Sunday, July 15, 2007

Kaldor On Neoclassical Economics

What could Alan Blinder have meant when the N. Y. Times quoted him as saying, "Mathematics is not scientific"? Perhaps Blinder was saying something like Kaldor did a over a quarter century ago:
"...unlike any scientific theory, where the basic assumptions are chosen on the basis of direct observation of the phenomena the behaviour of which forms the subject-matter of the theory, the basic assumptions of economic theory are either of a kind that are unverifiable - such as that producers 'maximize' their profits or consumers 'maximize' their utility - or of a kind which are directly contradicted by observation - for example, perfect competition, perfect divisibility, linear homogeneous and continuously differentiable production functions, wholly impersonal market relations, exclusive role of prices in information flows and perfect knowledge of all relevant prices by all agents and perfect foresight. There is also the requirement of a constant and unchanging set of products (goods) and of a constant and unchanging set of processes of production (or production functions) over time - though neither category, goods nor processes, is operationally defined: in other words, no attempt is made to show how these axiomatic concepts are to be defined or recognised in relation to empirical material.

While this pure theory is not intended to describe reality, it is put forward as the necessary conceptual framework - the necessary starting point - for any attempt at explaining how a 'decentralised' system works; how individuals guided entirely by the market, or rather by price information, sort themselves out between different activities and thereby secure the maximum satisfaction both to themselves and, in the specific Pareto-sense, to society as a whole.

Indeed it is the deep underlying belief, common to all economists of the so-called 'neo-classical' school, that general equilibrium theory is the one and only starting point for any logically consistent explanation of the behaviour of de-centralised economic systems. This belief sustained the theory despite the increasing (not diminishing) arbitariness of its basic assumptions - which was forced upon its practitioners by the ever more precise cognition of the needs of logical consistency. In terms of gradually converting an 'intellectual experiment' (to use Professor Kornai's phrase) into a scientific theory - in other words, into a set of theorems directly related to observable phenomena (*) - the development of theoretical economics was one of continual degress, not progress: the ship appears to be much further away from the shore now than it appeared to its originators in the nineteenth century. The latest theoretical models, which attempt to construct an equilibrium path through time with all prices for all periods fully determined at the start under the assumption that everyone foresees future prices correctly to eternity, require far mor fundamental 'relaxations' for their applicability than was thought to be involved in the original Walrasian scheme. The process of removing the 'scaffolding,' as the saying goes, - in other words of 'relaxing' the unreal basic assumptions - has not yet started. Indeed, the scaffolding gets thicker and more impenetrable with each successive reformulation of the theory, with growing uncertainty as to whether there is a solid building underneath.

Yet the main lessons of these increasingly abstract and unreal theoretical constructions are also increasingly taken on trust - as if in the social sciences, unlike the natural sciences, the problem of verification could be passed over or simply ignored. It is generally taken for granted by the great majority of academic economists that the economy always approaches, or is near to, a state of 'equilibrium'; that equilibrium, and hence the near-actual state of the world, provides goods and services to the maximum degree consistent with available resources; that there is full and efficient utilisation of every kind of 'resource'; that the wage of every kind and quality of labour is a measure of the net contribution (per unit) of these varying kinds and qualities of labour to the total product; that the rate of profits reflects the net advantage of substituting capital for labour in production, etc., etc. - all propostitions which the pure mathematical economist has shown to be valid only on assumptions that are manifestly unreal - that is to say, directly contrary to experience and not just 'abstract'. In fact, equilibrium theory has reached the stage where the pure theorist has successfully (though perhaps inadvertently) demonstrated that the main implications of this theory cannot possibly hold in reality, but that has not yet managed to pass his message down the line to the textbook writer and to the classroom.

(*) The difference between a scientific theory and an 'axiomatic' theorem has been well put by Einstein:
'Physics consitute a logical system of thought which is in a state of evolution, whose basis cannot be distilled, as it were, from experience by an inductive method, but can only be arrived at by free invention. The justification (truth content) of the system rests in the verification of the derived propositions by sense experiences.

The skeptic will say: "it may be true that this system of equations is reasonable from a logical standpoint. But it does not prove that it corresponds to nature." You are right, dear skeptic. Experience alone can decide on truth.'
A. Einstein, Ideas and Opinions, New York, 1960, pp. 322 and 355 (quoted by Kornai...)
The difference mainly resides in this. In the case of physics, any fundamental re-consideration of the basic 'axioms' of the system is the result of observations which could not be made consistent with existing hypotheses. Examples (chosen at random) are the observation that the amount of radiation emitted by Pitchblende was greater than could be accounted for by the absorption of sunlight; that a stream of light which passed through a glass and was directed at a mirror at some particular angle is not reflected by the mirror; or that there is a 'reddening' of the spectrum observed in distant stars. In economics, observations which contradict the basic hypotheses of prevailing theory are generally ignored: the 'theorist' and the 'empiricist' operate in two isolated compartments and the challenge of anomalous observations is ignored by the theorist - as something that could be taken into account at the stage of 'second approximation' without affecting the basic hypotheses. And where empirical material is brought into conjunction with a theoretical model, as in econometrics, the role of empirical estimation is to 'illustrate' or to 'decorate' the theory, not to provide support to basic hypothesis (as for example, in the case of numerous sudies purporting to estimate the coefficients of production functions)." -- Nicholas Kaldor (1972). "The Irrelevance of Equilibrium Economics", V. 82, N. 328 (Dec.): 1237-1255.


Gabriel M said...

Just because you can axiomatize it, it doesn't make it true or practically meaningful. For example, Sen's possibility of interpersonal comparisons claim.

Anonymous said...

I think my favourite von Mises quote should be shared:

"If a contradiction appears between a theory and experience, we must always assume that a condition pre-supposed by the theory was not present, or else there is some error in our observation. The disagreement between the theory and the facts of experience frequently forces us to think through the problems of the theory again. But so long as a rethinking of the theory uncovers no errors in our thinking, we are not entitled to doubt its truth"

My second favourite is:

"[It] cannot be denied that Fascism and similar movements aiming at the establishment of dictatorships are full of the best intentions and that their intervention has, for the moment, saved European civilisation. The merit that Fascism has thereby won for itself will live eternally in history."

Robert Vienneau said...

Von Mises probably sounds better in the original german.