-professional nature of many bondholders. As the Chief Cashier of the Bank pointed out to the Committee, It would, of course, be possible to force down the price of gilt-edged securities by direct sales on the market. But we take the view that if the Government, or the Bank, as the Government issuing bankers, were deliberately to sell the Government’s own securities in order to drive down the market price, it would greatly damage the Government’s credit. (RC, 1960C, Q. 1762)37 The existence of a close connection between monetary policy and national debt management was further emphasised by the Treasury: ‘[t]he choice and methods used to put the choice into effect depend upon considerations of National Debt and monetary policy’ (RC, 1960B, part II, n. 2, §21). On this latter point, the Treasury made a step further, considering what might happen if the government decided to change the whole structure of interest rates, and not only the short-term rate. The Treasury admitted that ‘[t]he Government can always, if it wishes, depress the price of gilt-edged securities, simply by placing a quantity of its own securities on the market’ (§35). However, the Treasury also pointed out that ‘[b]oth … methods of limiting people’s ability or willingness to raise money—control over the money supply through control over the supply of liquid assets, and the operation on security prices—are … subject today to much greater limitations than they have been in earlier periods’ (§37). These limitations came mainly from the need to roll over a significant part of Britain’s huge post-war public debt,38 and from the significant rise in capital expenditure, linked to reconstruction. Coupled with resurgent inflation, this left the Treasury and the Bank of England with little room for manoeuvre in issuance policy, which mainly aimed at replacing Treasury bills with short-term gilts (funding).39 Technically, the presence of the Government broker, trading actively in all maturities in the gilt-edged market, made it indeed possible to influence the entire term structure along the lines suggested by Kahn. In practice, however, an institutional set-up designed to secure finance for the Exchequer, combined with the reluctance of monetary authorities to provoke sharp rises in long-term interest rates and the possibility for private investors to abstain from rolling over maturing bonds, created a noticeable difficulty in following Kahn’s suggestions. By instructing its Issue Department to take up all the government bonds, which the public had not subscribed initially, the Bank of England provided an additional guarantee to the Exchequer.40 When absorbing government bonds in this way, however, the Issue Department was obliged to reduce its lending to the Exchequer by reducing its Treasury bill holdings. This, in turn, forced the Exchequer to issue more Treasury bills in the market, expanding liquidity. For the same reasons, bond sales by the Government Broker had the effect of putting the liquidity of the banking system under pressure. In this sense, the amount of liquid assets depended ultimately on the public’s willingness to absorb government bonds and on their expectations.41 To the extent that the Treasury and the Bank of England took those expectations as exogenously given, and chose to auction off all types of securities rather than resort to the tap issuance technique, this might explain their reluctance to adopt the measures recommended by the Radcliffe Committee: [T]he authorities have not been prepared to force interest rates on the longer bonds upwards in order to tempt holders of short bonds to switch to long bonds. … They have held the view that sales could only be made in any quantity on a rising market. … Their view that demand could not be stimulated by dropping prices—a view not easy to accept—was based mainly on the belief that the market for gilt-edged securities is dominated by expectations … and that it is the expectation of what prices will be tomorrow or next week, and not expectation of next year’s prices, that matters. (RC, 1960A, §551) The members of the Radcliffe Committee were not convinced by this interpretation and went so far as to lament the almost fatalistic attitude of monetary authorities in the face of movements of long-term interest rates (RC, 1960A, §552).42 In connection with this, echoing Kahn, they also claimed that Our view is that expectations have been overrated [by the authorities] as independent market forces, and that at time the influence of the authorities themselves on those expectations has been correspondingly underrated. (RC, 1960A, §563) In spite of these considerations, the members of the Radcliffe Committee could not ignore the position of the Treasury and the Bank of England and their dislike of variable long-term interest rates. This led the Committee to adopt a compromise position, which we can summarize as follows. Monetary policy and national debt management together can indeed determine the term structure of interest rates, by combining ad hoc open market operation, Bank rate policy and other controls. Financial stability and other considerations, however, make it advisable not to change the long-term interest rate too often and/or too abruptly. 5. Conclusions At the time of its publication and in subsequent years, the Radcliffe Report attracted considerable attention by supporters and detractors alike. This was true in particular of the idea that monetary policy should determine the entire term structure of interest rates with the aim of regulating the liquidity position of the economy and aggregate demand. As we reconstruct in this paper, the Radcliffe Committee shared Kahn’s indication to make the term structure of interest rates, rather than money supply, the centrepiece of monetary action. At the same time, we identify significant differences between the Radcliffe Report indications, which largely reflected the opinions of Sayers, and Kahn’s ideas. In particular, while Kahn was convinced that monetary authorities could determine the entire term structure and thereby influence aggregate demand through the demand for credit, Sayers did not share these beliefs entirely. On the one hand, he underlined how difficult it might be for the authorities to sway expectations about interest rates, even if he did not exclude this possibility altogether. On the other hand, he opined that the main influence of interest rates on aggregate demand worked through the supply, rather than the demand, for liquid funds. These differences of opinion, by themselves, go some way in reducing Kahn’s actual impact on the Radcliffe Report. But we think that the decisive factor that led to the emasculation of his ideas coincided with the opposition of the monetary authorities to countenance sharp policy-driven fluctuations in long-term interest rates. This opposition, combined with an institutional set-up designed to guarantee the exchequer financing needs in all cases, posed, in our opinion, the main obstacle to the full application of Kahn’s recommendations to the Radcliffe Committee. These recommendations are best understood in connection with Kahn’s views on interest rates and monetary policy as expressed in his 1954 paper on liquidity preference. The characterisation of different types of investors and Kahn’s insistence on the connection between precautionary and speculative motives led him to reject the idea of a stable well-defined liquidity function in view of a more problematic approach, where expectations about interest rates play a dominant role in the determination of the term structure. This introduces an element of indeterminacy in the setting of interest rates, which the Report’s insistence on the blurred distinction between money (notes and bank deposits) and near-money magnified. On the eve of the recent crisis, Tily (2006, p. 657) made this comment: ‘[w]hile the role of expectations and uncertainty in the determination of the rate of interest has been restored by Chick (1983), this restoration has not led to a recognition that expectations can be managed and the rate of interest brought under control’ (Tily, 2006, p. 657). We have shown in this paper that something similar happened during the workings of the Radcliffe Committee, as a result of the institutional and political obstacles which were in the way of implementing term-structure control, be it for its influence on the supply of liquid funds or on their demand. The Committee tried to overcome these obstacles by adopting a compromise position whereby monetary authorities should set the long-run path of interest rates without varying them too often and too abruptly. Evidently, this compromise took off much of the innovative edge from the Radcliffe Committee’s proposals, contributing to that impression of confusion and wooliness lamented by many contemporary observers. From this we conclude that, albeit that Kahn’s impact on the Radcliffe doctrine was clearly not decisive, his views on the centrality of managing expectations about long-term interest rates based on his theory of liquidity preference mark a significant contribution to post-Keynesian monetary economics, one which appears particularly relevant today, in connection with the application of unconventional monetary measures. The authors wish to thank Sylvie Diatkine, Ivo Maes, George Peden, Jan Toporowski and two anonymous referees for their useful comments on earlier draft of the paper. The usual disclaimers apply. The paper is the first product of a research project on Unconventional monetary policy ante litteram: the Radcliffe Committee and the debate on liquidity and the long-term interest as a monetary policy instrument. Financial support was provided by the ESHET (European Society for the History of Economic Thought) through the ‘ECB Grants’ programme 2015, sponsored by the European Central Bank. Carlo Cristiano was responsible for writing sections 1 and 2; Paolo Paesani for sections 3 and 4. The concluding Section 5 is a joint contribution by the two authors. Bibliography Allen , W. A . 2014 . Monetary Policy and Financial Repression in Britain, 1951–59 , London , Palgrave Macmillan Google Scholar CrossRef Search ADS Allen , W. A . 2016 . The British attempt to manage long-term interest rates in 1962–1964 , Financial History Review , vol. 23 , no. 1 , 47 – 70 Google Scholar CrossRef Search ADS Alon , T. and Swanson , E . 2011 . Operation Twist and the effect of large-scale asset purchases , FRBSF Economic Letter, vol. 13 Artis , M. J . 1961 . Liquidity and the attack on the quantity theory , Bulletin of the Oxford University Institute of Economics & Statistics , vol. 23 , 343 – 66 Google Scholar CrossRef Search ADS Artis , M. J . 1965 . Foundations of British Monetary Policy , Oxford , Basil Blackwell Cairncross , A . 1991 . Richard Sidney Sayers, 1908–1989 , Proceedings of the British Academy , vol. 76 , 545 – 61 Cairncross , A . 1996 . Prelude to Radcliffe: monetary policy in the United Kingdom, 1948–57 , in Cairncross , A ., Economic Ideas and Government Policy , London and New York , Routledge Google Scholar CrossRef Search ADS Cairncross , A . 1999 . Diaries: The Radcliffe Committee and the Treasury 1961–64, Institute of Contemporary British History, Cairncross Treasury Series , London , University of London Capie , F . 2010 . The Bank of England: 1950s to 1979 , Cambridge , Cambridge University Press Google Scholar CrossRef Search ADS Cardim de Carvalho , F. J . 2010 . Uncertainty and money: Keynes, Tobin and Kahn and the disappearance of precautionary demand for money from liquidity preference theory , Cambridge Journal of Economics , vol. 34 , no. 4 , 709 – 25 Google Scholar CrossRef Search ADS Chick , V . 1973 . The Theory of Monetary Policy , London , Gray-Mills Chick , V . 1983 . Macroeconomics After Keynes , Oxford , Philip Allan Chick , V. and Tily , G . 2014 . Whatever happened to Keynes’s monetary theory ?, Cambridge Journal of Economics , vol. 38 , 681 – 99 Google Scholar CrossRef Search ADS Cobham , D . 1992 . Radcliffe Committee , pp. 265 – 66 in Eatwell , J. , Milgate , M. and Newman , P . (eds), The New Palgrave Dictionary of Money and Finance , London, Macmillan. Committee on the Working of the Monetary System (Radcliffe Committee) . 1960A . Report, Cmnd. 827 , London , HMSO Committee on the Working of the Monetary System . 1960B . Principal Memoranda of Evidence , 3 vols., London , HMSO Committee on the Working of the Monetary System . 1960C . Minutes of Evidence , London , HMSO Dardi , M . 1994 . Kahn’s theory of liquidity preference and monetary policy , Cambridge Journal of Economics , vol. 18 , 91 – 106 Fellner , W. , Gilberts , M. , Hansen , B. , Kahn , R. F. , Lutz , F. and de Wolff , P . 1961 . The Problem of Rising Prices , Organization for European Economic Co-Operation Fforde , J . 1992 . The Bank of England and Public Policy: 1941–1958 , Cambridge , Cambridge University Press Goodhart , C . 2012 . Monetary policy and public debt , Financial Stability Review , vol. 16 , 123 – 30 Gurley , J. G . 