II. THE TREATISE ON MONEY

In fact, public debt is not much discussed in TM. Keynes likens “[w]ar-expenditure, not covered by taxation” to “a sudden increase of investment” (TM I, p. 283n; also II, p. 171). Since his analysis of saving/investment equilibration in the Treatise is in orthodox, marginalist terms (TM I, pp. 154–155, 185–220)—such that an exogenous increase in private investment ceteris paribus increases the “natural” (and sooner or later, the market) rate of interest—this implies that an increase in debt-financed government expenditure, at least of a substantial magnitude, raises interest rates. (Debt-financed war expenditure can be conceived of, alternatively, as an exogenous decrease in public saving: TM II, p. 167; also in Keynes Reference Keynes1936, p. 128n [hereafter, GT]; CW XIII, pp. 232, 458.) The significance of deflation and inflation for real public debt burdens is noted (TM II, pp. 181, 393), and the unorthodoxy of debt-financed public works (TM II, p. 170; cf. p. 376).^{Footnote 1}

The term structure has a more significant role in TM, particularly by way of Keynes arguing for the capacity of short rates to influence long rates, with a view to his overall purpose of providing a comprehensive account of the transmission mechanism. In Keynes’s Reference Keynes1930 theory the crucial element in explaining price-level behavior is the balance between aggregate investment and aggregate saving, that balance in turn being determinable by the level of market interest rates relative to “the natural rate of interest,” where the latter concept is close to that of Knut Wicksell.^{Footnote 2} Hence, with saving behavior regarded as relatively passive, the capacity of policy to achieve price stability requires policy to be able to sufficiently influence the interest rates that are understood by Keynes to regulate investment, and, most importantly, investment in fixed capital (TM II, p. 348). The term structure enters the picture at this point because for fixed capital it is long rates that matter: “How can we be sure that the long-term rate of interest will respond to the wishes of a Currency Authority which will be exerting its direct influence, as it must [cf. GT, p. 203], mainly on the short-term rate?” (TM II, p. 352; also pp. 187–188, 195, 347).

The point of departure for linking short and long rates is the proposition that if the sequence of short (say, ninety-day) rates over the course of, say, the next twenty years was known by market participants with confidence, then the twenty-year interest rate would reflect that sequence of short rates—implicitly, due to arbitrage (TM II, pp. 352–353; the Bank of England discount rate, “bank-rate,” applied to ninety-day bills: I, p. 200). That being so, a change in the current short rate, in and of itself, would exercise only slight influence on the long rate. Keynes argues that the influence is “much greater” than this for a number of reasons that need not detain us, except to note that they indeed include the capacity of current short rates to influence expectations of future short rates (TM II, pp. 353, 356–362).

In a further thread of argument pertinent to the yield curve, in the context of assessing “bank-rate policy” versus market operations, Keynes affirms the capacity of the latter to directly influence long rates (TM II, pp. 251–252; also pp. 231–232). And in the penultimate chapter of the book, there is a subsection on market operations “to the point of saturation”: “extra-ordinary methods,” “extreme measures,” likely required only if “conditions of acute slump or boom have been allowed to develop”—although it is slumps rather than booms that are more likely to “defy all normal methods of control.” What Keynes has in mind here is “taking measures which would have the effect of causing the total volume of bank-money [deposits] to depart widely from its normal volume, whether in excess or in defect,” with a view to accommodating abnormal demand for liquidity or for securities (TM II, pp. 369–374; cf. pp. 386–387). But all this, it may be emphasized, for Keynes in 1930, pertains to extreme circumstances, not normal conditions.

