Lionel Robbins on the Principles of Economic Analysis

The 1930s Lectures

Lionel Robbins, Susan Howson

• Publisher: Routledge
• Year: 2018
• eBook ISBN: 9781317225812
• Edition: 1
• Topic: Business
• Subtopic: Business General
• Index: Business

Lionel Robbins (1898–1984) is best known to economists for his Essay on the Nature and Significance of Economic Science (1932 and 1935). To the wider public he is well known for the 'Robbins Report' of the 1960s on Higher Education, which recommended a major expansion of university education in Britain. However, throughout his academic career – at Oxford and the London School of Economics in the 1920s, and as Professor of Economics at the School from 1929 to 1961 – he was renowned as an exceptionally gifted teacher. Generations of students remember his lectures for their clarity and comprehensiveness and for his infectious enthusiasm for his subject.

Besides his famous graduate seminar his most important and influential courses at LSE were the Principles of Economic Analysis, which he gave in the 1930s and again in the late 1940s and 1950s, as well as the History of Economic Thought, from 1953 until long after his official retirement. This book publishes for the first time the manuscript notes Robbins used for his lectures on the Principles of Economic Analysis from 1929/30 to 1934/40. At the outset of his career he took the advice of a senior colleague to prepare his lectures by writing them out fully before he presented them; the full notes for most of his pre-war lectures survive and are eminently decipherable.

Since he made two major revisions of the lectures in the 1930s the Principles notes show both the development of his own thought and the way he incorporated the major theoretical innovations made by younger economists at LSE, such as John Hicks and Nicholas Kaldor, or elsewhere, notably Joan Robinson. He intended to turn his lecture notes into a book, abandoning the project only when he was asked to chair the Committee on Higher Education in 1960. This volume is not exactly the book he wanted to write, but it is a unique record of what was taught to senior undergraduate and graduate economists in those 'years of high theory'. It will be of interest to all economists interested in the development of economics in the twentieth century.

Part I 1929–31

Introduction

Editorial Note: The lecture notes in this section were written to be given first in the Lent and Summer Terms 1930 and the Michaelmas, Lent and Summer Terms of 1930/31. With the exception of the first part of Lecture 8 (Production: Factors Flexible: Labour Supply) the first 15 lectures come from the notebook labelled ‘Principles of Economics I Old Notes 1929–31’ and Lectures 16–20 come from the notebook labelled ‘Principles of Economics II’. Lectures 8, 21 and 22, however, were filed in the notebook labelled ’1939–40’, as was lecture 15*. The lecture notes as a whole correspond to the description of the course in the LSE Calendars for 1930/31 and the next two years.

1 The framework of economic analysis

Today I want to discuss with you what you may call the analytical framework of Economics – questions of arrangement and division.

To some of you this may sound as if I am at once going back on my promise not to indulge in investigations of method. But this is not the case. I am not proposing to discuss logical method at all. What I want to do rather is simply to define a little the nature of our problem and sort out various ways in which we can approach it – to choose between tools – not to discuss the rationale of tool using.

For clearly there is a problem. We cannot do everything at once. Our business as I conceive it is to examine human behaviour as conditioned by the fact of scarcity – you may disagree with this if you like – and our subject matter is therefore the whole complex of human relations arising from this fact: Production, Exchange, Distribution, Localization, Trade and so on and so forth. Clearly if we are to do anything which takes us beyond the naive conclusions of experience and common sense we must attack the problem by stages. We must analyse. We must break up. And clearly if we are to do this, the way we are to do it is not a matter of indifference. There are some modes of approach which are better than others – better that is to say in that they lead to more significant results – in that they facilitate our understanding of the whole.

Now the traditional way of dividing the subject has been – or has pretended to be – division into sections on production and distribution.

The … fundamental questions of economics are why all of us taken together are as well off … as we are and why some of us are much better off and others much worse off than the average, says Cannan

(Wealth Int.)¹

And this procedure – the procedure of dividing our enquiries so as to provide answers to these questions has been followed by perhaps a majority of English and Continental Economists. There are differences of content under these two headings and differences as regards supplementing divisions. But this division has been the main ‘cut’ – as it were – into the body of the subject.