1960 . The Radcliffe Report and evidence , American Economic Review , vol. 50 , 672 – 700 Howson , S . 1987 . The origins of cheaper money, 1945–7, Economic History Review , New Series , vol. 40 , no. 3 , 433 – 52 Howson , S . 1988 . Cheap money and debt management in Britain 1932–51 , in Cottrell , P. L. and Moggridge , D. E . (eds), Money and Power: Essays in Honour of L. S. Pressnell , Basingstoke , Macmillan Google Scholar CrossRef Search ADS Howson , S . 1993 . British Monetary Policy, 1945–1951 , Oxford , Oxford University Press Kahn , R. F . 1952A . Comment on ‘Monetary policy and the crisis’ , by C. Kennedy, Bulletin of the Oxford University Institute of Statistics , vol. 14 , 147 – 53 Kahn , R. F . 1952B . Oxford studies in the price mechanism , Economic Journal , vol. 62 , no. 245 , 119 – 30 Google Scholar CrossRef Search ADS Kahn , R. F . 1954 . Some notes on liquidity preference , Manchester School of Economic and Social Studies , vol. 22 , 229 – 57 ; pp. 72–96 in Kahn, R. F. (1972), Selected Essays on Employment and Growth, Cambridge, Cambridge University Press Google Scholar CrossRef Search ADS Kahn , R. F . 1960 . Memorandum of evidence submitted to the Radcliffe Committee , pp. 124 – 52 in Kahn, R. F. (1972), Selected Essays on Employment and Growth , Cambridge , Cambridge University Press Kaldor , N . 1960 . The Radcliffe Report , Review of Economics and Statistics , vol. 42 , 14 – 19 Google Scholar CrossRef Search ADS Koutsobinas , T. T . 2007 . Monetary policy formation at the long-term margin: a Kahn-Tobin framework , pp. 43 – 58 in Arestis , P. , Hein , E. and Le Heron , E . (eds), Aspects of Modern Monetary and Macroeconomic Policies , London , Palgrave Google Scholar CrossRef Search ADS Kregel , J. A . 2013 . Was Keynes’s monetary policy à outrance in the treatise, the model for ZIRP and QE ?, pp. 135 – 48 in Hirai , T. , Marcuzzo , M. C. and Mehrling , P . (eds), Keynesian Reflections , New Delhi , Oxford University Press Laidler , D . 1989 . Radcliffe, the quantity theory and monetarism , in Cobham , D. , Harrington , R. and Zis , G . (eds), Money, Trade and Payments—Essays in Honour of Dennis Coppock , Manchester , Manchester University Press Laidler , D . 2007 . Monetary policy and its theoretical foundations , pp. 14 – 31 in Mayes , D. G. and Toporowski , J . (eds), Open Market Operations and Financial Markets , London and New York , Routledge Google Scholar CrossRef Search ADS Marcuzzo , M. C . 2002 . The demise of the quantity theory of money , History of Economic Ideas , vol. 10 , 49 – 62 Marcuzzo , M. C. and Rosselli , A . 2017 . Richard F. Kahn (1905–1989) , pp. 705 – 22 in Cord , R. A . (ed.), The Palgrave Companion to Cambridge Economics , vol. 2 , Basingstock, Palgrave Macmillan. Google Scholar CrossRef Search ADS O’Hara , G . 2012 . Governing Post-War Britain: The Paradoxes of Progress, 1951–1973 , London , Palgrave Macmillan Paesani , P . 2017 . On the relevance of Keynes to the contemporary world: post-crisis views on economic policy, capitalism and international monetary reform , Annals of the Fondazione Luigi Einaudi , vol. 51 , June, 249 – 68 Peden , G. C . (ed.) 2004 . Keynes and His Critics: Treasury Responses to the Keynesian Revolution: 1925–1946 , Oxford , Oxford University Press for the British Academy Robinson , J. V . 1951 . The Rate of Interest and Other Essays , London , Macmillan , pp. 3 – 30 Rousseas , S . 1998 . Post-Keynesian Monetary Economics , 3rd ed ., London , Palgrave Macmillan Google Scholar CrossRef Search ADS Sawyer , M . 2002 . Economic policy with endogenous money , in Desai , M. , Dow , S. and Arestis , P . (eds), Essays in Honour of Victoria Chick , vol. 1 , London , Routledge Sayers , R. S . 1951 . The rate of interest as a weapon of economic policy , pp. 1 – 16 in Wilson , T. and Andrews , P. W. S . (eds), Oxford Studies in the Price Mechanism , Oxford , Clarendon Press Sayers , R. S . 1955 . The determination of the volume of bank deposits: England 1955–56 , Banca Nazionale del Lavoro Quarterly Review , reprinted in Central Banking after Bagehot, pp. 92 – 107 Sayers , R. S . 1958 . Modern Banking , 4th ed ., Oxford , Clarendon Press Sayers , R. S . 1960 . Monetary thought and monetary policy in England , Economic Journal , vol. 70 , 710 – 24 Google Scholar CrossRef Search ADS Seldon , A . (ed.) 1960 . Not Unanimous: A Rival Verdict to Radcliffe’s on Money , London , Institute of Economic Affairs Tily , G . 2006 . Keynes’s theory of liquidity preference and his debt management and monetary policies , Cambridge Journal of Economics , vol. 30 , no. 5 , 657 – 70 Google Scholar CrossRef Search ADS Tily , G . 2007 . Keynes’s General Theory, the Rate of Interest and Keynesian Economics: Keynes Betrayed , London , Palgrave Macmillan Google Scholar CrossRef Search ADS Tily , G . 2009 . The policy implications of the general theory , Real-World Economics Review , vol. 50 , 16 – 33 Tobin , J . 1958 . Liquidity preference as behavior towards risk , Review of Economic Studies , vol. 25 , 65 – 86 Google Scholar CrossRef Search ADS Turner , P . 2013 . ‘ Benign Neglect of the Long-Term Interest Rate ’, BIS Working Paper Series no. 403 Turner , P . 2014 . Is the long-term interest rate a policy victim, a policy variable or a policy lodestar ?, pp. 19 – 55 in Chadha , J. S. , Durré , A. C. J. , Joyce , M. A. S. and Sarno , L . (eds), Developments in Macro-Finance Yield Curve Modelling , Cambridge , Cambridge University Press Footnotes 1 On these aspects, focusing on the UK and the USA, see Allen (2016), Alon and Swanson (2011), Turner (2013, 2014) and the references cited therein. 2 Cairncross (1991, p. 550) traces back the ideas that Sayers would later reproduce in the Radcliffe Report far back in time to the first edition of Modern Banking (1938). The fourth edition of this book saw light in 1958, during the period in which Sayers was an active member of the Committee. See also Sayers (1955, 1960). 3 On Cairncross’ Keynesianism, see O’Hara (2012, ch. 7); on the Sayers’ association with Keynes, see Cairncross (1991). 4 ‘The Radcliffe Report [1959] was a document of tremendous importance to the theory of monetary policy. … The submitted evidence and answers to their questions embody the thinking, in all its diversity, of a wide range of economists, bankers, and administrators on the subject at the time’ (Chick, 1973, p. 58). 5 In particular, we were inspired by Tily (2006, 2007) and Chick and Tily (2014), who insist on the relative neglect of Keynes’s theory of liquidity preference by modern and contemporary reappraisals of Keynesian economics. 6 For a survey of post-crisis contributions on Keynes and Keynesian economics, see Paesani (2017). 7 In chapter 37 of his Treatise on Money, Keynes proposed open market operations to be carried out ‘à outrance’ in order to force down interest rates at all maturities as an extreme anti-cyclical monetary measure in the context of the slump of 1930. The idea of managing the entire structure was then revived in the General Theory and during the National Debt Enquiry of 1945 (more on this below, in Section 3). 8 Cardim de Carvalho (2010) explores how Kahn and Tobin address the role of risk and uncertainty in determining the precautionary demand for money, seen as a key component of liquidity preference, whose fluctuations influence employment and monetary transmission. 9 For a full exposition of this doctrine, see Fellner et al. (1961, ch. 5), to which Kahn was one of the contributors. 10 The lock-in effect, also known as Roosa effect, consists in the reluctance of banks and other lending institutions, fearful of capital losses, to liquidate securities in view of expanding loans and anticipations in a context of rising interest rates. 11 In his obituary of Sayers, Cairncross (1991, pp. 553–54) would later confirm that the sections of the Report which deal with economic theory had been drafted by Sayers. Cairncross also reported of Sayers being ‘bitterly disappointed by the reception of the Report—“two years of my life—two years wasted”, he once exclaimed’—thus emphasizing that the Radcliffe doctrine is, largely, Sayers’s doctrine. Some overemphasis on the relevance of Sayers’s works might not be out of place in an obituary, but Cairncross’s account finds ample support in the fourth edition of Modern Banking (Sayers, 1958)—which was given to the press in the spring of 1957 (the preface is dated April 1957), just when the Radcliffe Committee was about to start—as well as in Sayers (1960). 12 Gurley (1960, p. 678), in particular, commented ironically on the idea that interest rates might depend on expectations as to what the interest rate would be: ‘It is fair to say, I think, that the Committee looks upon “expectations and confidence” of the public as the chief determinant of rate levels—though lip service is paid now and then to the role of liquidity. And these expectations are greatly affected by what the authorities say, by the public’s interpretation of their mood. That is, expectations are molded by “faces” made by the authorities’. Gurley himself noticed that ‘The Committee was also influenced in this view by R. F. Kahn’, but he suggested that the Committee had ‘gone somewhat further than Kahn did’ in his evidence (1960, pp. 679, 691). Gurley plainly affirmed that the argument based on what would nowadays be called forward guidance made no sense. Unfortunately, this latter argument, certainly one of the most unconventional of the Radcliffe Report, is one of the main influences that Kahn may have had on the Report. Kaldor (1960) seems to suggest that the Radcliffe Committee might have adopted Kahn’s doctrine for reasons that were not coherent with Kahn’s own view, but did not develop this point. Chick (1973, ch. 4) avoided any reference to Kahn in her critical review of the Radcliffe Report, presented as strongly influenced by the ideas of Sayers. 13 On this, see Howson (1988, 1993). 14 Data on inflation are from Allen (2014, appendix A, table B). 15 On this aspect, see Artis (1965, pp. 43–46). 16 Further archival evidence on this point is now available in Capie (2010, pp. 88–90). 17 Seldon (1960) collects the opinions of those who opposed the Radcliffe Committee’s conclusion based on traditional quantity theory arguments (on this, see also Laidler, 1989). 18 The absence of a clear definition of the concept of whole liquidity position was lamented by many observers at the time. On this, see Artis (1961, pp. 345–50) and Gurley (1960), among others. 19 On this, see Artis (1961, pp. 360–64) and the references therein. 20 As explained in the preface to the book edited by Andrews and Wilson in 1951, ‘because of war-time and other difficulties, these issues were small, and have been unobtainable for some time’. For this reason, some of these papers were selected for reprint in the volume of 1951. 21 Segmentation between trade unions combined with competition among them and employers causes a permanent inflation bias whereby each category tries to keep ahead of others and employers pass wage increases on to prices. This generates an inflation bias, which Kahn hopes to contain through the centralization of wage bargaining and adoption of specific wage policies together with measures capable of influencing inflation expectations. 22 On this, see Kahn (1960, p. 147). Kahn shares Sayers’s refutation of the idea of constant money velocity even if based on different arguments. While Kahn emphasizes the adjustment of money velocity to changes in economic activity, for a given money supply, Sayers (1960, p. 715) focuses on the endogenous adjustment of money supply, and more generally of liquid means. ‘The trouble [with the concept of velocity] begins when we jump to the conclusion that a rise in the ratio of payments to bank deposits means only that firms have decided to put their bank deposit to faster uses, whereas much of what has happened is that firms have decided to give more credit and people have shifted some of their savings into the hands of more nimble intermediaries. The artificiality of the concept [of velocity of circulation] lies, in short, in its reliance on a distinct and identifiable category of Money’ (Sayers, 1960, p. 715). 23 Kahn expressed his support for cheap money policy as a way of promoting investment on several occasions during the 1950s. He did so in the symposium on the reactivation of monetary policy published in the Bulletin of the Oxford University Institute of Statistics in 1952 (Kahn, 1952A) and again in two articles published in the Financial Times on 3–4 June 1954 under the title ‘The case for cheap money’. In this latter article in particular, Kahn described a situation in which unemployment was at a satisfactorily low level, the British economy was growing but productivity was not growing fast enough to keep pace with other countries (e.g. Germany). For Kahn, too large a share of aggregate income went to consumption, thus reducing the resources that remained available for increasing productivity via investment. In order to provide a remedy to this situation, Kahn strongly argued in favour of using monetary policy: ‘Monetary policy is surely the answer—in the shape of cheap money and credit expansion’. 