It is worth finally noting from TM that the 1930 slump is first and foremost attributed to market rates of interest since the mid-1920s having been held above the natural rate, precipitating a cumulative downturn of both the price level and aggregate activity (TM II, pp. 377–382; echoed in newspaper articles the same year—CW IX, pp. 132–133, 321). Again, the detail of the argument may be left aside. What is of interest here is that by way of this interpretation, Keynes seeks to reconcile his conviction that too high market rates of interest are a fundamental cause of the slump, with his then orthodox theoretical framework. His conclusions prefigure a persistent theme in his subsequent writings during the 1930s, that interest rates, short and long, are too high and are capable of being reduced by policy (but during those later years these views will break out of the orthodox straitjacket): the prospect for the next two decades is “a strong tendency for the natural-rate of interest to fall,” unless delayed by “Central Banking Policy preventing the market-rate … from falling as fast as it should” (TM II, pp. 207–208); lasting economic recovery requires “a very great fall in the long-term market-rate,” which needs to be “accelerated by deliberate policy” (TM II, p. 383); lenders have become “accustomed to high rates” because notions of “the normal and the permanent” are “mainly fixed by actual experience of the last fifteen years” (TM II, p. 384); for appropriate adjustment of rates, it might suffice “merely to produce a general belief in the long continuance of a very low rate of short-term interest” (TM II, p. 386).

III. THE MULTIPLIER AND DEBT FINANCING

The key theoretical elements in Keynes’s transition to GT can be summarized as follows:

1. 1. The fundamental breakthrough is the development of the multiplier mechanism and its application to the determination of aggregate activity levels. Movements in aggregate activity cease to be a subsidiary element in the dynamics of price-level disequilibrium, and are no longer a function of wage or price inflexibilities, as in TM, but rather are an expression of a theory of equilibrium aggregate activity in which full employment is a special (and unlikely) case.^{Footnote 3}

2. 2. This mechanism entails the coordination of (planned) investment and saving via the former determining the latter through multiplied changes in (actual) aggregate activity, including possible price-level changes (GT, pp. 63–65, 82–85, 117, 375)—so that there is no unique level of interest rates, functionally required for saving-investment (or capital supply-and-demand) equilibrium.

3. 3. This leads to Keynes’s abandoning the notion of a natural rate of interest (although somewhat incompletely) and positing a monetary theory of interest, or the notion of the level of interest rates as a mere convention—a customary norm resulting from the interplay between central bank behavior and money market sentiment (Aspromourgos Reference Aspromourgos2007, pp. 514–517).

4. 4. Just as the new theory of aggregate activity levels and income fundamentally changes the financing issue with respect to private investment, since saving is no longer a prerequisite for investment, so it also changes it for the economics of debt-financed public spending. With public sector expenditure and income incorporated in the theory, the multiplier mechanism whereby investment determines saving becomes a mechanism in which private investment plus government expenditure generates an equal magnitude of private saving plus tax revenues.

The first three sets of propositions are posited here without argument; and international economic issues (e.g., external constraints on interest rate policy) are also put aside, in order to focus on the more general closed-system issues.

As indicated above, it is a continuous theme of Keynes’s writings through the 1930s that policy-driven lower interest rates are desirable and feasible, with a view to economic recovery—and a theme often also connected with rolling over outstanding public debt at lower yields (e.g., CW XXI, pp. 79, 100, 106, 114–115, 240, 256, 319, 332, 454, 488, 582; these instances cover dates from 1932 to 1939). Open market operations to directly act upon long rates are a similarly consistent theme (e.g., CW XXIII, p. 366; XXI, pp. 200, 265, 271, 297, 327–328, 571; IX, p. 353; instances from 1931 to 1939). One may recall here also Keynes’s GT “euthanasia of the rentier” doctrine, favoring a zero real riskless interest rate (Aspromourgos Reference Aspromourgos, Aspromourgos and Lodewijks2004; Reference Aspromourgos2011, p. 641).

