Saturday, October 11, 2014

Axel Leijonhufvud’s “The Wicksell Connection”: A Review

Axel Leijonhufvud’s “The Wicksell Connection: Variations on a Theme” seems to be regarded as a classic paper in modern economics. In what follows I review it, though using Leijonhufvud (1979), not the later reprint Leijonhufvud (1981) (I am assuming there are no major differences between the two).

The central theme of “The Wicksell Connection” seems to be that Keynes’ liquidity preference theory was badly wrong, and caused modern macroeconomics to deviate from the “true path” of Wicksellian theory (Leijonhufvud 1979: 4).

For Leijonhufvud, Keynes’ economic ideas after the Treatise on Money but before the General Theory, which he dubs Keynes’ “Z-theory,” was the best and most defensible part of his economic thought: according to Leijonhufvud, it is that part of Keynes’ work that “should be preserved and developed,” while the liquidity preference theory and everything derived from it should be thrown out (Leijonhufvud 1979: 4).

Again, according to Leijonhufvud, the elimination of Wicksellian loanable funds theory was a mistake, since denying loanable funds means that the natural rate of interest cannot be defended (Leijonhufvud 1979: 5).

Leijonhufvud even commits himself to a long-run equilibrium model (Leijonhufvud 1979: 7–8), though recognises that the real world has coordination failures and incomplete information (Leijonhufvud 1979: 11–12). A great part of the paper is devoted to a review of the IS-LM and monetarist models, and the controversies pertinent to 1970s macroeconomics (Leijonhufvud 1979: 11–22). The upshot of Leijonhufvud’s analysis in this section is that the maladjustment of the interest rate and its effects on savings and investment is a central problem of macroeconomics (Leijonhufvud 1979: 23).

The concern with the savings and investment functions and their relationship to the interest rate emerged out of Knut Wicksell’s development of the quantity theory (Leijonhufvud 1979: 24).

Wicksell introduced credit-creating banking into his model of loanable funds, which was in essence a simple model where banks should be simple financial intermediaries between household savers and businesses:
“Household savings flow into, business investment finance flows out of banks. Here, banks are perceived in the first instance as loan intermediaries between the household and business sectors, rather than as money suppliers. In the short run, this theory of nominal income determination will focus on changes in the flow of bank-intermediated credit rather than in the stock of money. By lending more to the business sector than flows in as savings from the household sector, the banking system will cause the circular flow to expand. By lending less, they will make it contract. When nominal income is rising, investment exceeds saving by the net addition to loanable funds injected by banks. When nominal income is falling, banks let loanable funds "leak out" so that savings exceed investment. In income equilibrium, saving should equal investment; this requires that banks do no more and no less than intermediate the desired savings of the household sector. When they behave themselves ‘neutrally’ the excess demand for final goods at the prevailing level of money prices should be zero.” (Leijonhufvud 1979: 25).
In Wicksell’s model investment exceeds savings when banks extend credit over and above the quantity of money “saved” by households.

Leijonhufvud notes that Wicksell’s model is oversimplified in many ways too:
(1) bank-financed consumer credit is ignored;

(2) banks are assumed to be the major source of business finance: the securities markets are ignored (Leijonhufvud 1979: 26).

(3) firms do not finance investment out of retained earnings (Leijonhufvud 1979: 28).
Perhaps these were appropriate for Wicksell’s time, but they are not today. This does not stop Leijonhufvud from endorsing a modern version of loanable funds, however, and Wicksell’s idea of an interest rate that equates saving and investment at full employment called the “natural rate” (Leijonhufvud 1979: 27).

The natural rate is, as Leijonhufvud notes, the crucial concept:
“The central concepts of Wicksell’s analytical apparatus are, of course, the market rate and the natural rate of interest. The terms are names for two values of the same variable. The market rate denotes the actually observed value of the rate of interest, the natural rate the hypothetical value that the interest rate would take if and when the system is in equilibrium.

Equilibrium is used here in two senses at the same time. Their tight linkage is a result of the construction of Wicksell’s model that we have outlined; in other analytical settings we might well want to keep them distinct. First, the equality of the market rate with the natural rate is a condition of ‘monetary equilibrium’ in the sense of the system remaining at a stable price level. Second, it is also a condition for maintaining ‘real equilibrium’ in the sense of an allocation of current resources between consumption and investment such as to be consistent at the same time with the intertemporal consumption preferences of households and with the intertemporal production possibilities perceived by firms. The two are tightly linked in the sense that a divergence of market from natural rate will upset both at once and neither can be upset except by such a divergence occurring.” (Leijonhufvud 1979: 28–29).
For the two ways that Wicksell explicitly defined the natural rate in his writings, see my post here.

For Wicksell the system has a long-run tendency to monetary equilibrium, and the starting point of his analysis is an economy in actual monetary equilibrium.

Wicksell’s monetary equilibrium approach was adopted and developed in diverse ways by three important schools of economists who followed him, as follows:
(1) the Stockholm School;

(2) the Marshallian neoclassical school, and

(3) the Austrians.
This “legacy” of Wicksell can be seen in the diagram below (which should be opened in a separate window).

Keynes took over the Wicksellian framework by the time of the Treatise on Money, though by this stage Keynes was more concerned with deflation than inflation (Leijonhufvud 1979: 34). Fundamentally, Leijonhufvud sees the General Theory as a wrong turn in Keynes’ thought, because he rejected Wicksellian loanable funds theory.