24 Short-term interest rates mattered also for their influence on stock-holding, as discussed by many participants in the works of the Radcliffe Committee, including Kaldor (RC, 1960B, vol. 3, Memorandum, n. 20, §27) and Little, Neild and Ross (RC, 1960B, vol. 3, Memorandum, n. 23). Sayers (1958, pp. 167–69) refused this doctrine, which was presented in Hawtrey’s traditional version. Although without mentioning Hawtrey’s name, the Report (§386) follows Sayers’s argument in this respect. 25 As Kahn mentions at the beginning of the Memorandum, he shares this idea with I. M. D. Little, R. R. Neild and C. R. Ross, who discuss it in their Memorandum to the Radcliffe Committee titled ‘The Scope and Limitations of Monetary Policy’ (RC, 1960B, part 13, no. 23, pp. 159–67). 26 ‘It is not only a question of rates of interest. The ability of the banks to make advances has an influence independently of actual rates of interest. … Therefore I would argue that there should be imposed on the banks a restriction on the amount of advances which they can make’ (Kahn, RC, 1960C, Q. 10993). On this point, in §60 of his Memorandum, Kahn referred to Keynes’s ‘fringe of unsatisfied borrowers’ in chapter 37 of the Treatise on Money. 27 On this, see Collected Writings of J. M. Keynes, vol. 7, p. 206; Tily (2006) and Peden (2004, ch. 11). 28 On Keynes’s evolving ideas about the possibility that monetary authorities influence both short- and long-term interest rates and on connections between these ideas and unconventional monetary policy today, see Kregel (2013). 29 TDR, introduced as a war-time measure in July 1940, were non-marketable short-term debt instruments, which banks were obliged to buy but could not use as collateral. TDR were discontinued after the war and their reintroduction met with strong oppositions by the banking community during the 1950s (Fforde, 1992, pp. 635, 651). On TDR’s potential to curb inflation, see Sayers (1956, pp. 220–21). 30 On Dalton’s policy, see Howson (1987) and Fforde (1992, pp. 327–59). 31 As Kahn acknowledges at the start of his essay on liquidity preference, this assumption marks a significant dividing line between the Hicks school, which includes Kalecki and Kaldor, and the alternative school of thought, comprised of Robertson, Harrod and J. Robinson (Kahn, 1954, p. 72). Considerations of space prevent us from reconstructing the vast British literature on the subject between the 1930s and the 1950s. 32 For a formalization of Kahn’s theory in the case of three assets (money, irredeemable bond, short-term commercial paper), see Dardi (1994, section 3). Our notation is close to that used by Dardi. 33 The same result obtains if individual expectations are not unique but still held with complete certainty. The only difference with the respect to the case of unique expectations is that each agent, as long as she is sufficiently risk averse, will hold both money and bonds. On this and related aspects, see also Tobin (1958). 34 For a definition of these two types of risk, which Kahn draws from Hawtrey, see Kahn (1954, pp. 81–82). 35 ‘Bankers’ are expert investors who fear capital losses with particular intensity, prefer short-run liquid investments and enter into the bond market only when very confident that bond prices are rising, possibly in connection with bull speculation. On the other hand, ‘widows and orphans’ are inexpert investors who especially fear income losses, prefer long-run safe securities and leave the bond market only when sure that bond prices will fall, possibly in connection with bear speculation. The simultaneous presence of these and other types of investors, each with its own preferred habitat, contributes to making money, bills and bonds imperfect substitutes, enhancing the possibility of a separate but coordinated determination of interest rates at different maturity dates. On the notion of bankers vs. widows and orphans, see also J. Robinson (1951). 36 On the possibility of interpreting Kahn’s position and the ensuing conclusions of the Radcliffe Committee in the light of the post-Keynesian concept of endogenous money supply, see Sawyer (2002). 37 On the reluctance of the Bank of England to countenance funding (substitution of Treasury bills with government bonds) as a means of containing liquidity and inflation in a falling gilt market, see also Capie (2010, pp. 88–90). 38 Britain’s total National Debt was 175.2% of GDP in 1951 and 112.5% in 1959 (Allen, 2014, table A2, p. 245). 39 On this, see Allen (2014, table A4, pp. 248–49 in particular). 40 On the operations of the Government broker, see RC (1960A, §§581–82 and 1960C, Q. 11921–-12065). 41 When bond prices fell (as might occur in case of rising inflation), investors not only abstained from underwriting new bond offers but became net sellers of bonds. In this case, the Issue Department would be compelled to absorb government bonds, leading the amount of liquid assets to increase with adverse repercussions on inflation and bond prices. In his Memorandum and oral evidence in front of the Radcliffe Committee, Kahn gives the impression of downplaying the relevance of these institutional obstacles even if he does not ignore them completely. 42 On this official attitude about the behavior of investors, see also Goodhart (2012, p. 125). © The Author(s) 2017. Published by Oxford University Press on behalf of the Cambridge Political Economy Society. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Cambridge Journal of Economics Oxford University Press

Unconventional monetary policy ante litteram: Richard Kahn and the monetary policy debate during the works of the Radcliffe Committee

Loading next page...
 
/lp/ou_press/unconventional-monetary-policy-ante-litteram-richard-kahn-and-the-aBRrh3CS00
Publisher
Oxford University Press
Copyright
© The Author(s) 2017. Published by Oxford University Press on behalf of the Cambridge Political Economy Society. All rights reserved.
ISSN
0309-166X
eISSN
1464-3545
D.O.I.
10.1093/cje/bex072
Publisher site
See Article on Publisher Site

Abstract

Abstract The appointment of a Committee on the Working of the Monetary System in 1957 (Radcliffe Committee) in the UK gave the occasion to debate innovative ideas in the field of monetary policy. Many prominent economists participated in the Committee’s works, which ended with the publication of the Radcliffe Report in 1959. Making an exception to the rule of not mentioning individual contributions, the Radcliffe Report claims to ‘follow Professor Kahn … in insisting upon the structure of interest rates rather than some notion of the “supply of money” as the centrepiece of monetary action’. As we reconstruct in the paper, Kahn based his proposals on a conceptual framework, emphasising the link between inflation, wages and aggregate demand and the importance of combining incomes policy and term-structure control in view of achieving external equilibrium, full employment and price stability. One finding of our research is that Kahn’s recommendations regarding term-structure control—with their relevance to the present day—are best understood in connection with his theory of liquidity preference as presented in Kahn (1954). We also find that the Radcliffe Committee, while espousing Kahn’s policy recommendations, surrounded them with so many caveats as to render them de facto inoperative. As we suggest in our reconstruction, this depends partly on institutional considerations and partly on differences of opinion between Kahn and the Radcliffe Committee regarding monetary policy transmission and the causes of inflation. 1. Introduction In recent years, policymakers have deployed conventional and unconventional measures to combat financial instability, unemployment and risks of deflation. In this context, ‘[c]entral banks through their balance sheet policies and government through their debt management policies have sought to directly influence the long-term interest rate’ (Turner, 2014, p. 19). This has had repercussions on debt refinancing costs, monetary transmission and bank leverage, which a rapidly expanding literature is investigating. Neither attempts to use long-term interest rates as policy variables, however, nor debate about the relation between public debt management and monetary policy are completely novel,1 as we show in this paper, which focuses on the works of the Radcliffe Committee, established in the UK in 1957, and the subsequent publication of its Report, known as the Radcliffe Report (henceforth, RC, 1960A). The Committee on the Working of the Monetary System, chaired by Lord Radcliffe, included eight members. Only two of them, Richard S. Sayers and Alec Cairncross, were professional economists. As Cairncross (1999) recalls, it fell upon himself and Sayers to draft the final version of the Report. Sayers concentrated on the central, theoretical part, while Cairncross took on the introductory chapter, the capital market, external relations and statistics. The Committee met on 59 days to hear oral testimony, which appears in the Minutes of Evidence (henceforth, RC, 1960C). In addition to this, the written evidence fills three volumes (henceforth, RC, 1960B). Many of the ideas and policy recommendations presented in the Radcliffe Report came directly from Richard Sayers,2 who had expressed them in several papers and essays, which shared a general Keynesian orientation. The Radcliffe Report was regarded as one of the most important documents in Keynesian monetary economics (Chick, 1973; Cobham, 1992) and Sayers and Cairncross were generally identified as Keynesians.3 More specifically, the Radcliffe Report combined the widely held Keynesian idea that monetary policy influences aggregate demand through long-term interest rates and investment, with the insistence—which we trace back to Sayers—on the blurred boundary between monetary and financial assets and the influence of interest rates on the supply of liquid funds. The Report’s main recommendation was that monetary policy should determine the entire term structure, including long-term interest rates, with the aim of regulating liquidity in a broad sense and, through that, aggregate demand. Technically, term structure control required the combination of ad hoc open market operations, Bank rate policy and other controls (including the enforcement of cash and liquidity ratios on clearing banks and the regulation of bank advances). Coordination between monetary policy and national debt management were also part of this strategy, which aimed at reconciling full employment, price stability and external equilibrium. The experimental nature of these recommendations, which came at the end of an extensive consultation process involving some among the most prominent economists of the time,4 makes their reconstruction of particular interest and motivates our research. The resonance of the Radcliffe Report, at the time of its publication and in subsequent years, until Monetarism and Rational expectations submerged the Keynesian framework on which it was based, provides further motivation in this respect, especially in connection with the ongoing research on post-Keynesian monetary theory5 and the renewed interest in Keynes and Keynesian economics after the recent crisis.6 In this paper, in particular, we focus on the contribution by Richard Kahn to the works of the Radcliffe Committee. This choice can be justified as follows. Making an exception to the rule of not mentioning individual contributions, the Radcliffe Report claims to ‘follow Professor Kahn … in insisting upon the structure of interest rates rather than some notion of the “supply of money” as the centre-piece of monetary action’ (RC, 1960A, §395). This statement makes the investigation of Kahn’s position, as presented in his Memorandum and oral evidence in front of the Committee, of particular relevance to contextualize the Report’s conclusions. As we show in our paper, while connections between Richard Kahn’s ideas and the Radcliffe Committee are much more tenuous than the explicit reference made in the Report might suggest, Kahn’s contribution is worth investigating per se, as it provides a conceptual framework where the main task of the central bank consists in determining interest rates at different maturity dates with the aim of achieving simultaneously external and internal equilibrium. It is our main contention that Kahn’s policy proposals, with their clear Keynesian orientation,7 are best understood in connection with his theory of liquidity preference, as put forth in Kahn (1954), and in particular with his sophisticated view of the role agents’ expectations play in determining interest rates. By combining Kahn’s evidence to the Radcliffe Committee with his theory of liquidity preference, our approach complements Koutsobinas (2007, p. 56), who builds on Kahn (1954) (and Tobin) to support ‘the adoption by monetary authorities of some form of “bills-bonds” approach that is consistent with the importance of long term interest rates for the effectiveness of monetary policy’.8 While the Radcliffe Committee espoused Kahn’s recommendations regarding the need to make the term structure, rather than money supply, the centrepiece of monetary action, it differed from Kahn on the causes of inflation and monetary policy transmission. More specifically, Kahn regarded inflation as mainly caused by wages running ahead of labour productivity, with aggregate demand influencing price dynamics principally, but not exclusively, through its impact on wage negotiations.9 Based on this framework, Kahn recommended to keep long-term interest rates as low as possible, in order to stimulate investments and productivity, leaving it to incomes and fiscal policy to restrain any concomitant inflationary pressure. By contrast, the Radcliffe Committee envisaged inflation as mainly determined by the direct pressure of aggregate demand on available resources, to be restrained by raising long-term interest rates with the aim of discouraging lenders, instead of borrowers, through the operation of a lock-in effect10 (on this, see Rousseas, 1998, p. 121). Our findings go some way towards explaining how it came about that, in spite of the prominence given to Kahn, the secondary literature about the Radcliffe Report has largely neglected and sometimes misrepresented Kahn’s position. In particular, our research suggests that Kahn’s contribution has remained almost completely hidden behind that of Richard Sayers,11 and that reference to ‘Professor Kahn’ was in part misleading, as it created the impression of a complete harmony between Khan’s ideas and those of the Committee. We show in the paper that on the crucial issue of the term structure of interest rate, the Report attributed to Kahn ideas that were not entirely his own and were at least in part those of Sayers. In particular, while Kahn claimed that monetary authorities could determine long-term interest rates by swaying expectations about them through consistent actions, Sayers did not share this belief entirely. This fact, coupled with official opposition to any impression of tampering with the gilts market, would eventually lead to emasculation of Kahn’s proposals. While early reviews of the Radcliffe Report reflect this attitude,12Tily (2006, 2007 and, in particular, 2009) emphasises Kahn’s insistence, in front of the Radcliffe Committee, on recommending the advantages of low long-term interest rates to promote investment and full employment, in line with Keynes’s own indications. Rousseas (1998, pp. 119–20, 124), instead, presents Kahn’s views in opposition to the conclusions reached by the Radcliffe Report, although as part of larger, and less personal, views of ‘the economists’, and not in some relation to Kahn (1954), as we do in this paper (see also Marcuzzo and Rosselli, 2017, pp. 710–11). On the whole, however, Kahn’s active interest in anti-inflationary policies during the 1950s and 1960s and his original views on monetary policy as related to his monetary theory have remained largely neglected. It is especially on this latter point that we focus our attention in this paper. Our main conclusion is that, albeit that Kahn’s impact on the Radcliffe doctrine was clearly not decisive, his views on the centrality of managing expectations about long-term interest rates based on his theory of liquidity preference mark a significant contribution to post-Keynesian monetary economics, one which appears particularly relevant today, in connection with the application of unconventional monetary measures. Based on these considerations, the structure of the paper is as follows. Section 2 briefly describes the historical and analytical background and the main conclusions of the Radcliffe Report. Section 3 discusses Richard Kahn’s contribution to the works of the Radcliffe Committee, focusing on his Memorandum and oral testimony. Section 4 elaborates on the linkages between Kahn’s ideas and his theory of interest rates and monetary policy as presented in his 1954 paper on liquidity preference, and explores the institutional dimension of these ideas in connection with the analysis presented by the Bank of England and the Treasury. Section 5 concludes the paper. 2. Radcliffe Report: background and main policy recommendations Between 1932 and 1951, Bank rate, the chief instrument of British monetary policy, had been held at 2%. This contributed to keeping long-term interest rates at a low level, with positive repercussions on debt servicing costs and aggregate demand. This policy, known as ‘cheap money’, ended in 1951 amidst rising inflationary pressures and a negative and deteriorating balance of payments.13 The crisis in the balance of payments, and the concomitant fall in official reserves, led to the reactivation of monetary policy in 1951. This was followed by declining inflation until 1953. Inflation picked up again in 1954 and 1955, and a second balance of payments crisis occurred in 1956. This triggered a second cycle of restrictive monetary policy, which led Bank rate to peak at 7% in September 1957. Initially, monetary restriction seemed to yield the expected results, leading inflation to fall and the balance of payments to improve. However, after declining to 3.1% in 1956, during 1957 inflation picked up again, reaching 4.6% during the 12 months.14 This led to the emergence of doubts about monetary policy and its effectiveness in maintaining price stability. As Cairncross (1996) reconstructs, these doubts were part of a deeper contrast between the Treasury and the Bank of England on how to control inflation. The Treasury preferred direct controls on bank advances and recommended the enforcement of variable cash and liquidity ratios on banks as a means to contain the growth of money and prices. The Bank was sceptical about the effectiveness of these controls and preferred constant cash and liquidity ratios.15 It also favoured the containment of public expenditure in view of restraining inflationary pressures.16 Opinions on monetary policy and the appropriate level of interest rates were mixed and shifting. In this context, the government appointed the Radcliffe Committee, at the suggestion of the Bank of England, with the mandate ‘to inquire into the working of the monetary and credit system, and to make recommendations’. More details on the actual mandate of the Radcliffe Committee can be gleaned from its report. It is clearly stated in §5 that when the ‘Chancellor of the Exchequer announced the decision to appoint the Committee, the problem that was pressing itself most urgently on his attention was that of controlling inflation’. The ‘persistence of inflationary pressure’ after the measures adopted over the last two years had ‘encouraged fears that the inflation was not just a temporary maladjustment … but a fundamental and perhaps inseparable concomitant of the economic circumstances and state of opinion which emerged after the war’ (RC, 1960A, §5). The Radcliffe Committee regards regulation of total demand for goods and services with the aim of achieving full employment, price stability and external equilibrium as the main purpose of monetary policy, together with fiscal policy and other measures (RC, 1960A, §§382–83). Traditionally, monetary policy would be expected to work through the control of money supply, as envisaged by quantity theorists.17 However, the existence of non-bank financial intermediaries, which provide the economy with money substitutes and credit (RC, 1960A, §§523–24), implies that spending is not limited by the amount of money in existence but rather by the ‘overall liquidity position’ or ‘whole liquidity position’ of the economy, defined as ‘the amount of money people think they can get hold of, whether by receipt of incomes (for instance from sales), by disposal of capital assets or by borrowing’ (§390). 18 The Committee rejects the view that monetary policy would become more effective by substituting ‘for the traditional control of the supply of money a complex of controls over an indefinitely wide range of financial institutions’ (§394). Direct controls can be useful in case of emergency but, ordinarily, the central bank should control the overall liquidity position by manipulating ‘the entire structure of interest rates’, as distinct from action ‘confined to the short-end of the market’ (§§393 and 394). In relation to this specific issue, the Committee follows ‘Professor Kahn … in insisting upon the structure of interest rates, rather than some notion of the “supply of money” as the centrepiece of monetary action’ (§395). The Committee regards changes in interest rates as affecting aggregate demand in two ways. One is the ‘interest incentive effect’ by which a change in interest rates can induce a change in the incentive to hold inventories and/or purchase fixed capital goods. The other is the ‘general liquidity effect’, which works as follows. ‘A movement of interest rates implies significant changes in the capital values of many assets held by financial institutions; a rise in rates makes some less willing to lend because capital values have fallen, and others because their own interest structure is sticky. A fall in rates, on the other hand, strengthens balance sheets and encourages lenders to seek new business’ (§§393 and 394). In general, and coherently with the view expressed in Sayers (1958, ch. 7), the Committee was sceptical about the relevance of the first effect and more confident of the second (on this, see in particular RC, 1960a, ch. VI).19 Manipulation of the term structure requires monetary authorities to take a view on the appropriate level of short- and long-term interest rates and enforce this view by a combination of ad hoc open market operations, Bank rate policy and other controls. It also requires close coordination between monetary policy and national debt management (§§533–35), effective communication (§499) and wider access to economic statistics (§865), which together contribute to influencing ‘expectations and confidence’, regarded as prime determinants of interest rates. The Committee’s position, as summarised above, seems to envisage the possibility of reactivating monetary policy based on innovative conceptual premises, relying on interest rate control and synergy between monetary action and national debt management within a Keynesian policy aimed at full employment. As to that, the Radcliffe Report incorporated in the official view new and unconventional elements in monetary policy that had begun to circulate over the last decade, especially on the concept of the liquidity position being the new target of monetary policy. It also took into account elements of revision that had been introduced in economic theory since the end of the 1930s, in particular by the Oxford Economists Research Group. In his contribution to Oxford Studies in the Price Mechanism, edited by A. W. S. Andrews and T. Wilson in 1951, Sayers offered a survey of the literature on ‘The Rate of Interest as a Weapon of Monetary Policy’, in which he looked back at ‘[t]he belief that changes in the rate of interest powerfully influence economic activity’. This doctrine—Sayers argued in introducing his survey—had ‘reached its high-water mark in Keynes’s Treatise on Money’ (Sayers, 1951, p. 1). Afterwards came the inquiry by the Oxford Economists Research Group on the relationship between the rate of interest and investment decisions, the results of which were published in the earlier issues of Oxford Economic Papers.20 The ‘Oxford inquiry’ reached the conclusion that changes in the rate of interest had almost no effect on investment decisions. Although based on a very narrow basis of empirical evidence, the inquiry had a large impact on the profession, as Sayers’s survey clearly shows. In this paper (written early in 1949), Sayers remained relatively prudent on this issue, but he was nonetheless ‘astonished at the confidence and simplicity of Keynes’s argument’ (1951, pp. 1–2). In 1952, the Economic Journal published Kahn’s review of Oxford Studies in the Price Mechanism (Kahn, 1952B). While commenting favourably on Sayers’s survey of the literature, Kahn sounded very sceptical of the results of the ‘Oxford inquiry’ (as well as of Hall and Hitch’s full cost article of 1939, which was also reprinted in the book of 1951). By contrast, the fourth edition of Modern Banking (ch. 7) and §§386–87 of the Radcliffe Report bear witness to the fact that Sayers had eventually embraced the new ideas, strongly rejecting the theory that the price at which money could be borrowed was a decisive factor in determining the level of investment in favour of the idea that the decisive factor was the ‘availability’ of credit. On the other hand, the endorsement of a theory of interest rates as conventional phenomena and the parallel rejection of a stable relationship between money supply, liquidity preference and interest rates constitute as many elements of continuity between the theory underlying the Radcliffe Report and Richard Kahn’s ideas on the subject, as he presented them in front of the Radcliffe Committee. 