Saturday, November 03, 2012

The Historical Failure of Neoclassical Economics

"Both classical and marginalist economics provided accounts of the long-period (uniform rate of profit) theory of value and distribution, but whereas a classical economist could take the real wage as a datum for the purpose of such analysis (whatever the implicit ‘background’ theory of wages might be), the marginalist economist had to ‘close the system’ in some other manner. In effect, since ‘resource supplies’ were often taken as given, this meant that the ‘the supply of capital’ had to be taken as given, in one way or another. Just how the given supply of capital was to be represented was an issue which led to considerable heterogeneity amongst even those marginalist economists who shared the long-period method of analysis with the classical economists and with each other. That heterogeneity cannot be entered into here (see Kurz and Salvadori, 1995: 427-43) but it is now widely recognized that each version of such traditional long-period marginalist theory of value and distribution encountered insoluble problems (ibid: 443-8)." -- Ian Steedman (1998).
1.0 Introduction

The ‘neoclassical’ revolution is conventionally dated to the 1870s, with the works of William Stanley Jevons, Leon Walras, and Carl Menger. Neoclassical economists, from the 1870s to the 1930s, tried to develop neoclassical theory:

  • To include production, including production with previously produced means of production, within the scope of the theory.
  • To extend supply and demand-based reasoning to all runs, including the long run.

All long-run neoclassical models produced in this period failed; they were logically self-contradictory.

2.0 The Endowment of Capital in Early Neoclassical Theories

In long-run theory, relative spot prices are stationary in equilibrium1. In a long-run equilibrium, entrepreneurs have correctly anticipated effective demand, and the size of plants have been adapted to this demand. Furthermore, plants are being operated at their ideal capacity for which they have been designed2. These early economists can be said to have specified the given endowment of capital in two ways:

  1. As a vector of physical quantities of heterogeneous produced goods to be used as inputs in production.
  2. As a homogeneous quantity, given in value, but free to change its form.

Leon Walras adopted the first approach, even though his fully developed model contained a market for value capital. His approach comes to grief on the need to simultaneously assume a uniform rate of profit in all markets for produced goods, to equate supply prices and market prices of capital goods, and to impose the condition that capital goods are being produced in proportions that will allow the economy to be reproduced (perhaps on an expanded scale).

The second approach can be further subdivided. In the first subdivision, Jevons and Eugen von Böhm-Bawerk, for example, took the homogeneous stuff of which capital consists to be a fund of subsistence goods to maintain the workers while they labored. More capital somehow represents a longer period of production. These economists incorrectly thought that a meaningful physical measure of this period of production could be defined independently of prices and that a lower interest rate would necessarily encourage entrepreneurs to extend this period, given the available technology.

In the second subdivision, this homogeneous stuff consists of the value of a heterogeneous quantity of capital goods. This ignores price Wicksell effects, the variation of the value of a given collection of physical quantities of capital goods with distribution and prices. Knut Wicksell was both a prominent proponent of this approach and an early economist to realize why it does not work.

3.0 Later Developments

From around the end of the 1920s to the 1960s, neoclassical economists abandoned the long-run to concentrate on the refinement of certain general, logically consistent, although empirically empty, theories. I refer to the works of Erik Lindahl, J. R. Hicks, Friedrich Hayek, and Gerard Debreu and Kenneth Arrow on intertemporal and temporary equilibrium3. Towards the end of this second period, economists returned to the elaboration of long run theories of stationary and steady states. In the logically consistent multisectoral models in this trend, capital is not taken as given, either as a value quantity nor as a vector of physical quantities. Rather, the quantity of capital, in both senses, is found by solving the model4.

4.0 Conclusion

Economists have developed a logically consistent and empirically applicable theory of classical ‘natural prices’ (also known as Marxian ‘prices of production)’. As I and others have repetitively demonstrated, such prices are inconsistent with supply and demand-based reasoning. Since the endowment of means of production is not taken as given in such theories, these theories are not about the allocation of given resources among alternative ends.

Over, the last century economists have extensively explored the logic of models in which given resources are allocated among alternative ends. Although such models might be of use to a central planner, they seem to be unable to describe prices in actually existing capitalist economies.

The development of these claims have been available in the scholarly literature for about a third of a century. They have not been refuted. Most mainstream economists just ignore this collapse of neoclassical economics, in their teaching, in their applied work, in policy advice, and in their research.

  1. Long run equilibrium is compatible with slow, secular changes, such as improvements in technology and in the composition of output.
  2. Neoclassical economists, of course, did not claim that an actual economy would ever be in such a long-run equilibrium. The model was developed as an aid to analyze tendencies to equilibrium thought to be in existence at any given moment of time.
  3. I have seen some claim Irving Fisher as a forerunner for these theories.
  4. Many of the recent mainstream developers of models of endogenous growth (such as, Paul Romer) seem to be ignorant of this fact.


Tao Jonesing said...

The fact that neoclassical economic theory dominates economic thought and drive political decision-making, it is clear that neoclassical economic theory has been a wild success. The merits of the underlying theory are irrelevant to the power structure who paid for the development of neoclassical economic theory, and clearly the theory's lack of merit is irrelevant to most of the Western world.

Economics is politics. The point of neoclassical economics was to disappear rents from economic and political discussion (e.g., Henry George). Job well done.

Emil Bakkum said...

Perhaps the motivation is that neoclassical economics forms the backbone of contemporary economics as a science. Without it economics as a science would be similar to other social sciences.

Robert Vienneau said...

Thanks for the comments. I would not mind if economics were more similar to other social sciences.

windwheel said...

Does the earlier work of Graciella Chichilnisky & Heard offer a way forward? Put another way, by changing the focus to why and when preference diversity is too great or too little and/or to considering underlying the co-ordination problem for co-operative behaviour, it may be that the Neo-Classical model was somehow 'canonical', given what was available for Capital markets to price assets especially if they came from engineering or Actuarial backgrounds.
In this context, does Hilary Putnam's approach, or that he imputes to Amartya Sen, really point the way forward? Might it not transfer power to bureaucrats with Arts degrees and good sycophancy skills?
It is quite human, but not particularly helpful, to scold the Weather or the Economy or the Political System or Biology for not being nice- and Classical Economics did that just as well or better than Aristotle or the Church but would it really be able to sustain as good, if not better, a financial services industry as the (admittedly very bad) one that we have?
Grateful for any reading suggestions.