Tuesday, December 4, 2012

"More Heat than Light": the failure of modern economics

I'm going to turn now to Philip Mirowski's More Heat than Light: Economics as Social Physics, Physics as Nature's Economics, which is a quite devastating critique of mainstream neoclassical economics.

The heart of modern economics, he argues persuasively, was lifted wholesale from physics in the late nineteenth century. The trouble was that the main figures of the marginalist revolution, such as Walras, Jevons and Marshall, didn't quite understand all the math they imported into political economy.

The neoclassical founders almost all came from an engineering or natural science background. But they had a limited grasp of the state of the art of physics even at the time. Above all, says Mirowski, they failed to understand the importance of conservation principles in the math. To accurately measure change, something must stay the same. That means most of the edifice of neoclassical economics is based on stale physics contaminated by basic errors.


Substance and fields

What happened was basically this. Natural scientists struggled in the early nineteenth century with ideas of heat and motion, imagining fluids or ethers or substances. By the 1870s, that had given way to a unified view centered on energy, and the conservation of energy as it was transformed from one kind to another. Instead of fluids or other kinds of substance, physicists now thought of fields and forces, and worked out the vector math of kinetic and potential energy.

The heart of neoclassical economics, says Mirowski, is that economists replaced energy with utility in the same equations, and lifted the framework wholesale. The marginalist revoluton paralled the revolution in physics in preceding decades. Classical economists saw Value as a substance, such as the equivalent of wheat for the physiocrats or the labor theory of value for Ricardo and Marx. But for neoclassicals, Value was a field, like electromagnetism in physics. Kinetic energy was essentially spending and income; potential energy was utility.

He quotes several of the major marginalists who explicitly acknowledged that this is how they thought. But later economists mostly forgot these origins. The discipline has never been that historically self-conscious.

There were two main problems with all this, however. First, without a conservation principle, the math didn't work. Conserving energy implied income and utility were a constant - so essentially the same thing. That would mean utility would be superfluous as a separate measure to money, which was not at all desirable. The point was often lost in a technical debate about "integrability". Leading physicists tried to explain the point to economists, who appeared to have been mostly baffled and nonplussed at the argument.

Secondly, physics moved on from its 1870-vintage "proto-energetics" state, as Mirowski termed it. The second law of thermodynamics implied entropy was always increasing, so interactions were not easily reversible. Special relativity, general relativity and quantum theory all upset the mechanical "Laplacian dream" of 1870s physics, bringing frames of reference, probability, indeterminacy and the role of the observer into the picture. Symmetries and conservation principles could be broken. Matter could decay. Particles could pop in and out of existence. In contrast to notions of inherent "scarcity", the whole universe might be a "free lunch", something which came from a temporary variation in nothing.

None of these could easily be incorporated into the neoclassical framework. However, economists insisted all the more stridently that they were pursuing disciplined science, in contrast to sociologists or anthropologists, while actual scientists were increasingly doing something quite different.

As long as the Laplacian Dream was their dream, they clutched neurotically at their portrait of persons as irrotational mental fields suffusing an independent commodity space, as science ebbed ever further away toward a world subject to change, diversity and indeterminacy, and at one with the observer. P275

And the outcome?

In brief, the practical dissolution of the energy concept in advanced twentieth-century physics has painted neoclassical economics into a corner. p388

Mirowski wrote the book in 1989. Of course, the metaphor of utility as (potential) energy looks even more strained today. We now know that the visible universe of 1870s physics is only 4% of the universe. The rest is dark matter and dark energy that we cannot as yet observe and don't understand.

So what? Some mainstream economists concede to his argument about the origin of the neoclassical model, notes Mirowski, but they claim it is not relevant to the subsequent evolution of the discipline.

But it is. They still want the appearance of science, while being stuck with a model which is increasingly divergent from science in reality, he claims. To talk about analogies to entropy, said Samuelson, for example, is always the mark of a crank. But Mirowski points out that Samuelson frequently published articles with tenuous links to physics himself. Indeed, the key to Samuelson's career was maintaining the appearance of scientism.

Economists have produced various ad-hoc conservation principles in the twentieth century, according to Mirowski, "but in the final analysis this is all one big shell game, with the offending conservation principles passed from one assumption to another." p274



The metaphor of utility as an energy field is too embedded to be given up by neoclassical economics, he says The trouble is it is also a metaphor of instantaneous exchange, and as such it has proved consistently difficult to reconcile with production, which had been the focus of classical economics. Classical economics thought that value was created in production, circulated in trade, and consumed in consumption. Neoclassical economics was focused on exchange, and found it hard to explain production at all.

That inconsistency explains a proliferation of production functions in postwar economics, and difficulties with temporarily and the existence of firms through time.

Economists have effectively tried to reinvent a substance theory when it comes to production, says Mirowksi. But this is bound to be inconsistent with utility as a potential energy field. So the profession has not been able to settle on a satisfactory answer.

The situation was embarrassing, but no neoclassical was willing to come right out and say that production was superfluous or irrelevant in their scheme of things. (Lionel Robbins came the closest). p272


There are also implications for scarcity. The idea of scarcity as the heart of the economic (and human) condition was largely an artifact of the neoclassical approach, he says.

