Lionel Robins Methodology Lectures

Lionel Robins Methodology, Lionel Robbins Lectures

An Essay on the Nature and Significance of Economic Science (Robbins, 1932) is one of the great works of twentieth-century economics. This is clearly so if it is viewed solely as a statement of a problem area or class of problems, but the book also has greatness because of its attempt at solutions. The latter may now be regarded as a magnificent failure, but from the vantage point of the 1980s it looks less of a disaster than does the naive empiricism of the previous three decades.

What is the nature of the Robbins contribution? In the first place he asks the direct question, on what foundation do the basic propositions of economics (and especially value theory) lie? Secondly, he sees that these propositions do have and are meant to have empirical relevance. Thirdly, he rejects historical inductionism, or support via the notion of 'history shows'. Fourthly he recognises the limitation of controlled experimen­tation in economics. Fifthly, he argues that the correctness of microeconomics does not depend on the validity of particular psychological doctrines. (In this connection it is interesting to note his rejection of behaviourism partly on the grounds that the subjective theory of value cannot be properly understood without a psychical element.) Sixthly, he adopts the position that econ­omics is value free in the sense that it does not depend on individual valuations being regarded as ethically appropriate. Seventhly, in various parts of the book he emphasises the importance of abstraction, of the, particular assumptions made. (A good example of this is the use of the assumption of perfect foresight, although for other purposes, such as the theory of profit, uncertainty must be the dominant theme.)

Given this position he then argues that the propositions of economics are deductions from various postulates, 'and the chief of these postulates are all assumptions involving in some way simple and indisputable facts of experience relating to the way in which the scarcity of goods which is the subject-matter of our science actually shows itself in the world of reality'. In addition to these ultimate truths there are 'the assumptions of a more limited nature based upon the general features of particular situations or types of situations which the theory is used to explain'.2 It must be recognised that Robbins mentions both sorts of assumption. Equally, there can be no doubt that he is correct in stating that economics consists of theoretical assumptions and generalisations, particular propositions, and a wide range of inferences. The difficulty is the one he faces, namely whether all these are true, and what it is of an empirical nature that economists have established. In this connection it seems to the present writer that the usual objection to Robbins, his assertion of simple and indisputable facts, represents a mistaken emphasis. Many if not all of his facts may be indisputable but this does not mean that a great deal of significance can be derived from them.

His view on methodology did develop. An extremely interest­ing statement appeared in 1959. This was noteworthy partly for what Robbins did say and partly for what he did not. On the latter there is no mention of simple facts, intuitively known for certain. On the former there is a definite acceptance of the hypothetico-deductive method as enunciated by Popper. Robbins says succinctly, 'We devise theories to explain the world and we test them by asking whether they perform this function in specific instances ... as economic theory becomes more advanced and complicated, the need for testing becomes more and more imperative'.

Later on Professor Robbins went further still and argued that his methodological position had been too essentialist. He claims that he tried to get the problem of testing by reference to reality sorted out in the second edition, but agrees that he was really not successful. His justification, and it is a fair one, was his 'reaction . . . against the ridiculous claims of the institutionalists and the cruder econometricians and an attempt to persuade Beveridge and his like that their simplistic belief "in letting facts speak for themselves" was all wrong'.5 His move to the Popperian position and his willingness to accept sophisticated econometrics, while apparently reasonable, may have been a trifle premature. For while the logic of the hypothetico-deductive method may seem compelling, its practical usefulness is less so. Lionel Robbins's initial doubts about the ability of economists to establish empirical laws have surely been borne out by the experience of econometric estimation and testing in the past decade. Almost every empirical generalisation, above the level of the trivial, has a tendency to collapse, and almost all that remains is economics as an analytical method which can be applied tentatively and with much judgement to an extremely uncertain world.

Let us, however, revert to the question of whether there are 'simple and indisputable facts of (economic) experience'. What Lionel Robbins had in mind, for example, was scarcity, by which I do not believe he meant so much that economics was the science of scarcity, but rather that the rejection of the existence of scarcity would be to deny the most obvious facet of experience. He is not saying that he could not conceive of a world without scarcity, but that it was not this one. Moreover, he was not denying the existence of unemployment; nor is the existence of scarcity incompatible with unemployed labour with a low, even zero, opportunity cost.

He is arguing that all decision makers in society are constrained, and for most the constraints are effective in that their wants exceed their resources and they are not satiated. Note that he is aware of the existence of ascetics and altruists, but still insists that scarcity applies to them. It is simply that their tastes are different.

More generally, even if wants are not limitless, they need merely exceed the free bounty of nature for scarcity to exist. Even in those circumstances there remains the not totally uninteresting question of the allocation of the hours of the day and the days of a lifetime.