And no doubt there are advantages to be gained from the use of this method.

I do not question Cannan’s assertion that the two questions to which he makes the division correspond are the fundamental questions which we ask if we turn to Economics for guidance on matters of practical policy. What will be the effect on production What on Distribution Do the two effects work together or tend to cancel out – These are certainly the questions we ask when are considering the advisability of a tax or a bounty, or this form of control or that relaxation. We are out to better production and to better distribution. It is only natural that we should frame our analysis to assist these endeavours.

But in scientific investigation it does not always “pay” even in the most material sense of the word to have too exclusive an eye to the practical. Of course there is no guarantee that if we seek first accuracy and truth, if we seek light, fruit shall be added unto us. But it not infrequently happens that it is. We know that in the natural sciences enquiries which have seemed most recondite and abstract have ultimately proved to have the most important practical application. So that even if we regard the raison d’etre of economic study as amelioration of economic conditions rather than mere knowledge and understanding – and I am far from suggesting that this should be our attitude – even if you have practical ends in view, the fact that your analysis is framed with them in mind is not an infallible sign of its excellence. And as a matter of fact when we come to examine the production, distribution analysis not from the point of view of its suitability for providing broad answers to general questions but from the point of view of its suitability for affording a body of exact “laws” for generalizations we begin to be aware of certain deficiencies.

It is worth while to look at these closely.

Economic “laws” or generalizations must if they are to be exact relate to movements of quantities. They are concerned essentially with questions of more or less. They may not always be statistically verifiable. There may be profound difficulties in the way of experimental proof. But they should at any rate relate to quantities which are capable of more or less accurate definition. They need not have a numerical content. They should have a general quantitative form which is intelligible. This being so we should expect a law of production to conform to this general requirement. And no doubt there are laws of production – the law of diminishing returns in particular industries, e.g. which do succeed in doing so. But when we come answer the broad question what are the causes of increases or diminution of production quantitative exactitude is harder to conceive. It is easy enough to conceive broad qualitative generalizations such as that if men work harder or save more their power to produce will be greater. But it is when we come to give our generalizations more exact form that complications begin.

Let us examine the nature of these complications.

We want to discuss changes in the aggregate or average volume of production. How are we to conceive exactly of such changes[?]

It is easy enough to conceive of changes in the volume of any one kind of product. You simply measure in physical units.

But this plan is not available when you come to generalize about the movement of collections of commodities.

You might go some way if all the physical units moved in the same direction 100 A + 100 B + 100 C … becoming 200 A + 200 B + 200 C

Though here great precision would be difficult. For how could you say whether one increase was more than another?

• e.g. 150 A + 200 B + 250 C
• than 250 A + 200 B + 150 C

But of course it is notorious that the production of heterogeneous aggregates does not move in this simple way. It is a question of judging whether

• 100 A + 100 B is greater than, equal to or less than 50 A + 200 B

Some quantities move up, others move down. Obviously physical computations are out of the question.

We therefore fall back upon comparing aggregates of values. We reduce our products to a money denominator and add. Now I do not want to suggest that computations of this sort are useless. Clearly they do afford a more precise guide than more guesses. Clearly if they are used simply as evidence of direction of change they may be very helpful. But from the point of view of strict accuracy, they are open to grave strictures.

1. (1) In the first place come the well-known technical difficulties of eliminating variations in the measured. The value of money changes and if corrections are not introduced the measurement is vitiated. (See Keynes & Haberler)² (If capital is being measured it is also necessary to take account of changes in the rate of interest.)
2. (2) Secondly if the elasticity of demand for a commodity is less than unity, the aggregate value moves inversely to the volume of supply.
3. (3) These are the elementary cautions of first year economics.

And it could be argued perhaps that they are not very important, that changes in the value of money can be eliminated and that high inelasticity for important products is the exception rather than the rule. I am not clear that these would be valid replies. Still we may concede that they are respectable.