3. Richard Kahn’s contribution to the works of the Radcliffe Committee In the opening section of his Memorandum to the Radcliffe Committee (Kahn, 1960), Richard Kahn reviews the state of the British economy in the 1950s, focusing on the impact of monetary and fiscal policy on economic activity, inflation and the balance of payments. Two aspects of his analysis are of particular interest in the context of the present study: his theory of inflation and his considerations on monetary policy and the supply of money. Kahn regards the level of prices as determined by three interacting factors: money wages, productivity and the level of aggregate demand. The price level raises to the extent that nominal wages—determined through negotiations between trade unions and employers—rise faster than productivity.21 Under normal conditions, demand influences wages directly and the price level indirectly, unless it is so high as to cause severe labour shortages, which drive nominal wages above negotiated levels. It is the rise in wages which is the cause of prices rising. In that sense all price inflations are ‘cost inflations’ (or wage inflation). But the speed at which wages, and therefore prices rise is likely in some degree to depend on demand. In that sense there is nearly always some measure of ‘demand inflation’. (Kahn, 1960, p. 138) Kahn’s position on inflation goes together with his refutation of the quantity theory of money as a causal process, a long-held position, as Marcuzzo (2002) among others recalls. In the Memorandum, Kahn reiterates that no direct link exists between money and the general level of prices. He also stresses that money velocity, far from being constant, adjusts to changes in money supply and subsequent changes in economic activity.22 A lower quantity of money causes interest rates to rise in connection with banks reducing their holdings of securities and their advances to private investors and industry. This leads security prices and private investment to fall, reducing inflationary pressures at the cost of rising unemployment. This side effect, together with the concomitant reduction in aggregate investment, is one of the principal reasons that led Kahn to be wary about monetary policy restriction as a cure against inflation.23 Following on these considerations, Kahn opines that the whole structure of interest rates matters for monetary transmission. Short-term interest rates count mainly for their influence on the international capital movements and the exchange rate while long-term interest rates matter principally for their influence on investment and domestic demand.24 As such, monetary authorities should adjust the former in view of balance of payments considerations and the latter to balance domestic demand and supply, bearing in mind connections between internal and external equilibrium.25 The Memorandum, while not entering into the analysis of the relationship between short- and long-term interest rates, treats their separate, but coordinated, determination as a technical problem, whose solution requires ad hoc operations. Within wide limits it is possible to achieve any desired structure of interest rates by a suitable combination of monetary policy with management of the National debt. The Exchequer, by issuing short-dated securities in the place of long-dated, or vice versa, can secure the desired shift in relative rates of interest against the background of a monetary policy which operated on rates in general. Of course, the interest charge on the National debt will thereby be affected and this consideration will play an important part in determining the objectives of policy. (Kahn, 1960, pp. 148–49) The Minutes of evidence (RC, 1960C) clarify the conceptual framework underlying Kahn’s recommendations. The questions put to Kahn by the members of the Radcliffe Committee revolve around the link between money supply and interest rates and the impact of changes in interest rates on investment and domestic demand, leaving inflation out of the picture. In oral evidence, Kahn reiterates his position about the primacy of interest rates over the supply of money as a monetary policy instrument capable of influencing spending decisions. Money is a stock, it is not a flow. The things which influence people’s conduct are flows, profits, incomes, that sort of thing: things you measure as rates per unit of time; not a stock which is simply a matter of how at any particular moment people decide to dispose of their personal wealth and companies of their liquid assets. (Kahn, RC, 1960C, Q. 10986) If a stock matters for business decisions, that is overall liquidity, rather than bank deposits only. In a developed financial system, companies have many alternative financing sources at their disposal (banks, non-bank financial institutions, bill market, bond markets, Stock Exchange), which monetary authorities must control simultaneously in order to regulate corporate spending decisions. This requires coordination between Bank rate policy, open market operations, minimum cash and liquidity ratios and direct regulation of bank advances,26 capital issues and investment allowances (on this, see also Kahn, 1960, pp. 150–52). In this context, any residual significance of money supply control per se rests on its relationship with interest rates and investment, which Kahn regards as substantial although slow in coming into effect and unpredictable in its results (Kahn, 1960, p. 158). Moreover, Kahn considers it impossible to quantify a priori the level of money supply required to achieve the desired term structure of interest rates, as the latter depends substantially on expectations of investors and their reaction to monetary decisions: If the market likes to help by altering its expectations, because it thinks that this is the way the cat is jumping, then the authorities will not have to do much. If the market proves very obstinate and does not believe that they are going to succeed in doing what they are trying to do, then they will have to do much more. (Kahn, RC, 1960C, Q. 10988) Based on these considerations, which reflect an approach heedful of liquidity preference and national debt management constraints, it is natural for Kahn to indicate the term structure of interest rates rather than the quantity of money as the centrepiece of monetary actions. As seen above, Kahn sees no obstacle, within very broad limits, to the possibility that monetary authorities may achieve their objectives, provided they are ready to operate both in money and capital markets, combining open market operations à outrance (Kahn, 1952A, p. 148) with different issuing techniques. This requires firm ideas about the appropriate structure of interest rates and consistent actions, carried out by skilled technicians, aimed at altering the relative balance between bills and bonds in order to achieve the desired term structure. Kahn insists on the fact that as long as markets perceive that the government is determined, ‘It is very difficult to conceive of a situation in which it would not be possible to arrive at any particular price, provided that, as I did say, one is considering this within reasonable limits. … [O]nce the market realizes that the authorities are serious they will dash in and help the authorities’ (Kahn, RC, 1960C, Q. 10988, 10993). Naturally, this requires that the authorities are ready to accept any variation in the amount of money and/or bills and bonds which may be required in order to achieve the desired term structure. Kahn’s recommendations to the Radcliffe Committee bear close resemblance to Keynes’s own proposals on national debt management as presented in the General Theory and again in 1945 to National Debt Enquiry (NDE).27 As Tily (2006) and Peden (2004, pp. 330–31) reconstruct, in the documents prepared for the NDE Keynes regarded it possible for monetary authorities to set short- and long-term interest rates at any desired level, provided they were ready to combine appropriate issuing techniques for gilts (tap issuance) with an elastic supply of Treasury bills. Keynes also favoured issuance of a wider range of bonds to meet different savings requirements.28 The NDE first draft report (dated 15 May 1945) quoted extensively from p. 203 of Keynes’s General Theory to explain the conventional element of the long-term rate of interest and the need to make it as clear as possible that the authorities are strongly committed to their target. It also emphasized that the tap issuance policy fitted in well with a government that was, in most of the cases, a net borrower. Kahn’s views are consistent with these indications, even if differences exist between the two approaches. First, the NDE report insists on the need to pose no limits to the possibility for the Exchequer to issue Treasury bills, something which occurred on an unprecedented scale in the 1940s. It also draws attention to the inflationary impact of this form of financing, which increases the liquid assets of banks and their ability to lend, and on how to limit it (e.g. through Treasury Deposits Receipts, TDR).29 Kahn seems not to be worried about this possibility, coherently with his refutation of any direct causal link from aggregate liquidity to inflation. Second, Kahn appears to share the confidence shown in the NDE report as to the possibility of modifying prevailing market conventions and long-term interest rates with them. As it seems, he was also more confident than Sayers (1958, pp. 172–76), who did not deny that changing the long-term expectations of the public was possible, but warned that it was also very difficult to do so. With the hindsight provided by Hugh Dalton’s recent experiment of ‘ultra-cheap money’, that Sayers (1958, p. 175) took as an example of the potentially disastrous consequences in case of failure, this came not unexpected.30 The determination of interest rates at different maturity dates raises a series of interesting analytical problems concerning the impact of monetary policy on investors’ expectations and portfolio allocation decisions, in the presence of uncertainty and imperfect substitutability between money, bills and bonds. Exploring how Kahn addressed these problems in the context of his own theory of liquidity preference, as presented in 1954, contributes to understanding the theoretical underpinnings of his recommendations to the Radcliffe Committee. 4. Kahn’s theory of interest rates and liquidity preference One of the key ideas Kahn presents in his 1954 Notes on Liquidity Preference is that the long-term interest rate depends on expectations about the long-term interest rate itself rather than about the short-term interest rate, as theorised by John Hicks and others since the 1930s.31 When a bill is bought rather than a bond the only relevant expectation determining the decision is what the bond rate will be when the bill matures. That expectation is certainly related to the expectation of what the bill rate will itself be at that same date. ... All this is, however, a very different thing from saying that the bond rate depends on expectations about the future of the bill rate itself, rather than of banking and monetary policy generally. Professor Hicks tries to reconcile the two different views, but only by appealing to the necessity for consistency between the two sets of expectations, which brings us back to an imaginary world of expectations held with unanimous and complete conviction. (Kahn, 1954, p. 78) Differently from Hicks and his school, Kahn considers knowledge about the future level of interest rates as highly uncertain, subject to doubt and disagreement, with different persons formulating different expectations about the future course of interests, which they hold with different degrees of conviction and different sensitivity to risk. This provides the basis of Kahn’s analysis of the precautionary demand for money, which he conducts focusing on the case of an investor choosing between money and long-term securities (fixed-rate irredeemable bonds for simplicity), assuming Treasury bills away. Kahn is aware that this assumption involves a serious departure from reality but, if one agrees with him that decisions to go long rather than short depend primarily on expectations about what the long-term interest rate will be when the bill matures, and only secondarily on expectations about the future bill rate, nothing fundamental is lost by focusing on money and bonds only. As Kahn adds, ‘[o]f course the short-term interest rate will still remain lurking in the picture in a certain sense, in the guise of expectations about the behaviour of the long-term rate itself, but these are subjective and personal to each individual concerned and