Keynes’s conviction in favor of “cheap money” is bolstered by a more or less axiomatic belief that if the level of profitable aggregate investment is less than full-employment saving, the long rate must be too high—which is, after all, just a corollary of the notion of a well-defined and robust inverse relation between aggregate investment expenditure and the interest rate (a notion that persists in GT): “So long as there is serious all-round unemployment I consider this proves that the equilibrium rate of interest is lower than the ruling rate” (CW XXI, p. 345, emphasis added; also pp. 200–201). While lower rates are no longer rationalized, as in TM, in terms of bringing sticky market rates down towards a lower natural rate, and Keynes explicitly repudiates the natural rate concept in GT, the interest/expenditure functional relation implies the existence of a full-employment rate of interest—as he also makes explicit in GT—a ghost of the discarded natural rate concept, persisting in the new theory (GT, pp. 31, 183, 202, 220, 242–244, 267, 309). Of course, public investment is also, and increasingly, advocated throughout the 1930s, a fact so well known as to hardly require supporting documentation (Moggridge and Howson Reference Moggridge and Howson.1974, pp. 237–239; for examples, CW XXI, pp. 60, 394). The considerable criticism of the efficacy of monetary policy in GT (pp. 163–164, 202–208, 315–321) may also be recalled here.

Keynes’s general position that lower interest rates are desirable and feasible then finds application in the further proposition that there is no intrinsic necessity for the general level of interest rates to rise in the face of economic recovery or expansion, whether due to increased private expenditures or expanded debt-financed public investment (e.g., CW XXI, pp. 314–315, 534–537). Certainly this is so for systems with excess production capacity—and even for systems without excess capacity, via recourse to policy instruments other than interest rates, notably, taxation (CW XXI, pp. 377–378, 536, 549, 557; Moggridge and Howson Reference Moggridge and Howson.1974, pp. 231, 243–244). Keynes’s conviction that expansion based upon debt-financed public investment need not be accompanied by upward pressure on rates is greatly reinforced by the multiplier analysis developed by Richard Kahn, with important input from Meade (Kahn Reference Kahn1931, pp. 187–191). Skidelsky (Reference Skidelsky1992, p. 451) reports an 18/12/31 Keynes letter (not published in CW) as the first known written recourse by Keynes to the multiplier mechanism. That letter certainly outlines the investment-determines-saving logic (King’s College Keynes Papers, L/31/180–181; hereafter cited as JMK), including the associated inter-sectoral balance, discussed immediately below.

That logic is itself a corollary of the spending-determines-income logic of a demand-side theory of activity levels: “It is often said by wiseacres that we cannot spend more than we earn. That is … true enough of the individual, but … misleading if … applied to the community as a whole. For the community as a whole it would be much truer to say that we cannot earn more than we spend” (CW XXI, p. 126, 20/7/32; also pp. 194–195, 26/4/33). In the March–April 1933 Means to Prosperity this multiplier mechanism is explicitly applied to both private and public spending (CW IX, p. 349).^{Footnote 4} Similarly, in a fragment of GT draft (tentatively dated 1933), Keynes argues that the “maxims of ‘sound’ public finance” are largely based on the supposition that saving causes investment, going on to sketch a version of the paradox of thrift and the underlying multiplier mechanism (CW XXIX, pp. 102–111; the paradox of thrift, also at XXI, pp. 287–288, 19/9/33; GT, pp. 84, 104–106, 110–112; XXVII, pp. 390–391, March 1945). As he sums up his contrary view: “It is … the act of investing which ‘finds its way’ into saving, rather than the other way round. On the other hand, individual acts of saving not only do not necessarily ‘find their way’ into investment, but are liable to have precisely the opposite effect” (CW XXIX, p. 106; also at p. 107, applied to war expenditure).

It is also a corollary of the demand-side multiplier logic that a debt-financed public expenditure, via the resulting new expenditure-income equilibrium, generates additional private saving plus tax revenues of equal magnitude. Thereby the notion that increased debt-financed public investment necessitates higher interest rates (in part, to supposedly induce additional saving) appears to be decisively undermined. This is made very clear in three documents of April, May, and July 1939 that essentially provide variants of the same argument. For a closed system, “the income of the community will be equal to what the Government spends plus what individuals spend. … Thus the excess of the community’s aggregate income over what individuals spend, which is left over and available to pay taxes and loans to the Government, must be exactly equal to what the Government spends” (CW XXI, p. 515). Given a balance of aggregate expenditure and income, the public sector deficit will be matched by a private sector surplus of the same magnitude: “The savings will come into existence pari passu with the expenditure. The only question which arises is as to the ultimate form in which they are held—whether as balances at the Bank of England, in Treasury bills and bonds, or in longer-dated Government debt” (CW XXI, p. 516; also p. 523, and similarly, p. 398, from 1937).^{Footnote 5}

