Short-Period and Long-Period Analysis: Neoclassical Versus Historical
by Philip Pilkington
In my previous post I was concerned with summarising Lawson’s argument regarding the term “neoclassical” for an audience that was not going to read his paper in full. Thus, I did not wish to insert too much of my own thoughts on the matter. However, I would now like to deal with something which I only mentioned in passing in yesterday’s post.
Toward the end of the paper Lawson makes a distinction between three different groups. He writes:
In short, I am suggesting that there are three basic divisions of modern economics that can be discerned in the actual practices of modern economists. These are:
1) those who both (i) adopt an overly taxonomic approach to science, a group dominated in modern times by those that accept mathematical deductivism as an orientation to science for us all, and (ii) effectively regard any stance that questions this approach, whatever the basis, as inevitably misguided;
2) those who are aware that social reality is of a causal-processual nature as elaborated above, who prioritise the goal of being realistic, and who fashion methods in the light of this ontological understanding and thereby recognise the limited scope for any taxonomic science, not least any that relies on methods of mathematical deductive modelling; and
3) those who are aware (at some level) that social reality is of a causal-processual nature as elaborated above, who prioritise the goal of being realistic, and yet who fail themselves fully to recognise or to accept the limited scope for any overly-taxonomic approach including, in particular, one that makes significant use of methods of mathematical deductive modelling.
Lawson points out that it is the first group that makes up much of the mainstream; the second group that makes up what Lawson calls the “core of heterodoxy”; and the third group that would fall into the neoclassical camp if we understand Veblen’s original use of the term.
Now, although any labels we apply to these groups will be to some extent arbitrary I think that by considering the three groups from a different historical angle might help to get a grip on how we might begin to get our heads around the new taxonomy which Lawson has unearthed. This historical angle will be that of Joan Robinson who, in retrospect, was a serial offender in the misuse of the term neoclassical but who also had an extremely good intuition regarding the differences Lawson has uncovered.
In her book Economic Heresies: Some Old-Fashioned Questions in Economic Theory Robinson tries to give some coherence to what she sees as the different trends in economics over the course of its history (since Ricardo, anyway). In this book she draws a clear distinction between the doctrines of Walras and those of Marshall and indicates that it is out of the latter that Keynesianism proper (what we would today call “Post-Keynesianism”) emerged. Robinson finds the key difference in this regard in how Walras and Marshall treat time. She writes:
The essential idea is that a short-period situation [in Marshall] is one in which productive capacity happens to be whatever it is. But a situation with a specific plant in existence today is not to be identified with the Walrasian concept of a given stock of factors of production; its role in analysis is quite different. Unlike the Walrasian concept, Marshall’s short-period is a moment in a stream of time in which expectations about the future are influencing present conduct, and it belongs to a monetary economy in which the division of proceeds between wages and profits emerges from the relation of money prices to money-wage rates. With the aid of this concept, we can analyze price policy in imperfect competition, the effects in the present of uncertainty about the future, and the meaning of equilibrium in a process of growth, all of which are ruled out by the assumption of a Walrasian market. (Pp17-18)
A number of comments are here in order. Clearly Robinson is recognising in Marshall what Veblen thought to be the hallmark of neoclassical economics: namely, its partial recognition of the “stream of time in which expectations about the future are influencing present conduct” but its remaining in the realm of static analysis where various equilibria must be compared. This is, of course, what is often called the Marshallian partial equilibrium approach. Walras, by contrast, is no good for even partial equilibrium analysis. The Walrasian system is one of general, not partial, equilibrium.
Thus we can recognise in Walras a characteristic that Veblen assigned to the classicals; that is, a tendency to think in terms of pure teleology. While we can recognise in Marshall the very thing that led Veblen to coin the term “neoclassical”; that is, the uneasy existence of a recognition of historical time together with the use of a methodology suited to a static analysis.
We also see this “neoclassical” aspect of Marshall in how he thinks of the “long-period”. Robinson continues:
We can make use of the distinction between the long- and the short-period concepts without being committed to any faith in equilibrium being established in the long run. Indeed, it is absurd to talk of “being in the long period,” or “reaching the long period,” as though it were a date in history. (Marshall himself thought of the economy as tending toward long-run equilibrium but never actually being there.) It is better to use the expressions “short period” and “long period” as adjectives, not as substantives. The “short period” is not a length of time but a state of affairs. Every event that occurs, occurs in a short-period situation; it has short-period and long-period consequences. The short-period consequences consist of reactions on output, employment and, perhaps, prices; the long-period consequences concern changes in productive capacity. (Pp18)
This was written in 1971 in a period of her life where Robinson was coming to reject the vast amount of economic theory that she had done in the past. Disappointed by the results and effects of the Capital Debates, she began to realise that she had merely been fighting symptoms and the sickness in economic theory ran much deeper. You get a sense of this especially in the first of these lectures she gave in Stanford in 1974.
I think in this passage you see Robinson breaking away from what Veblen would have called “neoclassical” analysis. This especially so when she says that it is better that we understand “the expressions “short period” and “long period” as adjectives, not as substantives”. Here she is highlighting the “short” and “long” aspects rather than the “periods” per se. These concepts should not be thought of as short or long in then sense that we can assign them a numerical unit of time — say, one year and ten years respectively — but merely as being short and long relative only to each other. Just as we might say a “short man” who is only short in relation to his peers (among the Inuit, for example, he might be rather tall). As we can see Robinson develops such arguments out of the Marshallian tradition and suggests that such analyses would be impossible in Walras.
So, back to Lawson’s taxonomy. Where does all this fit in? I would suggest that the Walrasian system fits in to Lawson’s first group. Most economics practiced and taught to today is, in this sense, Walrasian. Marshall broadly fits into Lawson’s third group which we might call neoclassical. I would suggest that much actual policy analysis undertaken today fits into this category; wherein the policy analyst keeps at the back of their mind an ultimately fixed or static model and then analyses policy based on this. All mathematical deductive analysis that is not strictly general equilibrium/Walrasian, whether modelling or econometrics, also fits into this category. Finally, Robinson’s own mature views fit nicely into Lawson’s second category. Robinson is engaged in exorcising any notion of non-historical time out of the concepts she is using; something only hinted at in Marshall.
The moment I finished reading Lawson’s paper I knew that the distinctions he had made were not simply of rhetorical value. In fact, they provide a taxonomy with which we can sort out various paradigms based on their underlying metaphysics (Lawson would say: ontology). This is enormously advantageous. For example, we know that both Marshall and Walras are marginalists, but only the former is neoclassical in the proper sense of the term. There is a lot of fruitful research that can be conducted along these lines. I, for one, will certainly be using the term “neoclassical” differently — or should I say: correctly? — from now on.