It seems to me, therefore, that Robbins is not mistaken in insist­ing on the indisputable fact of scarcity, both in saying that it is indisputable and in insisting on its empirical character. I think also that one can cite other facts of similar standing, the most noteworthy of which is the existence of uncertainty. I do not mean by this the Popper proposition concerning the self-contradictory nature of a statement purporting to predict the whole of the future, but the much simpler point that even the most elementary characteristics of the next day cannot be known for certain.

What must be discussed is whether these facts are as important as Robbins thinks. It may well be that he is in error not so much in suggesting that they exist, but in exaggerating what can be derived from them alone. There are two sides to this. One is that we need a great deal more than these facts to get anywhere in economics at all; the other is that we can make progress in economics without paying explicit attention to them. (I ignore the well-known criticism of Robbins that some indisputable intuitions turn out to be highly disputable. The most obvious of these is that some economists are perfectly convinced that they have well defined wants expressible in a utility function while others are equally convinced of the opposite. It is also worth adding that the doubtful nature of some 'Robbins facts' is also characteristic of 'Friedman facts', i.e. those observations that determine the correctness of a theory as shown by its predictions. Here too what is more usual is the contentious character of testing rather than the agreement of 'objective observers'.)

Even the most simple examples clarify the matter. Robbins did not and would not doubt the validity of the theory of the firm; but that theory in the form in which it appeared and was discussed in the thirties tended to ignore uncertainty. And in this respect (if not in any other) it has made scant progress since then. Equally, the existence of scarcity does not tell us that 'the price of pork fluctuates with variations in supply and demand'. It does not even tell us that there will be a price of pork. Of course, it is possible to say that ex definitione if pork is scarce, a unit increase in the quantity available must imply a decrease in the availability of something else, and that decrease is the price of pork. But that is a purely theoretical consideration adding nothing to the notion of scarcity. If Robbins is saying anything at all, it must be about the price of pork as commonly understood or something similar; and this need not be determined by supply and demand. Despite its scarcity pork may be distributed to people by dividing the total quantity produced by the number of people, everybody receiving an equal share which they regard as fair and satisfactory. It is possible, therefore, for Robbins's sentence to be false, and 'the price of pork is fixed by government decree' to be the true one.

For the price of pork to depend on demand and supply a great deal more must be the case than the existence of scarcity, and much of that extra, experiential though it may be, is neither simple nor indisputable. We require the existence of a market (which is not necessarily a market place, but something more abstract) involving suppliers and demanders. Something must be said about how the price is fixed and how it comes to change and whether it is absolute price or relative price. It is difficult to avoid questions about the fulfilment of supply plans and demand plans, and one is led inexorably to that elusive concept, equilibrium. In other words, even to begin to assess what is apparently the most elementary 'law' of economics, one becomes involved with theory going beyond the fact with which one started.

And, of course, one is already in deep methodological water, for 'market', 'supply', and 'demand' are the names of concepts in economics and 'price' may be. Certainly, they are not Robbins's facts, but were, of course, invented by philosophers or political economists. They may have empirical counterparts, but these are to be sought and agreed upon, and may remain doubtful. Indeed, were this not so there arises the ultimate paradox of doubting the value of economic theory which Robbins initially set out to defend. If the theory cannot be wrong what significance lies in it being right? If supply, demand and all that were not additions to thought why is all the fuss made about discovering them?

The answer is that at the very least the theory draws our attention to and helps us to interpret experience. It guides us to look at the supplies of pork coming forward at the existing price and to the quantities that people are endeavouring to buy. It claims:

(i) that if the former is reduced compared with an earlier
occasion or the latter increased, the price will rise, (ii) that if the price is rising, the former must have been reduced or the latter increased.
Experience (not particularly simple) may confirm all this to the point of it being indisputable, but it is theory that is being, and needs to be, tested. The rise in price may not be open to doubt and the reduction in supply may certainly have happened, but that they go together is a contingency, and one which substantiates the theory. For their existence to be a necessity is of no help to the theory, but rather renders it nugatory.

Let us, however, now proceed and accept for the sake of argument that demand and supply are forces that indisputably act on price, doing this precisely in the way that Robbins claims. In particular let us accept that proposition (i) has been substantiated on many occasions and is undoubted. What does this remind us of in natural science? Surely it is such low level propositions as 'the sun always rises' or 'the seasons always follow each other'. They are undisputed facts of experience and are not incompatible with physical theory. But they play little, if any, part in such theory and go unnoticed. Thus, even if no other difficulties arose with Robbins's view, does it not cause us to set our sights extremely low, limiting us to little more than commonsense? Robbins's facts in themselves do not get us very far and devotion to them may prevent our progressing further.

There is yet one final card to be played. If one enquires as to which part of economics is most closely geared to the simple and indisputable, notably to supply, demand, price, and scarcity, the answer is surely general equilibrium theory. Yet this is the most abstract area of economic discourse, the one the conclusions of which are most distant from ordinary experience and which, while seemingly irrefutable, and in Robbins's terms entirely soundly based, is more or less irrelevant to the real world.