In the May 1939 document presenting a variant of the same argument, Keynes comments:

The qualification that the saving will come “mainly” from “increased output” is tacitly allowing for some possible price-level change as well. The July 1939 document, a two-part article in the Times, is a revised version of this May memorandum, very similarly worded with respect to the above-quoted text (CW XXI, p. 556). The logic of aggregate expenditure-income balance and associated public/private inter-sectoral balance is subsequently systematically applied to war finance and, in particular, to the British public sector budgetary framework from 1941 forward—with the balance partly brought about by inflation in a near-full-employment economy, although Keynes is firmly opposed to using inflation as a policy instrument.^{Footnote 6}

IV. DEBT FINANCING AND THE LEVEL OF INTEREST RATES

There are two difficulties that can be raised with respect to the multiplier resolution of the coordination of saving and investment, difficulties that might stand in the way of dismissing a role for interest rates: the character of the disequilibrium path to re-equilibration when investment, private or public, increases; and the question of provision of a means of payment or medium of exchange to enable increased investment spending—finance that must be made available prior to, or at least simultaneous with, the investment spending.

The former issue seems to cause no problem: as Meade formalized the saving dimension of the multiplier process in 1930–31—in a manner allowing for time lags between investment and the consequent induced consumption and associated desired saving—there will be an excess of actual saving over planned saving during the temporal process, with that discrepancy diminishing and approaching zero as the process approaches completion (see Meade Reference Meade1993; cf. GT, pp. 122–125; Chick Reference Chick, Harcourt and Riach1997, pp. 166–169, 176–179). The financing issue is more significant; but here there is an asymmetry between private and public investment. We may leave aside the question of possible difficulties a means-of-payment constraint might pose for the autonomy of private investment. But with respect to public investment (or public recurrent expenditure for that matter), the state of course can issue means of payment itself, most notably, fiat currency.

Keynes is well aware of this difference. In a string of 1939 commentaries he makes the point, in relation to debt-financed public expenditure, that government can (and should) issue the debt after it has done the spending. In the Times article already quoted above:

The prior funding mechanism Keynes has in mind here is partly monetary financing, as is clarified in a letter to the Times two weeks later:

Keynes goes on to argue that eventually such liquid savings will be redirected towards long-term securities, placing downward pressure on yields; “the Treasury should postpone the issue of new loans, other than Treasury bills, until this process is well advanced” (CW XXI, p. 525). It is worth noting that the context here is a crisis situation, with war only months away.

The same argument is presented in the previously mentioned 28/5/39 memorandum: “[W]ith modern representative money [see TM I, pp. 6–9] and a modern banking system, we know that the necessary ‘finance’ can be created by a series of ‘book’ or ‘paper’ transactions. The Treasury can ‘pay’ in effect by ‘book’ entries and the book entries can be transformed into a regular loan at a much later date” (CW XXI, p. 540). Keynes allows that if “these ‘book’ entries … such as Treasury bills” (cf. GT, p. 167n) take “an unlimited scale,” then this liquidity might prove “dangerous,” in enabling a later “uncontrolled expansion of private enterprise.” But he remains confident of a manageable process: “[I]f the Treasury is moderately patient, the weight of natural market forces will by themselves render a funding policy possible at a reasonable cost. It is simply a question of waiting and of making it clear that loans will only be available at a modest rate of interest” (CW XXI, p. 540). The argument is elaborated further, Keynes also making explicit the contrast on this issue between public and private debt financing (CW XXI, pp. 542–545).^{Footnote 10}