But there are yet more serious theoretical objections. In thinking of the validity of these conceptions of aggregates we have to remember the fundamental relativity of the value concept. The whole system of valuations on which we base our conception of an aggregate depends essentially upon a given distribution of wealth. Alter the distribution of wealth and the whole system of values changes also. You can see that quite clearly when you are thinking of calculations regarding the amount of the national income available for redistribution. You are all acquainted with Professor Bowley’s discussion of this matter in his Division of the Product of Industry.³ No doubt, calculations of this sort do provide some sort of basis for discussion. But we shall be tempted to go too far on this basis if we neglect Professor Bowley’s own caution that if great redistribution actually took place the basis of measurement would be gone.

This may sound abstract but it is really quite simple. it is common knowledge that the high incomes of some of the professional classes depend on the presence of other people in the community with high incomes who are willing to bid high for their services. If income were redistributed on an equalitarian basis Charlie Chaplin might go on collecting his £100,000 p.a. But it is doubtful whether famous surgeons would get huge fees for operations. It is certain that firms of solicitors could not make high profits by advising wealthy people on how to obtain quick and inconspicuous divorces.

1. (4) Finally, even if you ignore this you must remember that your computations of changes depend upon the assumption of fixed tastes and habits on the part of your consumers. It is really just an accident of history so to speak that certain things are demanded in certain quantities. Let there be a change in tastes an alteration of fashion and you are no longer comparing social dividends of the same composition. No doubt there is a sense in which you can say that the social dividend now is greater than the social dividend of the time of Queen Anne. But it is not sufficiently precise to provide a basis for extensive quantitative generalization.

For all these reasons, then, discussion of production from this point of view does not lend itself to exact analytical treatment. And there is yet a further reason why this general mode of approach does not always commend itself. It involves duplication of treatment and overlapping. For it is clear – is it not – that under capitalism at any rate there is an intimate relationship between the institutions of production & the institutions of distribution. We divide the product by means of money incomes. And the money incomes are the incentive to further production. It is a commonplace of elementary exposition that the processes of production and distribution are not really discreet [sic] – that the division made in the class room is only an expositional division.

It follows therefore that in dealing with the economic system from these two angles, while there are certain fields which are not common to the field of vision, there is a large field – the price system which is. If you adhere rigidly to the division you are bound to treat certain things twice over.

Because of all this – because of the difficulties of quantitative precision and because of the inelegance of a division which involves so great a degree of overlap – in recent years it has become the custom to approach the problem from rather a different direction. Instead of regarding the economic system as a machine for producing and distributing and inquiring concerning its potentialities and tendencies in these respects, it is considered simply as a system of quantities of goods of different descriptions – quantities of productive services – and analysis is directed to discovering under what conditions this system exhibits equilibrium – under what conditions there is no tendency for the quantities to alter or under what conditions they exhibit characteristics of orderly or periodic growth.

I say this is a recent habit – and it is true that it is only recently that this has come to be the avowed and central preoccupation of important treatises. See e.g. Pareto Cassel Knight etc.⁴ But of course it is a treatment which has been implicit in many earlier systems. The idea of different tendencies to equilibrium. The idea of natural or normal prices etc., has dominated whole areas of economic theory since the days of Adam Smith and the Physiocrats. The whole theory of price has been developed essentially in terms of tendencies to equilibrium. The theory of distribution has been to a large extent on the same basis. The moderns simply make explicit the methodological assumptions of these earlier theories and generalize them.

This method has very important advantages over the other.

1. (1) First, it considers always quantities which – in conception at any rate – are precise. It contemplates the flow of quantities of commodities of like kind and qu...

Table of contents

• Cover
• Title
• Copyright
• Dedication
• Contents
• Editor’s introduction
• Part I 1929–31 Introduction
• General outline of equilbrium analysis
• Special topics
• PART II 1932/33–1934/35 Introduction
• General outline of equilibrium analysis
• Part III 1935/36–1939/40 Introduction
• Statics
• Comparative statics
• Dynamics
• References
• Index