are obscured by risk and uncertainty’ (on this, see Kahn, 1954, p. 79).32 This investor bases her decision on two elements: the current long-term interest rate i, and the rate at which she expects i to rise (or fall) over the relevant investment horizon. On the assumption that the coupon paid to bond holders is equal to 1 euro, the current price of the bond p is equal to the reciprocal of the long-term interest rate (p = i– 1). Defining v as the expected price of the bond at maturity of the first coupon, the investor will be indifferent between bonds and money, whose return is equal to zero, when 1+v–p=0. (1) Dividing both sides by p and isolating i, equation (1) becomes i=v(1/v–1/p), (2) where v (1/v – 1/p) = g represents the expected rate of variation in the long-term rate of interest. As long as v is unique and held with complete certainty, the investor’s speculative demand for money will be equal to zero if g < i and to her entire portfolio if g > i. In this case, the precautionary motive is inoperative and only marginal investors, whose expectations are consistent with equation (2), will hold both money and bonds. On the assumption of heterogeneous agents—each of them holding unique expectations with absolute certainty—the aggregation of individual positions yields the market demand for securities and its obverse, the speculative demand for money. Speculative demand for money, which varies inversely with i, together with that part of money supply which agents hold for investment purposes, determines the equilibrium level of the long-term interest rate.33 If individuals hold expectations with less than perfect confidence, however, uncertainty enters into the analysis. In this context, precautionary demand for money becomes relevant and all investors hold money and securities simultaneously, in variable proportions depending on their aversion to income and capital risk34 and on their lack of conviction about their own expectations. In this case, speculative and precautionary demand for money become almost indistinguishable from one another, introducing an element of indeterminacy in the relationship between money and interest rates. As Dardi (1994, p. 102) observes, this indeterminacy makes it impossible to know in advance the measure of the variation in money supply, required to bring about a certain variation in interest rates, just as argued by Kahn in his evidence to the Radcliffe Committee. Kahn does not exclude that, even assuming uncertainty, changes in money supply may cause the long-term interest rate to move in the opposite direction, but the effect is unpredictable, due to heterogeneity in investors’ reactions. Observing a change in money supply, some investors will not modify the composition of their portfolios at all, either because they have no clue about how that change may affect interest rates or because precaution overrides the speculative motive completely. Other investors will revise their expectations but will not re-allocate their portfolios as long as they are not sufficiently confident about their views. Others still will react very quickly, causing brusque oscillations in interest rates if their number prevails in the market. The coexistence of different groups of investors, ranging from ‘bankers’ to ‘widows and orphans’,35 shifting in terms of size and composition and depending on current circumstances, magnifies the indeterminacy of the relationship between liquidity preference and interest rates (on this aspect in particular, see Kahn, 1954, pp. 81–90), weakening the case in favour of a quantitative approach to monetary policy, as supported by prominent opinion in front of the Radcliffe Committee (see footnote 17). To the extent that liquidity preference is, a priori, undetermined, targeting money supply would provoke permanent oscillations in interest rates, with negative repercussions on investors’ expectations and the economy at large. If, on the contrary, monetary authorities put interest rates at the centre of their strategy, this will reduce uncertainty and stabilize expectations, enabling them to achieve the desired term structure. Doing so, evidently, turns money supply into an endogenous variable—a concomitant factor, as Kahn put it to the Radcliffe Committee—best left to itself,36 which ceases to be a causal influence on economic activity and has no obvious connection with price stability. Kahn regards expectations of price stability as essential for effective policy action, which combines incomes policy, aimed at maintaining nominal wages in line with labour productivity, with adequate aggregate demand management: The really important thought which underlies advocacy of price stability relates to the harm which is done to the economic system once it becomes generally recognized as a certainty that the price level will go on progressively rising without interruption and without limit. It is continuous and complete confidence in that expectation which does the harm. (Kahn, 1960, pp. 144–45) Monetary restriction, while technically possible, is inferior to incomes policy in restraining inflation as it reduces investment, productivity and productive capacity and may increase unemployment in the absence of efficient coordination between trade unions and manufacturers associations (e.g. §§53–54). Moreover, the apparent political neutrality of monetary actions, if compared to fiscal policy, builds an upward bias into long-term interest rates, which may damage the economy’s growth prospects in the medium to long run. These positions, coupled with his theory of liquidity preference, lead Kahn to refute the idea of controlling money supply in view of achieving price stability. In the same vein, he observes that it is not the overall budget deficit that matters for the control of inflation but the instruments used to finance it, their maturity and the size of their respective markets, as they impact on the shape of the term structure. Kahn’s recommendations gave rise to not only substantial analytical questions but also institutional and practical problems. In oral evidence, Kahn expressed his confidence in the ability of the Treasury and the Bank of England to achieve, within very wide limits, any desired structure of interest rates. This would require cooperation between the two institutions and deliberate if gradual actions to influence market expectations in the desired direction. However, neither cooperation nor deliberate action could be taken for granted. Indeed, official reluctance to countenance sharp fluctuations in government bond prices, combined with an institutional set-up designed to guarantee full coverage of the Exchequer financing needs, posed a formidable obstacle to any concrete attempt to vary long-term interest rates in a deliberate fashion. The Bank of England, for one, opposed the idea that monetary policy should influence directly the level of long-term interest rates, on substantial rather than technical grounds. In the Bank’s opinion, systematic intervention in the market for government bonds, different from simply following the market, would risk undermining the confidence in the government’s credit, particularly in view of the non-professional nature of many bondholders. As the Chief Cashier of the Bank pointed out to the Committee, It would, of course, be possible to force down the price of gilt-edged securities by direct sales on the market. But we take the view that if the Government, or the Bank, as the Government issuing bankers, were deliberately to sell the Government’s own securities in order to drive down the market price, it would greatly damage the Government’s credit. (RC, 1960C, Q. 1762)37 The existence of a close connection between monetary policy and national debt management was further emphasised by the Treasury: ‘[t]he choice and methods used to put the choice into effect depend upon considerations of National Debt and monetary policy’ (RC, 1960B, part II, n. 2, §21). On this latter point, the Treasury made a step further, considering what might happen if the government decided to change the whole structure of interest rates, and not only the short-term rate. The Treasury admitted that ‘[t]he Government can always, if it wishes, depress the price of gilt-edged securities, simply by placing a quantity of its own securities on the market’ (§35). However, the Treasury also pointed out that ‘[b]oth … methods of limiting people’s ability or willingness to raise money—control over the money supply through control over the supply of liquid assets, and the operation on security prices—are … subject today to much greater limitations than they have been in earlier periods’ (§37). These limitations came mainly from the need to roll over a significant part of Britain’s huge post-war public debt,38 and from the significant rise in capital expenditure, linked to reconstruction. Coupled with resurgent inflation, this left the Treasury and the Bank of England with little room for manoeuvre in issuance policy, which mainly aimed at replacing Treasury bills with short-term gilts (funding).39 Technically, the presence of the Government broker, trading actively in all maturities in the gilt-edged market, made it indeed possible to influence the entire term structure along the lines suggested by Kahn. In practice, however, an institutional set-up designed to secure finance for the Exchequer, combined with the reluctance of monetary authorities to provoke sharp rises in long-term interest rates and the possibility for private investors to abstain from rolling over maturing bonds, created a noticeable difficulty in following Kahn’s suggestions. By instructing its Issue Department to take up all the government bonds, which the public had not subscribed initially, the Bank of England provided an additional guarantee to the Exchequer.40 When absorbing government bonds in this way, however, the Issue Department was obliged to reduce its lending to the Exchequer by reducing its Treasury bill holdings. This, in turn, forced the Exchequer to issue more Treasury bills in the market, expanding liquidity. For the same reasons, bond sales by the Government Broker had the effect of putting the liquidity of the banking system under pressure. In this sense, the amount of liquid assets depended ultimately on the public’s willingness to absorb government bonds and on their expectations.41 To the extent that the Treasury and the Bank of England took those expectations as exogenously given, and chose to auction off all types of securities rather than resort to the tap issuance technique, this might explain their reluctance to adopt the measures recommended by the Radcliffe Committee: [T]he authorities have not been prepared to force interest rates on the longer bonds upwards in order to tempt holders of short bonds to switch to long bonds. … They have held the view that sales could only be made in any quantity on a rising market. … Their view that demand could not be stimulated by dropping prices—a view not easy to accept—was based mainly on the belief that the market for gilt-edged securities is dominated by expectations … and that it is the expectation of what prices will be tomorrow or next week, and not expectation of next year’s prices, that matters. (RC, 1960A, §551) The members of the Radcliffe Committee were not convinced by this interpretation and went so far as to lament the almost fatalistic attitude of monetary authorities in the face of movements of long-term interest rates (RC, 1960A, §552).42 In connection with this, echoing Kahn, they also claimed that Our view is that expectations have been overrated [by the authorities] as independent market forces, and that at time the influence of the authorities themselves on those expectations has been correspondingly underrated. (RC, 1960A, §563) In spite of these considerations, the members of the Radcliffe Committee could not ignore the position of the Treasury and the Bank of England and their dislike of variable long-term interest rates. This led the Committee to adopt a compromise position, which we can summarize as follows. Monetary policy and national debt management together can indeed determine the term structure of interest rates, by combining ad hoc open market operation, Bank rate policy and other controls. Financial stability and other considerations, however, make it advisable not to change the long-term interest rate too often and/or too abruptly. 