What Keynes evidently has in mind with this kind of process is that the initial financing, largely via issuing Treasury bills, will be partly monetary financing, to the extent that the bills are taken up by public sector agencies. This will involve injection of additional outside money into the private sector—although when the take-up of the securities is by public sector agencies other than the central bank, it is a sort of substitution from “idle balances,” so to speak, held by one part of the public sector, to active balances expended by another part. During the 1930s public sector agencies other than the central bank were regularly purchasing parts of government securities issues, and then subsequently often selling the stock, gradually, to the private sector (Howson Reference Howson1975, pp. 160–166; also Sayers Reference Sayers1956, chs. V, VII, more widely on the public financing methods employed through the 1930s and subsequent war years). But the private financial sector is also understood to be partly, perhaps substantially, absorbing the initial Treasury bill issue (cf. Keynes Reference Keynes and Moggridge1923, pp. 141–146, on the role of Treasury bills in banks’ liquidity management). Probably Keynes is tacitly supposing that the private financial sector has a capacity to absorb the stock, up to some level, without adverse pressure being placed on the prices and yields of the stock, although the possibility of pressure on yields in the absence of accommodative monetary policy is explicitly acknowledged elsewhere (e.g., note 7 above), certainly in relation to longer-dated securities.

V. MANAGING THE TERM STRUCTURE

Keynes’s notion of “waiting” before securing the longer-term debt finance for public spending is motivated by a concern about the term structure, a concern to avoid placing upward pressure on long rates. There is a very extensive set of extant commentaries evidencing Keynes’s sensibility regarding managing the psychology of the market, with a view to keeping debt servicing costs as low as possible. We may present a range of these commentaries, to convey a sense of this sensibility and the issues involved, under four heads.

1. False Consciousness

In an echo of the problem of “false consciousness” that Keynes later raised with respect to functional finance (Aspromourgos Reference Aspromourgos2014b, p. 419), he comments: “There are enormous psychological advantages in the appearance of economy. It … tends to lower the long-term rate of interest. … But that does not prevent economy from being deflationary and probably injurious to business profits” (CW XXI, p. 110; also pp. 107, 126, all 1932). The Means to Prosperity repeats this notion that the “financial confidence” from “budget policy approved by public opinion,” for “psychological reasons,” enables “the transition to a lower long-term rate of interest”; but this justifies only “temporary reduction of loan-expenditure,” since “the whole object of the policy is to promote loan-expenditure” (CW IX, pp. 353–354). Discussing public works in GT (p. 120), Keynes notes: “With the confused psychology which often prevails, the Government programme may, through its effect on ‘confidence’, increase liquidity-preference or diminish the marginal efficiency of capital.” Nearing the end of the war, he sees an element of a false consciousness in interest rate expectations: an “expectation of higher rate … after the war … based on the false belief that it will be necessary to stimulate and encourage saving and that cheap money during the war has been the result of controls” (CW XXVII, p. 391).^{Footnote 11}

2. Psychology of the Market

The capacity of policy to manage rates presupposes an element of indeterminacy, which in the GT is attributable to a degree of malleability of the psychology underpinning liquidity preference and conventional interest rate beliefs. This is evident in incipient form earlier: “there is a large conventional or psychological element in the market rate of interest which needs firm and skilful management” (CW XXI, pp. 116–117; also p. 123, both 1932). In 1934, advocating the feasibility of a low (and lower) long rate, Keynes nevertheless acknowledges, but rejects, “a grave doubt in the mind of the market as to whether the existing price of long-term securities will be maintained,” based on “the evidence of past experience” of economic recovery, “the expectation that Consols will fall when trade recovers” (CW XXI, pp. 313–314). Discussing whether new debt-financed defence expenditure need place upward pressure on interest rates, Keynes emphasizes the importance of “the psychological atmosphere towards gilt-edged and other securities.” It is “not a shortage of savings which will impair the position of gilt-edged securities, but a change in psychological expectations as to their future prospects.” Hence the importance of policy “maintaining stability in the gilt-edged market”; recent “weakness” in the market is partly due to public debate “about future policy” (CW XXI, p. 399; also p. 392, both 1937).