What were Leijonhufvud’s theoretical and policy positions? Leijonhufvud rejected the neoclassical synthesis and Milton Friedman’s monetarism. But, as noted above, at the same time Leijonhufvud rejected Keynes’ liquidity preference theory of interest and other important aspects of the General Theory (Leijonhufvud 1979: 73). Instead, he committed himself to a “D. H. Robertson loanable funds theory” and to only accepting “half of the Keynesian revolution” (Leijonhufvud 1979: 73) and even calling his view the “Revised Standard Version of the General Theory (with the Liquidity Preference omitted as apocryphal)” (Leijonhufvud 1979: 77).

This is stated by Leijonhufvud as follows:
“The remaining, unexplored possibility of steering clear of both Cambridge and Chicago is that the interest rate mechanism (contra Mrs. Robinson) is capable of coordinating saving and investment but (contra Friedman) cannot be depended upon to do so always. This means a Loanable Funds theory with either speculation and/or the Central Bank capable of preventing or holding up adjustments of the real interest rate that is required for the equilibration of the system. Add temporary wage-inflexibility with consequent short-run quantity-adjustments and we are back with Z-theory.” (Leijonhufvud 1979: 60–61).
Leijonhufvud even accepted a natural rate of interest:
“The immediate consequence of retaining a basic Loanable funds position is clear …: the rate of interest will go to the new ‘natural’ level, and thus equate full employment saving and investment, unless bearish speculators (with or without the connivance of the Central Bank) intervene to prevent it. Consequently, it is possible for the interest rate mechanism to coordinate saving and investment decisions.” (Leijonhufvud 1979: 73).
In what Leijonhufvud called “normal cases” (when speculation does not affect interest rates) he even endorsed a monetarist policy position (Leijonhufvud 1979: 74).

It was only in exceptional circumstances that Leijonhufvud thought that economies require the Keynesian policy of discretionary fiscal policy (Leijonhufvud 1979: 75–77).

Leijonhufvud, A. 1979, “The Wicksell Connection: Variations on a Theme,” UCLA Working Paper No. 165. USA.

Leijonhufvud, Axel. 1981. “The Wicksell Connection: Variations on a Theme,” in Axel Leijonhufvud, Information and Coordination: Essays in Macroeconomic Theory. Oxford University Press, Oxford and New York. 131–202.


  1. I have been meaning to read this paper for a while and found your short encapsulation interesting, so thanks for posting.

    I have a question.Several times you refer to Wicksell's "loanable funds theory". I suppose this is to contrast it to "liquidity preference" theories. However Wicksell's theories clearly incorporate endogenous money concepts. And in a world of endogenous money don't the differences between loanable funds and liquidity preference becomes fudged ? For example: If at a given interest rate people increase their demand to hold money won't this be accommodated to some extent by an increase in the money supply in Wicksell's model that wouldn't happen under pure loanable funds ?

    1. (1) I call it "Wicksellian Loanable funds theory", because, yes, Wicksell took account of the banking system's ability to increase the money supply, as opposed to cruder versions that assume only a stock of commodity money (that is, no endogenously created credit money).

      (2) yes, Wicksell's theories do incorporate endogenous money concepts. Rogers (in Money, Interest and Capital 1989, pp. 25-26) notes this, but points out Wicksell's model is ambiguous: at some times it seems to be a pure credit money economy with perfectly elastic supply, and at other times a more realistic model of exogenous money, with credit money and the need for reserves restraining banks.

      (3) " If at a given interest rate people increase their demand to hold money won't this be accommodated to some extent by an increase in the money supply in Wicksell's model that wouldn't happen under pure loanable funds ?"

      Yes. Leijonhufvud discuses this on pp. 31-33.
      Wicksell seems to think that the money interest rate, with credit creating banks, can be determined by the "excess flow demand for loanable funds", and so excess demand for money governs the price level.

      But Wicksell thinks the banks should stick to a "neutral" policy "of just intermediating household saving without generating any net injection (or leakage) of loanable funds on its own" (p. 32)

      Download the paper here:

    2. Also further modern developments of Wicksellian loanable funds theory:

      Robertson, Dennis H. 1934. “Industrial Fluctuation and the Natural Rate of Interest,” The Economic Journal 44.176: 650–656.

      Robertson, Dennis H. 1966. Essays in Money and Interest. Collins, London.

  2. Good to see this paper being reviewed. His book Information and Coordination, is one of the most important macro books written in the post war period.

  3. "When nominal income is rising, investment exceeds saving by the net addition to loanable funds injected by banks"

    I think what he means here is that investment in nominal terms exceeds the supply of real savings (i.e. the accumulations of real things, such as iron, wood, etc, which are not directly consumed).

    This means that nominal investment demand exceeds the supply of the real goods which are demanded, causing their prices to rise.

    Investment in nominal terms equals savings in nominal terms, so I don't think he's talking about investment of money exceeding savings of money, as that wouldn't really make sense.

    I've seen this referred to elsewhere as 'forced saving'. The idea is that investors get the scarce goods by offering to pay a higher price, causing consumers to go without, as a result of them being outbid (i.e. consumers are 'forced' to save).