Prior to that time, scarcity as some sort of primordial state of mankind did not play any signficant role in the value theory of classical political economy. Only with the dominant impression that Nature enforced a general state of dearth, say, rather than the physiocratic notion of Nature's bounty, could it become possible to even think of economic equilbrium as a state of psychological counterpoise, hemmed in by the urgent necessity to clear markets in a state of stringent limitations. p240

This caused obvious problems.

The metaphor of utility as potential energy was predicated upon a Weltanschauung of a closed, bounded system that exemplified the natural state of mankind as enduring ineluctable scarcity. If and when production was to be introduced into this morality play, it had to be done in such a way as to prevent the contravention of the scarcity principle, all the while maintaining the field theory of value. p293

Of course, I think this is fascinating given I think one of our main challenges now is thinking through the implications of abundance.

Keynes, he says. succeeded in introducing a kind of value substance in the guise of "national income". That allowed the idea of an economic process to be reintroduced. There has been an immense effort in recent decades to link macro with rational choice "microfoundations." But this is doomed to failure, says Mirowski.

Keynes generated a theory of an unstable economic process by the instrumentality of his reversion to a substance theory of value, a tactic that allowed the joint conceptualization of production, growth and the passage of time in (relatively) internally consistent manner. In contrast, it is the avowed intention of the microfoundations school to renounce all value substances and to recast all macroeconomic analysis in the format of production and utility fields. It is precisely this choice that prohibits the logical modeling of process in priduction, in growth, and in exchange, as explained earlier in this chapter. The field metaphor cannot represent a circular economy where outputs become inputs and so on, ad infinitum. It cannot specify precisely what it is that grows in an economy. p346

Fields are just not suitable when time is involved.

.. the formalism of the field is useful only in cases where one can safely abstract away all considerations of process and the passage of time. p346


Mid-twentieth century neoclassical economics would have found an alternative to utility if it could, he says. The mainstream forgot how widespread concern about utility had become in the profession before the second world war. Mirowski quoted Viner as saying economists' understanding of utility was comparable to "the knowledge of heat prior to the discovery of the thermometer." It could not be satisfactorily measured. There had always been concern at how locating value in a purely mental framework came close to idealism or solipsism.

The development of the indifference curve approach and Samuelson's theory of revealed preference were not durable responses, either, and only served to obscure the origin of the utility metaphor. Revealed preference was not empirically tractable, for one thing, and confused preferences and behavior.

In the absence of the metaphor of utility as nineteenth-century potential energy, there is no alternative theory of value, no heuristic guide to research, no principle on which to base mathematical formalism, no causal invariant in the Meyersonian sense, and most threatneing, no basis for the claim that economics has finally become scientific. p368

So what does this lead to? There is, Mirowski says, no scientific method that can guarantee economics' scientific status.

This lesson is the legacy of the decline of positivist philosophies of science in the late twentieth century. Juxtapose this fact with the hypothesis that economic research has always met with the greatest difficulties in establishing the credibiliy of its results and fending off charges of charlatanism and quackery. p357

It is hard to revise the neoclassical framework without undermining it. New approaches do not have that problem. This means, he says, neoclassical economics will be vulnerable to new contenders for the role of social physics.

Theories of Value

And is there an alternative theory of value? Mirowski says there are two main alternatives. One is to deny any separate value, rarely advocated, but as represented by someone called Samuel Bailey (who I have never heard of).

This position argues that no economic phenomenon is conserved through time, and therefore scientific analysis is impossible. Whatever one might think of the truth of this option, it should be clear than the nihilism inherent in the program assures that in this instance there can be no legitimate research program called economics. p400

The other alternative is a "social theory of value", he says, based not on scientific or social metaphors , but in social institutions such as accounting conventions or property rights.

I doubt myself whether this is true. I imagine evolution is the main contender for an alternative scientifc framework, together with the notion of adaptability and "fitness" of some kind or another.


Overall, it is a very bracing read. It seems, at least to me, highly persuasive - but I would want to read some reviews and responses to make sure I am not overlooking flaws in Mirowski's own analysis. .

What it underlines is that utility and scarcity were chosen not so much because of their psychological or social accuracy, but because the math "worked". And if the math worked there was more scientific respectability. I have always had the firm impression that this was the driving force of major parts of the discipline, which is likely the main reason I did not become an academic economist. It did not ring true. It was about the aesthetics of models rather than genuine insight. It was about a particular quasi-religious view of rationality rather than solving problems.

The book is also highly illuminating , not to say shocking, about the origins of utility in modern economics. I've often talked before about how ethical theory went off the rails in the eighteenth and nineteenth centuries, dropping the older tradition of the virtues and the good life for a more utilitarian, neutral and welfarist approach. Mirowski excavates a much deeper layer of intellectual history underlying current economics. It was not a matter of an import from Bentham. It was an import from physics, and just more or less happened to be called utility.

The metaphor of potential energy as a utility field locked economics into an increasingly less productive path for a century - and to a large extent still does. I knew most of the arguments about indifference curves, production functions and revealed preference, of course, but I was much less familiar with the intellectual history of the arguments. It is fascinating. And disturbing.

Perhaps most of all, it shows how value theory is the great unsolved problem at the heart of economics. That is what I have been grasping toward in my own terms on this blog. To understand the future of the economy , we have to be back up into ethics and the question of the good life and human flourishing. That, after all, is the only place a valid notion of value can come from.


No comments:

Post a Comment