In February 1936 Keynes bemoans that the Treasury conducts itself in a manner that can only encourage a lack of market confidence in low rates: “Short-term money today is extremely cheap. But it is confidence in the future of short-term rates which is required to bring down long-term rates” (CW XXI, p. 375). After noting that the Treasury has most recently issued five-year debt at 1.5% and twenty-five-year debt at 2.75%, partly to retire short-term debt costing 0.5%, Keynes comments: “There can be no rational explanation of the longer-dated issue except that they themselves have no confidence in the short-term rate of interest remaining low. Since they largely control the situation, it is natural that humbler folk should be influenced by what the Treasury seem to expect” (CW XXI, p. 376). Similarly in 1939, Keynes writes that policy must promote “a sense of confidence in what the future borrowing policy of the Treasury is going to be” (CW XXI, p. 559), and of the importance of

3. Policy Gradualism

The need to manage—one could say, massage—market sentiment and expectations also points to the possible need for caution and gradualism. Writing of the Treasury and Bank of England, Keynes comments: “it lies within their power, by the exercise of the moderation, the gradualness, and the discreet handling of the market of which they have shown themselves to be masters, to make the long-term rate of interest what they choose within reason” (CW XXI, p. 395, 1937, emphasis added; also p. 317, 1934; XXII, pp. 159–160, 1940).^{Footnote 12} Similar sentiments are expressed concerning the US (CW XXI, pp. 327–328, 1934). In early 1935 Keynes sees obstacles in the way of further long-rate reductions, the key problem being “the attitude of British institutional investors to the future of the rate of interest.” In a degree of deference to the prevailing market psychology, Keynes recommends the monetary authorities “consolidate the [interest rate] position which has been won, rather than … aim at an immediate further advance”; adjustment of other rates that are lagging behind the decline in government yields is more important (CW XXI, pp. 350–351). Keynes sums up his position at this point in time: “I feel not less strongly than before [CW XXI, pp. 312–317] the importance of a declining long-term rate of interest, but a greater degree of confidence than now exists in the maintenance of the rates of interest we already have at a level not above their present figure is our most pressing need” (CW XXI, p. 351). The Treasury can contribute to that confidence by itself showing “confidence in the expectation of a declining rate of interest in the future,” rather than thinking in terms of “trapping the investor, so to speak, into lending to them for an indefinite period on terms which he will subsequently regret” (CW XXI, p. 351; cf. pp. 25n, 106, 1931–32).

4. Spectrum of Maturities

To underpin the success of a cheap money policy, Keynes wants the authorities to accommodate the market’s preferences with respect to maturities, ending the British practice of tending to offer the market too stark a choice between only either very short or very long (notably, perpetual) securities, which naturally tends to heighten long rates:

Keynes is still prosecuting the argument in 1937. The failure of the British authorities to offer a range of maturities is compared unfavorably with American practice; in Britain “the greater part of the debt … has no fixed date of repayment within the next 25 years,” whereas the US has “notes or bonds falling due for repayment in almost every year,” so that “every taste is suited.” This “allows the American Treasury to borrow at a materially lower average rate.” The British Treasury “should profit from the anxieties of the public and save interest by supplying them with the potential liquidity which they demand” (CW XXI, pp. 402–403; also pp. 517, 541, 544–545, 559, all from 1939)—where “liquidity” here includes shorter-dated debt (see section VI). The same view is pressed throughout the war (CW XXII, pp. 158, 410–420, documents from 1940–44), although by then the British authorities were issuing a wide range of maturities.

VI. THEORY OF THE TERM STRUCTURE AND INTEREST RATE POLICY

What, then, is Keynes’s mature theory of the term structure, enabling his conviction that policy can manage and shape it? All of the above-documented Keynes commentaries concerning the role of expectations of future interest rates in determining current interest rates—and hence the desirability of policy managing expectations—are consistent with the TM point of departure for explaining the term structure, that current long rates are regulated by future short rates. For most of GT Keynes theorizes in terms of “the” rate of interest, although making explicit that this is serving as a proxy for “the complex of the various rates of interest current for different periods of time, i.e. for debts of different maturities” (GT, p. 167n). Nevertheless he does address the determination of the term structure, in the framework of his liquidity-preference theory of interest.