5. Conclusions At the time of its publication and in subsequent years, the Radcliffe Report attracted considerable attention by supporters and detractors alike. This was true in particular of the idea that monetary policy should determine the entire term structure of interest rates with the aim of regulating the liquidity position of the economy and aggregate demand. As we reconstruct in this paper, the Radcliffe Committee shared Kahn’s indication to make the term structure of interest rates, rather than money supply, the centrepiece of monetary action. At the same time, we identify significant differences between the Radcliffe Report indications, which largely reflected the opinions of Sayers, and Kahn’s ideas. In particular, while Kahn was convinced that monetary authorities could determine the entire term structure and thereby influence aggregate demand through the demand for credit, Sayers did not share these beliefs entirely. On the one hand, he underlined how difficult it might be for the authorities to sway expectations about interest rates, even if he did not exclude this possibility altogether. On the other hand, he opined that the main influence of interest rates on aggregate demand worked through the supply, rather than the demand, for liquid funds. These differences of opinion, by themselves, go some way in reducing Kahn’s actual impact on the Radcliffe Report. But we think that the decisive factor that led to the emasculation of his ideas coincided with the opposition of the monetary authorities to countenance sharp policy-driven fluctuations in long-term interest rates. This opposition, combined with an institutional set-up designed to guarantee the exchequer financing needs in all cases, posed, in our opinion, the main obstacle to the full application of Kahn’s recommendations to the Radcliffe Committee. These recommendations are best understood in connection with Kahn’s views on interest rates and monetary policy as expressed in his 1954 paper on liquidity preference. The characterisation of different types of investors and Kahn’s insistence on the connection between precautionary and speculative motives led him to reject the idea of a stable well-defined liquidity function in view of a more problematic approach, where expectations about interest rates play a dominant role in the determination of the term structure. This introduces an element of indeterminacy in the setting of interest rates, which the Report’s insistence on the blurred distinction between money (notes and bank deposits) and near-money magnified. On the eve of the recent crisis, Tily (2006, p. 657) made this comment: ‘[w]hile the role of expectations and uncertainty in the determination of the rate of interest has been restored by Chick (1983), this restoration has not led to a recognition that expectations can be managed and the rate of interest brought under control’ (Tily, 2006, p. 657). We have shown in this paper that something similar happened during the workings of the Radcliffe Committee, as a result of the institutional and political obstacles which were in the way of implementing term-structure control, be it for its influence on the supply of liquid funds or on their demand. The Committee tried to overcome these obstacles by adopting a compromise position whereby monetary authorities should set the long-run path of interest rates without varying them too often and too abruptly. Evidently, this compromise took off much of the innovative edge from the Radcliffe Committee’s proposals, contributing to that impression of confusion and wooliness lamented by many contemporary observers. From this we conclude that, albeit that Kahn’s impact on the Radcliffe doctrine was clearly not decisive, his views on the centrality of managing expectations about long-term interest rates based on his theory of liquidity preference mark a significant contribution to post-Keynesian monetary economics, one which appears particularly relevant today, in connection with the application of unconventional monetary measures. The authors wish to thank Sylvie Diatkine, Ivo Maes, George Peden, Jan Toporowski and two anonymous referees for their useful comments on earlier draft of the paper. The usual disclaimers apply. The paper is the first product of a research project on Unconventional monetary policy ante litteram: the Radcliffe Committee and the debate on liquidity and the long-term interest as a monetary policy instrument. Financial support was provided by the ESHET (European Society for the History of Economic Thought) through the ‘ECB Grants’ programme 2015, sponsored by the European Central Bank. Carlo Cristiano was responsible for writing sections 1 and 2; Paolo Paesani for sections 3 and 4. The concluding Section 5 is a joint contribution by the two authors. Bibliography Allen , W. A . 2014 . Monetary Policy and Financial Repression in Britain, 1951–59 , London , Palgrave Macmillan Google Scholar CrossRef Search ADS Allen , W. A . 2016 . The British attempt to manage long-term interest rates in 1962–1964 , Financial History Review , vol. 23 , no. 1 , 47 – 70 Google Scholar CrossRef Search ADS Alon , T. and Swanson , E . 2011 . Operation Twist and the effect of large-scale asset purchases , FRBSF Economic Letter, vol. 13 Artis , M. J . 1961 . Liquidity and the attack on the quantity theory , Bulletin of the Oxford University Institute of Economics & Statistics , vol. 23 , 343 – 66 Google Scholar CrossRef Search ADS Artis , M. J . 1965 . Foundations of British Monetary Policy , Oxford , Basil Blackwell Cairncross , A . 1991 . Richard Sidney Sayers, 1908–1989 , Proceedings of the British Academy , vol. 76 , 545 – 61 Cairncross , A . 1996 . Prelude to Radcliffe: monetary policy in the United Kingdom, 1948–57 , in Cairncross , A ., Economic Ideas and Government Policy , London and New York , Routledge Google Scholar CrossRef Search ADS Cairncross , A . 1999 . Diaries: The Radcliffe Committee and the Treasury 1961–64, Institute of Contemporary British History, Cairncross Treasury Series , London , University of London Capie , F . 2010 . The Bank of England: 1950s to 1979 , Cambridge , Cambridge University Press Google Scholar CrossRef Search ADS Cardim de Carvalho , F. J . 2010 . Uncertainty and money: Keynes, Tobin and Kahn and the disappearance of precautionary demand for money from liquidity preference theory , Cambridge Journal of Economics , vol. 34 , no. 4 , 709 – 25 Google Scholar CrossRef Search ADS Chick , V . 1973 . The Theory of Monetary Policy , London , Gray-Mills Chick , V . 1983 . Macroeconomics After Keynes , Oxford , Philip Allan Chick , V. and Tily , G . 2014 . Whatever happened to Keynes’s monetary theory ?, Cambridge Journal of Economics , vol. 38 , 681 – 99 Google Scholar CrossRef Search ADS Cobham , D . 1992 . Radcliffe Committee , pp. 265 – 66 in Eatwell , J. , Milgate , M. and Newman , P . (eds), The New Palgrave Dictionary of Money and Finance , London, Macmillan. Committee on the Working of the Monetary System (Radcliffe Committee) . 1960A . Report, Cmnd. 827 , London , HMSO Committee on the Working of the Monetary System . 1960B . Principal Memoranda of Evidence , 3 vols., London , HMSO Committee on the Working of the Monetary System . 1960C . Minutes of Evidence , London , HMSO Dardi , M . 1994 . Kahn’s theory of liquidity preference and monetary policy , Cambridge Journal of Economics , vol. 18 , 91 – 106 Fellner , W. , Gilberts , M. , Hansen , B. , Kahn , R. F. , Lutz , F. and de Wolff , P . 1961 . The Problem of Rising Prices , Organization for European Economic Co-Operation Fforde , J . 1992 . The Bank of England and Public Policy: 1941–1958 , Cambridge , Cambridge University Press Goodhart , C . 2012 . Monetary policy and public debt , Financial Stability Review , vol. 16 , 123 – 30 Gurley , J. G . 1960 . The Radcliffe Report and evidence , American Economic Review , vol. 50 , 672 – 700 Howson , S . 1987 . The origins of cheaper money, 1945–7, Economic History Review , New Series , vol. 40 , no. 3 , 433 – 52 Howson , S . 1988 . Cheap money and debt management in Britain 1932–51 , in Cottrell , P. L. and Moggridge , D. E . (eds), Money and Power: Essays in Honour of L. S. Pressnell , Basingstoke , Macmillan Google Scholar CrossRef Search ADS Howson , S . 1993 . British Monetary Policy, 1945–1951 , Oxford , Oxford University Press Kahn , R. F . 1952A . Comment on ‘Monetary policy and the crisis’ , by C. Kennedy, Bulletin of the Oxford University Institute of Statistics , vol. 14 , 147 – 53 Kahn , R. F . 1952B . Oxford studies in the price mechanism , Economic Journal , vol. 62 , no. 245 , 119 – 30 Google Scholar CrossRef Search ADS Kahn , R. F . 1954 . Some notes on liquidity preference , Manchester School of Economic and Social Studies , vol. 22 , 229 – 57 ; pp. 72–96 in Kahn, R. F. (1972), Selected Essays on Employment and Growth, Cambridge, Cambridge University Press Google Scholar CrossRef Search ADS Kahn , R. F . 1960 . Memorandum of evidence submitted to the Radcliffe Committee , pp. 124 – 52 in Kahn, R. F. (1972), Selected Essays on Employment and Growth , Cambridge , Cambridge University Press Kaldor , N . 1960 . The Radcliffe Report , Review of Economics and Statistics , vol. 42 , 14 – 19 Google Scholar CrossRef Search ADS Koutsobinas , T. T . 2007 . Monetary policy formation at the long-term margin: a Kahn-Tobin framework , pp. 43 – 58 in Arestis , P. , Hein , E. and Le Heron , E . (eds), Aspects of Modern Monetary and Macroeconomic Policies , London , Palgrave Google Scholar CrossRef Search ADS Kregel , J. A . 2013 . Was Keynes’s monetary policy à outrance in the treatise, the model for ZIRP and QE ?, pp. 135 – 48 in Hirai , T. , Marcuzzo , M. C. and Mehrling , P . (eds), Keynesian Reflections , New Delhi , Oxford University Press Laidler , D . 1989 . Radcliffe, the quantity theory and monetarism , in Cobham , D. , Harrington , R. and Zis , G . (eds), Money, Trade and Payments—Essays in Honour of Dennis Coppock , Manchester , Manchester University Press Laidler , D . 2007 . Monetary policy and its theoretical foundations , pp. 14 – 31 in Mayes , D. G. and Toporowski , J . (eds), Open Market Operations and Financial Markets , London and New York , Routledge Google Scholar CrossRef Search ADS Marcuzzo , M. C . 2002 . The demise of the quantity theory of money , History of Economic Ideas , vol. 10 , 49 – 62 Marcuzzo , M. C. and Rosselli , A . 2017 . Richard F. Kahn (1905–1989) , pp. 705 – 22 in Cord , R. A . (ed.), The Palgrave Companion to Cambridge Economics , vol. 2 , Basingstock, Palgrave Macmillan. Google Scholar CrossRef Search ADS O’Hara , G . 2012 . Governing Post-War Britain: The Paradoxes of Progress, 1951–1973 , London , Palgrave Macmillan Paesani , P . 2017 . On the relevance of Keynes to the contemporary world: post-crisis views on economic policy, capitalism and international monetary reform , Annals of the Fondazione Luigi Einaudi , vol. 51 , June, 249 – 68 Peden , G. C . (ed.) 2004 . Keynes and His Critics: Treasury Responses to the Keynesian Revolution: 1925–1946 , Oxford , Oxford University Press for the British Academy Robinson , J. V . 1951 . The Rate of Interest and Other Essays , London , Macmillan , pp. 3 – 30 Rousseas , S . 1998 . Post-Keynesian Monetary Economics , 3rd ed ., London , Palgrave Macmillan Google Scholar CrossRef Search ADS Sawyer , M . 2002 . Economic policy with endogenous money , in Desai , M. , Dow , S. and Arestis , P . (eds), Essays in Honour of Victoria Chick , vol. 1 , London , Routledge Sayers , R. S . 1951 . The rate of interest as a weapon of economic policy , pp. 1 – 16 in Wilson , T. and Andrews , P. W. S . (eds), Oxford Studies in the Price Mechanism , Oxford , Clarendon Press Sayers , R. S . 1955 . The determination of the volume of bank deposits: England 1955–56 , Banca Nazionale del Lavoro Quarterly Review , reprinted in Central Banking after Bagehot, pp. 92 – 107 Sayers , R. S . 1958 . Modern Banking , 4th ed ., Oxford , Clarendon Press Sayers , R. S . 1960 . Monetary thought and monetary policy in England , Economic Journal , vol. 70 , 710 – 24 Google Scholar CrossRef Search ADS Seldon , A . (ed.) 1960 . Not Unanimous: A Rival Verdict to Radcliffe’s on Money , London , Institute of Economic Affairs Tily , G . 2006 . Keynes’s theory of liquidity preference and his debt management and monetary policies , Cambridge Journal of Economics , vol. 30 , no. 5 , 657 – 70 Google Scholar CrossRef Search ADS Tily , G . 2007 . Keynes’s General Theory, the Rate of Interest and Keynesian Economics: Keynes Betrayed , London , Palgrave Macmillan Google Scholar CrossRef Search ADS Tily , G . 2009 . The policy implications of the general theory , Real-World Economics Review , vol. 50 , 16 – 33 Tobin , J . 1958 . Liquidity preference as behavior towards risk , Review of Economic Studies , vol. 25 , 65 – 86 Google Scholar CrossRef Search ADS Turner , P . 2013 . ‘ Benign Neglect of the Long-Term Interest Rate ’, BIS Working Paper Series no. 403 Turner , P . 2014 . Is the long-term interest rate a policy victim, a policy variable or a policy lodestar ?, pp. 19 – 55 in Chadha , J. S. , Durré , A. C. J. , Joyce , M. A. S. and Sarno , L . (eds), Developments in Macro-Finance Yield Curve Modelling , Cambridge , Cambridge University Press Footnotes 1 On these aspects, focusing on the UK and the USA, see Allen (2016), Alon and Swanson (2011), Turner (2013, 2014) and the references cited therein. 2 Cairncross (1991, p. 550) traces back the ideas that Sayers would later reproduce in the Radcliffe Report far back in time to the first edition of Modern Banking (1938). The fourth edition of this book saw light in 1958, during the period in which Sayers was an active member of the Committee. See also Sayers (1955, 1960). 3 On Cairncross’ Keynesianism, see O’Hara (2012, ch. 7); on the Sayers’ association with Keynes, see Cairncross (1991). 4 ‘The Radcliffe Report [1959] was a document of tremendous importance to the theory of monetary policy. … The submitted evidence and answers to their questions embody the thinking, in all its diversity, of a wide range of economists, bankers, and administrators on the subject at the time’ (Chick, 1973, p. 58). 5 In particular, we were inspired by Tily (2006, 2007) and Chick and Tily (2014), who insist on the relative neglect of Keynes’s theory of liquidity preference by modern and contemporary reappraisals of Keynesian economics. 6 For a survey of post-crisis contributions on Keynes and Keynesian economics, see Paesani (2017). 7 In chapter 37 of his Treatise on Money, Keynes proposed open market operations to be carried out ‘à outrance’ in order to force down interest rates at all maturities as an extreme anti-cyclical monetary measure in the context of the slump of 1930. The idea of managing the entire structure was then revived in the General Theory and during the National Debt Enquiry of 1945 (more on this below, in Section 3). 8 Cardim de Carvalho (2010) explores how Kahn and Tobin address the role of risk and uncertainty in determining the precautionary demand for money, seen as a key component of liquidity preference, whose fluctuations influence employment and monetary transmission. 9 For a full exposition of this doctrine, see Fellner et al. (1961, ch. 5), to which Kahn was one of the contributors. 10 The lock-in effect, also known as Roosa effect, consists in the reluctance of banks and other lending institutions, fearful of capital losses, to liquidate securities in view of expanding loans and anticipations in a context of rising interest rates. 11 In his obituary of Sayers, Cairncross (1991, pp. 553–54) would later confirm that the sections of the Report which deal with economic theory had been drafted by Sayers. Cairncross also reported of Sayers being ‘bitterly disappointed by the reception of the Report—“two years of my life—two years wasted”, he once exclaimed’—thus emphasizing that the Radcliffe doctrine is, largely, Sayers’s doctrine. Some overemphasis on the relevance of Sayers’s works might not be out of place in an obituary, but Cairncross’s account finds ample support in the fourth edition of Modern Banking (Sayers, 1958)—which was given to the press in the spring of 1957 (the preface is dated April 1957), just when the Radcliffe Committee was about to start—as well as in Sayers (1960). 12 Gurley (1960, p. 678), in particular, commented ironically on the idea that interest rates might depend on expectations as to what the interest rate would be: ‘It is fair to say, I think, that the Committee looks upon “expectations and confidence” of the public as the chief determinant of rate levels—though lip service is paid now and then to the role of liquidity. And these expectations are greatly affected by what the authorities say, by the public’s interpretation of their mood. That is, expectations are molded by “faces” made by the authorities’. Gurley himself noticed that ‘The Committee was also influenced in this view by R. F. Kahn’, but he suggested that the Committee had ‘gone somewhat further than Kahn did’ in his evidence (1960, pp. 679, 691). Gurley plainly affirmed that the argument based on what would nowadays be called forward guidance made no sense. Unfortunately, this latter argument, certainly one of the most unconventional of the Radcliffe Report, is one of the main influences that Kahn may have had on the Report. Kaldor (1960) seems to suggest that the Radcliffe Committee might have adopted Kahn’s doctrine for reasons that were not coherent with Kahn’s own view, but did not develop this point. Chick (1973, ch. 4) avoided any reference to Kahn in her critical review of the Radcliffe Report, presented as strongly influenced by the ideas of Sayers. 13 On this, see Howson (1988, 1993). 14 Data on inflation are from Allen (2014, appendix A, table B). 15 On this aspect, see Artis (1965, pp. 43–46). 16 Further archival evidence on this point is now available in Capie (2010, pp. 88–90). 17 Seldon (1960) collects the opinions of those who opposed the Radcliffe Committee’s conclusion based on traditional quantity theory arguments (on this, see also Laidler, 1989). 18 The absence of a clear definition of the concept of whole liquidity position was lamented by many observers at the time. On this, see Artis (1961, pp. 345–50) and Gurley (1960), among others. 19 On this, see Artis (1961, pp. 360–64) and the references therein. 20 As explained in the preface to the book edited by Andrews and Wilson in 1951, ‘because of war-time and other difficulties, these issues were small, and have been unobtainable for some time’. For this reason, some of these papers were selected for reprint in the volume of 1951. 21 Segmentation between trade unions combined with competition among them and employers causes a permanent inflation bias whereby each category tries to keep ahead of others and employers pass wage increases on to prices. This generates an inflation bias, which Kahn hopes to contain through the centralization of wage bargaining and adoption of specific wage policies together with measures capable of influencing inflation expectations. 22 On this, see Kahn (1960, p. 147). Kahn shares Sayers’s refutation of the idea of constant money velocity even if based on different arguments. While Kahn emphasizes the adjustment of money velocity to changes in economic activity, for a given money supply, Sayers (1960, p. 715) focuses on the endogenous adjustment of money supply, and more generally of liquid means. ‘The trouble [with the concept of velocity] begins when we jump to the conclusion that a rise in the ratio of payments to bank deposits means only that firms have decided to put their bank deposit to faster uses, whereas much of what has happened is that firms have decided to give more credit and people have shifted some of their savings into the hands of more nimble intermediaries. The artificiality of the concept [of velocity of circulation] lies, in short, in its reliance on a distinct and identifiable category of Money’ (Sayers, 1960, p. 715). 23 Kahn expressed his support for cheap money policy as a way of promoting investment on several occasions during the 1950s. He did so in the symposium on the reactivation of monetary policy published in the Bulletin of the Oxford University Institute of Statistics in 1952 (Kahn, 1952A) and again in two articles published in the Financial Times on 3–4 June 1954 under the title ‘The case for cheap money’. In this latter article in particular, Kahn described a situation in which unemployment was at a satisfactorily low level, the British economy was growing but productivity was not growing fast enough to keep pace with other countries (e.g. Germany). For Kahn, too large a share of aggregate income went to consumption, thus reducing the resources that remained available for increasing productivity via investment. In order to provide a remedy to this situation, Kahn strongly argued in favour of using monetary policy: ‘Monetary policy is surely the answer—in the shape of cheap money and credit expansion’. 24 Short-term interest rates mattered also for their influence on stock-holding, as discussed by many participants in the works of the Radcliffe Committee, including Kaldor (RC, 1960B, vol. 3, Memorandum, n. 20, §27) and Little, Neild and Ross (RC, 1960B, vol. 3, Memorandum, n. 23). Sayers (1958, pp. 167–69) refused this doctrine, which was presented in Hawtrey’s traditional version. Although without mentioning Hawtrey’s name, the Report (§386) follows Sayers’s argument in this respect. 25 As Kahn mentions at the beginning of the Memorandum, he shares this idea with I. M. D. Little, R. R. Neild and C. R. Ross, who discuss it in their Memorandum to the Radcliffe Committee titled ‘The Scope and Limitations of Monetary Policy’ (RC, 1960B, part 13, no. 23, pp. 159–67). 26 ‘It is not only a question of rates of interest. The ability of the banks to make advances has an influence independently of actual rates of interest. … Therefore I would argue that there should be imposed on the banks a restriction on the amount of advances which they can make’ (Kahn, RC, 1960C, Q. 10993). On this point, in §60 of his Memorandum, Kahn referred to Keynes’s ‘fringe of unsatisfied borrowers’ in chapter 37 of the Treatise on Money. 27 On this, see Collected Writings of J. M. Keynes, vol. 7, p. 206; Tily (2006) and Peden (2004, ch. 11). 28 On Keynes’s evolving ideas about the possibility that monetary authorities influence both short- and long-term interest rates and on connections between these ideas and unconventional monetary policy today, see Kregel (2013). 29 TDR, introduced as a war-time measure in July 1940, were non-marketable short-term debt instruments, which banks were obliged to buy but could not use as collateral. TDR were discontinued after the war and their reintroduction met with strong oppositions by the banking community during the 1950s (Fforde, 1992, pp. 635, 651). On TDR’s potential to curb inflation, see Sayers (1956, pp. 220–21). 30 On Dalton’s policy, see Howson (1987) and Fforde (1992, pp. 327–59). 31 As Kahn acknowledges at the start of his essay on liquidity preference, this assumption marks a significant dividing line between the Hicks school, which includes Kalecki and Kaldor, and the alternative school of thought, comprised of Robertson, Harrod and J. Robinson (Kahn, 1954, p. 72). Considerations of space prevent us from reconstructing the vast British literature on the subject between the 1930s and the 1950s. 32 For a formalization of Kahn’s theory in the case of three assets (money, irredeemable bond, short-term commercial paper), see Dardi (1994, section 3). Our notation is close to that used by Dardi. 33 The same result obtains if individual expectations are not unique but still held with complete certainty. The only difference with the respect to the case of unique expectations is that each agent, as long as she is sufficiently risk averse, will hold both money and bonds. On this and related aspects, see also Tobin (1958). 34 For a definition of these two types of risk, which Kahn draws from Hawtrey, see Kahn (1954, pp. 81–82). 35 ‘Bankers’ are expert investors who fear capital losses with particular intensity, prefer short-run liquid investments and enter into the bond market only when very confident that bond prices are rising, possibly in connection with bull speculation. On the other hand, ‘widows and orphans’ are inexpert investors who especially fear income losses, prefer long-run safe securities and leave the bond market only when sure that bond prices will fall, possibly in connection with bear speculation. The simultaneous presence of these and other types of investors, each with its own preferred habitat, contributes to making money, bills and bonds imperfect substitutes, enhancing the possibility of a separate but coordinated determination of interest rates at different maturity dates. On the notion of bankers vs. widows and orphans, see also J. Robinson (1951). 36 On the possibility of interpreting Kahn’s position and the ensuing conclusions of the Radcliffe Committee in the light of the post-Keynesian concept of endogenous money supply, see Sawyer (2002). 37 On the reluctance of the Bank of England to countenance funding (substitution of Treasury bills with government bonds) as a means of containing liquidity and inflation in a falling gilt market, see also Capie (2010, pp. 88–90). 38 Britain’s total National Debt was 175.2% of GDP in 1951 and 112.5% in 1959 (Allen, 2014, table A2, p. 245). 39 On this, see Allen (2014, table A4, pp. 248–49 in particular). 40 On the operations of the Government broker, see RC (1960A, §§581–82 and 1960C, Q. 11921–-12065). 41 When bond prices fell (as might occur in case of rising inflation), investors not only abstained from underwriting new bond offers but became net sellers of bonds. In this case, the Issue Department would be compelled to absorb government bonds, leading the amount of liquid assets to increase with adverse repercussions on inflation and bond prices. In his Memorandum and oral evidence in front of the Radcliffe Committee, Kahn gives the impression of downplaying the relevance of these institutional obstacles even if he does not ignore them completely. 42 On this official attitude about the behavior of investors, see also Goodhart (2012, p. 125). © The Author(s) 2017. Published by Oxford University Press on behalf of the Cambridge Political Economy Society. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Journal

Cambridge Journal of EconomicsOxford University Press

Published: Oct 31, 2017

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off