From a paper by Arash Vassei:
Hayek attacks the Böhm-Bawerkian formulation of intertemporal choice in “Utility Theory and Interest”, published in 1936. Hayek attributes the ‘confusion’ characterizing the subsequent controversies about the proper relationship between Böhm-Bawerk’s three ingredients, especially the relation between subjective and objective data, to the original formulation of the problem. The ‘pseudo-problem’ of a positive net return on capital in equilibrium, so Hayek, roots in the improper conceptualisation of intertemporal choice. He criticizes the Böhm- Bawerkian framework for treating the utility function as primitive by resorting to the comparison of absolute marginal utilities (Hayek, 1936: 48). Marginal utilities are not simultaneously determined by subjective and objective data, including, besides preferences, the state of technology (the book of blueprint) and the levels and structure of endowment (ibid.: 44-5).
Also in Mises’s outline, marginal utilities are purely intrinsic phenomena. They are introduced independent of relative physical scarcities and are as such given data to the economic system. Indeed, even though Mises categorizes action as a choice among alternative means to a variety of ends, scarcity is introduced as an absolute. It is because any quantity of a given good ‘can only produce a limited effect, some things are considered scarce and treated as means’ (Mises 1998: 120). Mises’s ordinal measure of this ‘limited effect’, his utility function, is negative. He refers to ‘felt uneasiness’, which is to be minimized. Utility means in this context simply this: causal relevance for the removal of felt uneasiness. (Mises 1998: 120)
The term ‘causal’ is meant to exclude simultaneous determination. Marginal utilities are disproportionally attached to ‘the various portions of a supply of homogenous means (diminishing marginal utility). Each portion is valued separately’ (ibid.: italics mine). Small quantities of any given stock satisfy the most important pleasures or reduce the most unbearable pain. High quantities can additionally satisfy less urgent ends or reduce more tolerable disutility.
This, of course, affects the conceptualization of general equilibrium. Prices coordinate the excess demand relations of consumers. In an exchange economy, each consumer’s excess demand relation is directly derived from the primitive utility function. Single prices are thus independent of each other, and relative prices therefore also primitive. The only mechanism needed is the law of one price, in operation only in disequilibrium. Thus, market process theory is treated as a substitute for general equilibrium analysis. This is also true for Mises’s treatment of intertemporal preferences. Indeed, it is exactly here where we find the very hawkish prediction of his business cycle theory: time preference as a primitive, directly determining optimal intertemporal allocation independent of any other economic data, is responsible for his overemphasis on consumer sovereignty and mean-reversion.
Hayek’s critique highlights the consequences of such utility analysis for Böhm- Bawerk’s theory of intertemporal choice, especially given the choice to assume total consumption over time to be homogenous. Thus, according to the underlying utility framework, intertemporal exchange is reduced to analysis of the law of one price. At zero rate of time preference, general equilibrium is restored on a single market for a given total quantity of homogenous supply. Because physically identical commodities at different dates are treated as perfect substitutes, the equilibrium condition is fulfilled at equal absolute marginal utilities of present and future consumption for each date, indicating stationary equilibrium consumption at constant prices (Hayek, 1936: 44-8).
Mises theory of interest attempts to provide reason for the fact that the law of one price fails to establish a single price and resumes to Böhm-Bawerk’s second cause. The relative intertemporal price, the rate of interest, is directly imposed by the rate of time preference, the exogenous subjective rate of discount on future consumption. The ‘originary’ rate of interest is thus a pure value phenomenon, independent of the level and growth rate of consumption, identical with the prevailing rate of time preference. Objective data is unable to alter intertemporal equilibrium allocation. Whatever the supply of total consumption, and whatever its initial distribution in time, the long-run equilibrium allocation is unidirectionally determined by an unexplained rate of time preference.
In stationarity, Böhm-Bawerk’s first cause of interest is absent. Right because its sign is contingent on the data of the system, Mises dismisses the rate of time preference due to changing future consumption (Mises 1940: 442–4). Instead the second source, the systematic undervaluation of future consumption, is introduced as a categorical aspect of human action so that “the very act of gratifying a desire implies the gratification at the present instant is preferred to that at a later instant” (Mises 1998: 481). Thus, Mises superimposes the intertemporal preference relation. In fact, however, the dominant role of a positive rate of time preference is an embarrassment to Mises’s analysis (Hayek 1984: 55). In marginalist analysis it is illegitimate to presume that agents categorically discount future consumption at a constant rate whatever other primitives are, as it is illegitimate to assume ad hoc that agents always prefer apples to bananas, independent of other data. As Hayek points out:
“The fact is that the assumption of identical utility curves at all successive moments states not merely the general postulate that the choice between present and future will be made in the same way at different moments, but implies beyond that that the choice will be made in a particular way. […] It is impossible to decide on a priori grounds what the actual attitude of the person in any given position will be.”
(Hayek, 1936: 48, 51; his italics)