Keynes makes “the existence of uncertainty as to the future of the rate of interest, i.e. as to the complex of rates of interest for varying maturities which will rule at future dates,” the “necessary condition” for liquidity preference. In doing so he essentially repeats the logic of the TM point of departure on the term structure: if future short rates are known with certainty, then current long rates can be straightforwardly inferred from the zero-profitable-arbitrage condition.^{Footnote 13} With future rates uncertain, “if a need for liquid cash may conceivably arise before the expiry of n years, there is a risk of a loss being incurred in purchasing a long-term debt and subsequently turning it into cash, as compared with holding cash” (GT, pp. 168–169).^{Footnote 14} Hence also the following assertion of future rates anticipated and at least partially factored into current rates: “the expectations, which are held concerning the complex of rates of interest for various terms which will rule in the future, will be partially reflected in the complex of rates of interest that rule to-day” (GT, p. 143; also p. 145n).^{Footnote 15} Exposure to variability of bond prices gives a role to expectations of future long rates, as well as expectations of future short rates.

It follows from this understanding of the term structure that to succeed, a policy of persistent or permanent low(er) long rates must shift average market opinion as to the level of the normal (or “safe”) rate of interest: the short rate is “easily controlled”; but the long rate “may be more recalcitrant when once it has fallen to a level which, on the basis of past experience and present expectations of future monetary policy, is considered ‘unsafe’ by representative opinion.” Hence the very same policy may fail if perceived as “experimental” or “easily liable to change,” but “easily succeed” if seen as “reasonable … practicable … rooted in strong conviction, and promoted by an authority unlikely to be superseded.” Keynes goes on to discuss the interest rate as a “highly conventional” phenomenon, concluding with cautious optimism that a monetary policy “of persistence and consistency of purpose” will be able to shift the conventional perception as to the safe or normal rate of interest towards lower levels (GT, pp. 201–204). He also endorses a policy of open market operations across the whole range of maturities, in the spirit of the “tap” system (more on this below), and something he’d been more or less advocating since TM, with a view to enhancing the influence of monetary policy across the yield curve (GT, pp. 205–206). This approach to the term structure is also evident in policy commentaries by Keynes over subsequent years (CW XIV, p. 153; XXII, pp. 63–64, 84; XXIX, p. 266; 1937 to 1939).

The term structure is again systematically addressed by Keynes as a consequence of his membership of a 1945 government in-house committee to consider postwar debt management, the National Debt Enquiry. His handwritten notes for the Enquiry restate the liquidity preference doctrine, that interest is the compensation for “depart[ing] from liquidity.” For short versus long rates it then becomes a question of “[w]hat determines the reward the individual requires to surrender his liquidity for a long or short period” (emphasis added), implying that holding a longer-dated security is more illiquid a position. But why, if both securities are equally tradeable from day to day? It is individuals’ “expectation … or … uncertainty about the future changes in r. of i.”; if they “just don’t know” and seek to protect themselves “from possible loss in the event of … desiring liquidity, then the shorter are preferable and you need to earn a risk premium to lock yourself up longer.” In other words, the greater risk exposure at longer maturities is due to greater asset price variability at the longer end, for any given percentage variation in yields, across the board—a result now well known, as the basis for a normally positive term spread (cf. GT, p. 203). Keynes goes on to argue that so long as the authorities have no “counter-liquidity preference”—i.e., they are “indifferent about [short versus long] funding”—then “they can make both the short and long-term [rates] whatever they like” (emphasis added), although this is qualified by considerations of “whatever they feel to be right” for “employment and other social reasons [e.g., ‘how much reward to saving is socially desirable’]” (CW XXVII, pp. 390–392; also XXI, pp. 544, 558, 563, all 1939).

