The Clower Leijonhufvud version

Greenfield and Yeager's monetary disequilibrium story, which finds its roots in the early American monetarists, is paralleled by the work of Robert Clower and Axel Leijonhufvud that emerged in the 1960s in their attempts to come to grips with the Keynesian revolution. In many ways, the theoretical stories each group offers share some fundamental characteristics, including placing money as the medium of exchange at center-stage (and by implication the banking system as well) and emphasizing the stickiness of prices in response to excesses or deficiencies in the supply of money. One notable difference between Leijonhufvud's work and Yeager's is the former's more prominent emphasis on the Wicksellian natural rate/market rate mechanism. For Leijonhufvud, it is this 'Wicksell connection' (1981b) that links his perspective with Keynes' work, albeit Keynes of the Treatise rather than Keynes of the General Theory. In this section, I want to explore the Clower-Leijonhufvud version of the deflationary cumulative rot and the ways it builds off of the very Wicksellian monetary equilibrium framework developed in Chapter 3.

Clower's key contribution to this line of thought is the notion of the dual-decision hypothesis and its corollary dictum that 'money buys goods and goods buy money, but goods do not buy goods in an organized market' (1984b [1970]:100). The dual-decision hypothesis simply recognizes the fact that decisions to buy are intimately linked with decisions to sell in both equilibrium and disequilibrium. In equilibrium, planned sales constrain planned purchases, as the former are the source for the latter. Given that equilibrium is defined in terms of complete compatibility of plans, this 'constraint' does not prevent anyone from executing their desired plans: we sell what we plan to and we buy what we plan to. However, in disequilibrium, this is no longer the case. Here, some, if not many, plans will not be completed and it is not necessarily true that one is able to do what one plans. In disequilibrium the constraint upon one's ability to actually realize particular purchases is one's realized sales. Once one excess demand appears somewhere in the system, there will be implications for the realization of all other planned purchases and sales. Some, if not many, people will be unable to realize their plans.

As Clower carefully points out, this point does not matter in the timeless and instantaneous world of equilibrium. However, when one takes a more process-oriented perspective, and sees that decisions to buy and sell are not made simultaneously but in sequences, then the fact that realized sales constrain planned consumption poses some interesting problems. To take Clower's (1984a [19631:48) example, suppose we have a person who is unable to realize sales of his labor services, that is, someone who is involuntarily unemployed in Keynes' sense. The problem is that he cannot consume without first selling his labor services. If sellers knew of his desires to buy, they would be willing to hire him and if he could get hired, he would, in turn, consume. Clower asks how this sort of signal gets sent through the marketplace. His answer is that it happens in a variety of ways, from lowering our reservation wage, to drawing on savings, to reducing our consumption substantially. Clower concludes from this that orthodox equilibrium analysis cannot explain the situation at hand:

For if current receipts are considered to impose any kind of constraint on current consumption plans, planned consumption as expressed in effective market offers to buy will necessarily be less than desired consumption as given by the demand functions of orthodox analysis.

Because we might choose to lower our consumption, or are forced to because we accept a wage lower than that which corresponds to our real productivity, our effective demands in the market are less than the demands we would have in equilibrium, where such demands would be driven by our being able to sell our labor services at their 'full' value. This condition is sometimes expressed as our effective demands being less than our 'notional' demands, where notional demand is based on the actual value of our labor's marginal product, rather than what the wages we are actually able to command at a particular instant in the market.5

The importance of this insight, according to Clower, is that these excess demands (the difference between our notional and effective demand) will not be registered anywhere in the market. Thus there is an informational breakdown between the sellers of labor services and the sellers of commodities. If those notional demands could be made effective, and therefore 'register' in the market, the buyers of labor services/sellers of commodities would be willing to hire the labor that was signaling its willingness to buy goods and services. This signaling process is further complicated by money's role as a medium of exchange. In barter, of course, all acts of demand are simultaneous acts of supply. The dual-decision hypothesis would still be relevant in a barter world, but in a money-using world, the ability to temporarily store the value of one's realized sales in the form of money may well weaken the ability of commodity buyers to signal their notional demands to commodity sellers.

Leijonhufvud's work, particularly the two essays 'Effective Demand Failures' (1981a) and 'The Wicksell Connection: Variations on a Theme' (1981b), takes Clower's analysis of the dual-decision hypothesis and places it in a Wicksellian framework. More specifically, in line with Keynes' analysis in the Treatise, Leijonhufvud is interested in describing the scenario where the market rate of interest is above the natural rate, what we have termed the deflationary monetary disequilibrium.6 One fundamental concept in Leijonhufvud's work is the notion of the 'corridor'. He argues that within a certain corridor, market price signals generally work well and that deviations from monetary equilibrium will be self-corrected in a relatively short period of time, although not without some real effects. Outside of that corridor, when the exogenous changes impinging on the system are large enough, the Clower-like signaling failure comes into play and, in these scenarios, we cannot be confident at all that market mechanisms will be able to correct the problem. As a result, Keynesian-type unemployment equilibria may occur. Leijonhufvud argues that his analysis in 'The Wicksell Connection' is largely about what happens inside the corridor, while the 'Effective Demand Failures' paper explores what happens outside of the corridor.

Leijonhufvud's own theory, what he terms the 'Z-theory' interpolation between Keynes of the Treatise and Keynes of the General Theory, attempts to analyze the effects of a deflationary impulse. The focus of that theory is the Wicksellian insight that the central macroeconomic phenomenon is the relationship between the natural and market rates of interest. That relationship depends, in turn, on the relationship between savings and investment. As Leijonhufvud cogently argues, the Wicksellian tradition is defined by its concern about what conditions are necessary for the savings—investment equality to hold and what happens when it does not. This is in contrast to what he terms the Quantity Theory tradition, leading from Fisher up to Friedman, which ignores these interest rate questions because it 'assumes that the interest rate mechanism can be relied upon to coordinate the intertemporal decisions of households and firms' (1981b:132).7 Leijonhufvud also argues that this Wicksellian insight has been lost not only due to the rise of interest rate-free monetarism, but also because Keynes himself progressively eliminated the central role of the banking system in determining interest rates in his intellectual shift from the Treatise to the General Theory. Specifically, it is the liquidity preference theory of interest that drives the final nail in the Wicksellian coffin. As we have noted earlier, it is Keynes' assumption that saving and investment are continually equal that removes the banking system and leaves only liquidity preference as the interest rate determiner. As Leijonhufvud (1981b:170-1) puts it:

If saving and investment are continually equal, the rate of interest cannot possibly be governed by any difference between them. The possibility of a corresponding excess flow demand for loanable funds has then also been defined away. The loanable funds interest rate mechanism is gutted. Hence, the flow part of the Treatises stock-flow analysis should be erased. The speculative element remaining from it now has to make do as a complete interest theory.

This is the result that Leijonhufvud wishes to avoid.

The Z-theory perspective begins with a hypothesized decline in the marginal efficiency of capital, that is, entrepreneurs turn pessimistic. The demand for loanable funds (investment) declines, while saving remains the same (a shift from I to I' in Figure 5.3). This corresponds to a decline in the natural rate of interest (from nr to nr'), as the pessimism of investors means that they will require lower interest rates in order to undertake projects they would have previously undertaken at higher rates. In the Z-theory model, this decline in the natural rate is not immediately transferred into a corresponding fall in the market rate, leaving the market rate above the natural rate. A fall in income and employment ensues, for either the kinds of reasons outlined in the previous section or a more Keynesian multiplier mechanism. In a truly Wicksellian framework, the fall in income and prices would eventually restore monetary equilibrium at the lower natural rate, but only after a painful adjustment process.

Where Leijonhufvud sees the Z-theory as breaking with Wicksell is over whether this endogenous adjustment need always occur. One central difference is that Leijonhufvud argues that the fall in income drags down savings with it (from S to S'), enabling the securities market to clear at a market interest (mr') rate still above the natural rate (nr'), i.e., the rate that would have held had market mechanisms fully adjusted. The problem here is that the securities market is now cleared, thanks to the income-induced decline in savings, and there is no 'excess flow supply of loanable funds whose accumulating timeintegral progressively distorts the balance-sheets of banks and/or bearish speculators' (1981b:166). Whereas in the truly Wicksellian model, the excess supply of savings at the too-high market rate generates pressure on banks to

Figure 53 Leijonhufvud's Z-theory in the loanable funds market

lower those market rates in response, in the Z-theory model of Leijonhufvud, the income-induced decline in saving can cause the money/securities market to clear at a too-high market rate, with no consequent pressure on anyone to make the needed interest rate adjustments. The key to this result is that the original decline in investment is not immediately offset by price adjustments in the form of the banking system promptly reducing its market rate in response. Absent such price adjustments, output, employment, and income (and therefore saving) will bear the burden of adjustment.

Leijonhufvud does argue that it would take a sizeable shock to generate a reaction adverse enough to throw the economy that far out of whack. The problem faced when deflationary monetary disequilibria get underway is that quickly multiplying effective demand failures can take place. As Clower's work emphasized, realized sales are a constraint on purchases, the unemployment and general excess supply of goods that will accompany the scenario Leijonhufvud lays out may create further problems. To the extent that the current inability to sell labor services prevents workers from purchasing commodities, the sellers of those commodities see their own buying power constrained by the lack of output sales, further constraining other sellers. If income falls enough at the outset, the situation can rapidly turn into a cumulative rot, one which Leijonhufvud believes will require activist fiscal policy to remedy.

However, one cannot neglect the role of cash-balances here. Leijonhufvud (1981a:117ff.) argues that stocks of liquid assets, such as cash, 'allow expenditures to be maintained when receipts fall off. These asset stocks are referred to as 'buffers' and they prevent burgeoning effective demand failures from turning into a Wicksellian rot by providing a source of expenditures for the un- or underemployed. The key is that actors hold such buffer stocks based on their expectations of the system's stability. If the size or length of the disturbance is so large that the buffer stocks being held are insufficient, then the rot will set in and policy will be necessary, according to Leijonhufvud. It is these buffers that define what he calls the 'corridor'. As long as such stocks are sufficient, normal market processes will be sufficient to avoid massive excess demand failures, even in the face of exogenous events that cause monetary disequilibria. That is not to say that such disequilibria will be costless, only that eventually, and with some pain, they will be able to self-correct. Outside of the corridor defined by those buffer stocks, the self-correction mechanisms weaken and disappear and systemic breakdowns are possible. This implies 'a variable width of the corridor' as actors will hold buffer stocks based on their past experiences and expectations of the future (1981a:123).

It is here where Leijonhufvud's perspective overlaps with Yeager's. Both are concerned about the cumulative downturn that can result from a monetary disequilibrium. For Yeager, the originating impulse is normally some error by the banking system that inappropriately reduces the supply of money. For Leijonhufvud, the problem seems to lie more in the banking system's inability to respond appropriately to changes in the factors underlying the natural rate. Despite this very important difference, and it is one that explains Yeager's focus on the role of money and Leijonhufvud's emphasis on the loanable funds market, their explanations of how a problem in the money/loanable funds sector can spill over into the real sector are quite similar.

There will be more to say about Leijonhufvud's theory in the next chapter's discussion of the labor market. For now, one important point is in order. On the surface, the Z-theory approach does not seem like a monetary theory of depression. Leijonhufvud assumes, as Keynes did, that the originating factor is a real disturbance in the form of entrepreneurial pessimism. This disturbs the natural rate, setting up the monetary disequilibrium. From the perspective of this study, however, this is indeed a monetary problem. It has been our contention that it is the responsibility of the banking system to maintain the market rate/natural rate equality. If the natural rate should move for any reason, a properly functioning banking system should respond with a corresponding adjustment in the market rate. Lying hidden in Leijonhufvud's theory is the question of why the banking system does not respond appropriately. Why, for example, does the banking system not, in response to balking borrowers, offer them lower interest rates to buy off their pessimism? This strategy would also reduce the quantity of loanable funds supplied, closing the gap between saving and investment. If the banking system could respond this way well before speculators began to worry about the direction of interest rate movements, or better yet, if speculators knew that a market rate response was quickly forthcoming, then the cumulative process could be choked off, perhaps with only a minimum of buffer stocks necessary. What Leijonhufvud does not point out sufficiently is that the size of the corridor might also depend upon the monetary regime and its perceived ability to make appropriate responses to changes in the natural rate, which could also affect the quantity of buffer stocks that the public holds.

Continue reading here: Austrian economics and 100 percent reserve banking

Was this article helpful?