With a view to this rate-setting objective he endorses the “tap” issuing system, whereby the authorities set the rates at which they will supply securities, across the range of maturities, and allow the private sector to choose the quantities of the various maturities that they wish, over time, to take up—and defends a 3% bond issue, partly on the basis that “the euthanasia of the rentier should not take place just yet.” But it’s also clear here that Keynes is happy to see securities not taken up, so that funding can occur via the private sector’s holding idle money balances. The policy cautiousness of which Meade speaks (in our epigraph quotation) is evident here: “The essence of our interest policy should be to give a sufficient immediate reward to saving, so not to run prematurely against public psychology” (this is the month following Meade’s comment). Keynes’s subsequent 18/4/45 written summary of his views, for the benefit of the other Enquiry members, again endorses the tap system—whereby “the preferences of the public … determine the distribution of the debt between different terms”—but makes explicit that the rates set should be constrained by a variety of considerations, including responding to changing private sector preferences between maturities by altering the rates set. And there is that caution again: “continuity of policy and gradualness of changes should be ensured unless in exceptional circumstances and for grave cause” (CW XXVII, pp. 392–398; also p. 400).

Hence Keynes’s above-quoted comment that the authorities can make short and long rates “whatever they like,” in fact, is subject to rather substantial qualification. In the tap system the authorities set the rates they offer the market across new issues of maturities (in effect, exogenously fixing the term structure), and allow the market to determine the quantities taken up, with monetary financing making up any shortfall between the quantities of securities issued and the private sector take-up. Putting aside the self-imposed constraints that Keynes mentions (e.g., the above-mentioned “social reasons”), if the policymaker can set any rates and spreads, then, for example, why not offer 2% and 1% on short and long securities, respectively? There are limits to the authorities’ capacity to set term spreads that, to take the most obvious consequence, make a negative term spread non-credible as a persistent or permanent policy. Hence the levels and spreads Keynes proposes could not have been other than with a view to constraints derived from the psychology of the market—even if with influence on those constraints from policy (along the lines of GT, pp. 201–204)—that psychology, in turn, being determined by objective phenomena (e.g., relative liquidity, differential risk, objective factors shaping future short rates) and subjective factors (e.g., false consciousness, attitudes towards risk). All the rate structures proposed by Keynes in the 1930s and 1940s entail a positively sloped yield curve.

The notion of private lenders’ liquidity preference requiring a risk premium in longer interest rates implies that a policy of shifting the composition of debt towards a longer average maturity—e.g., rolling ten-year debt over into twenty-year debt—will tend to steepen the yield curve, in order to induce the private sector to hold a larger proportion of longer-dated public debt. This possibility is consistent with Keynes’s opposition to government “counter-liquidity preference”: if the demand side of the debt market ceteris paribus prefers shorter securities (liquidity preference), a preference by the supply side of the market ceteris paribus to issue longer securities (counter-liquidity preference) must tend to make the yield curve steeper than it otherwise would be. Keynes wants the authorities to be more or less indifferent to maturity composition, and just aim to minimize the average cost of debt servicing (CW XXII, p. 418, 1943).^{Footnote 16}

If, at some point in time, Keynes’s preferred “euthanasia” policy were to be implemented—a zero real riskless rate of interest as a permanent policy—then to the extent that this was associated with a permanently constant nominal short rate, term spreads should go to zero. (The same logic applies to setting a very low, rather than strictly zero, real rate, or, indeed, any constant level.) The zero-profitable-arbitrage condition would equalize longer rates with that constant short rate. At least this would be the case, so long as the policy was credible to the markets into the more or less distant future; and putting aside: (a) variability of the inflation rate or of risk premia, which, by leading to variability of nominal rates, could provide further grounds for positive term spreads; (b) any differential risk due to causes other than interest rate variability (e.g., default risk); and (c) random disturbances to nominal rates that could also provide still some slight reason for a positively sloped yield curve. This would be the ultimate conquest of the term structure. Of course, such a permanent policy would entail rejecting use of the interest rate as a short-run countercyclical or anti-inflation policy instrument, as Keynes indeed suggested in GT (e.g., pp. 320–329) and argued in the 1940s, against the views of Meade and others.^{Footnote 17} His opposition to discretionary monetary policy in this sense is precisely due to the consequent unsettling of interest rate expectations undermining the maintenance of low long rates: