The First Cambridge Years

Robin’s earliest work was on the historical and theoretical aspects of the trade cycle in the UK and advanced capitalism generally. His first book, A Study in Trade Cycle History (Matthews 1954a), was published by Cambridge

University Press in 1954. The Preface is dated November 1952 so perhaps the manuscript of the book played a role in his promotion to a University Lectureship in 1951. He tells us in the Preface that his chosen method of enquiry—‘quantitative-historical’—was only one of many by which empirical research on the trade cycle may be carried out. He subjected a

single brief period [running from the late 1820s to the early 1840s] to close study so as to be able to do as full justice as the evidence permits to the complexity of the fluctuations experienced while avoiding the dangers of oversimplification through imposing too uniform a pattern on the history of fluctuations in different periods (ibid.: xiii).

His procedure did not allow an assessment of the place of fluctuations in the longer-run evolution of the national economy nor to discover changes in the cyclical process itself over time.

One of his major findings (ibid.: 224) was that the imperfect synchronisation of fluctuations as between industries and between home and export markets, in conjunction with the high proportion of the population dependent upon an industry, where outputs and prices were governed more by weather than by the level of demand, meant that changes in the distribution of income between social classes and between industry groups were a feature of short-period fluctuations scarcely less important than changes in the aggregate of incomes.

As an astute scholar of Keynes, Robin had a keen sense of the importance of finance in economic processes. He noted in his study that ‘even entirely agricultural regions were affected by the boom in joint stock bank formation ... [L]ocal traders were often induced to expand their operations by availability of finance during the boom’ (ibid.). These and other passages document that Robin was already bringing to his analysis a rich mixture and understanding of institutions, evolutionary changes, incentives, and the need to understand in detail what was happening in different sectors, in order to see how they interrelated and also to obtain an understanding of the impact of their linked behaviour on systemic processes.

Robin’s next book was his Cambridge Economic Handbook, The Trade Cycle (Matthews 1959). It is still one of the best introductions to the theories of this inherent characteristic of capitalism. As with all the best research, it was closely connected to his teaching. Dalyell writes: As a young lecturer specialising in trade cycle theory, Robin was a beacon of clarity, [explaining] in understandable language the intricacies of the trade cycle theories of such heavyweights of the 1950s as Sir Roy Harrod, Nicholas Kaldor, Michai Kalecki and Franco Modigliani’ (Dalyell 2010).

In the Preface, Robin writes that he was much indebted to Richard Goodwin, Kaldor, Joan Robinson, and James Duesenberry when Duesenberry was on leave in Cambridge. He thanks Aubrey Silberston, Joan Robinson, Kenneth Berrill, and the General Editors of the series, Claude Guillebaud and Milton Friedman (!), for reading the manuscript. Would we now accept the claim by the General Editors (in Matthews 1959: vi) that ‘there is again a large measure of agreement among economists on the fundamental theoretical aspects of their subject’, adding the proviso that there is also ‘a wide divergence of views on policy’? Would we also accept what Maynard Keynes wrote in his Introduction to the series soon after the end of the First World War: ‘[T]he theory of economics does not furnish a body of settled conclusions immediately applicable to policy. It is a method rather than a doctrine, an apparatus of the mind, a technique of thinking which helps its possessor to draw correct conclusions’ (ibid.: vii)?

The titles of the 14 chapters in the table of contents are witness to the extraordinarily comprehensive coverage of the subject. The author saw his task as ‘mainly theoretical analysis’. The aim was to ‘provide a framework within which any particular phase of historical experience of fluctuations...can be usefully studied’. The subject matter was ‘advanced or relatively advanced countries’ (ibid.: 2). Robin thought then ‘that much of the theory of the cycle [could] be stated more or less independently of the exact nature of the underlying growth process’ (ibid.) so that a more systematic discussion of the relation between growth and fluctuations was left to near the end of the book. This separation is not a view that Dick Goodwin would have accepted even then and which, like Kalecki, he discussed comprehensively in his later writings (see, for example, Goodwin 1967; Kalecki 1968).

Robin (Matthews ibid.: 3) noted the general agreement that changes in the level of demand cause fluctuations in national income, that the rise and fall of demand is the essence. A typically wise insight, so characteristic of all his writings, is that output and prices move together whereas if fluctuations arose chiefly from supply, they would move in opposite directions (ibid.: 4). He singles out fluctuations in investment, the most volatile component of aggregate expenditure, as the principal focus, so that the main topic in trade cycle analysis is an explanation of investment fluctuations, themselves associated with the mutual interaction between consumption and investment.

The most difficult formal part of the book is in Chapter 2, ‘the aim of which is to provide a rigorous background to more realistic discussion later on where the chapters are less abstract and mathematical’ (ibid.: 7, fn. 1). Robin started with analysis of the accelerator principle as a means to his end of setting out the capital stock adjustment principle as a looser formulation of a more satisfying and illuminating general model of the cycle.[1] By having investment decisions varying directly with the level of income and indirectly with the stock of capital inexistence, most of the difficulties with the strict form of the accelerator principle are overcome. It is stressed that ‘inertia in expectations.. .serves to protect the system from the more extreme versions of cyclical instability’ (ibid.: 48).

The next chapters cover crucial phases of the cycle—upper and lower turning points, inventory cycles and so on—as well as the role of consumer behaviour. In Chapter 13, the author returns to the trend and the cycle. He starts by reminding us that some economists regard them as intimately related phenomena and that it is impossible to treat them in isolation from one another. Others, though, hold that they are independent; the cycle represents a movement of demand relative to supply while the trend is concerned with growth in demand and supply pari passu (ibid.: 227).

For Robin a large part of the problem is to evolve an analysis of the demand side of the growth process that is consistent with the analysis of fluctuations around it: ‘We cannot assume that effective demand, which gives so much trouble in the short period, simply looks after itself in the long period’ (ibid.: 229)—that wise insight has fallen on deaf ears in modern times. Technical progress and population growth are singled out, Robin tells us, for imparting a long-run upward trend to the consumption function and/or the inducement to invest (ibid.: 253).

With prescience, Robin argues (ibid.: 148) that monetary and financial factors may strengthen, not weaken, tendencies to real fluctuations, that they are sometimes responsible for magnifying the effects of relatively minor real disturbances into major movements in income and activity— Matthews’s Minsky moment? While Robin thinks it unrealistic to expect to be able to achieve complete stability in investment expenditure, he does not think that a policy of maintaining stable full employment will have an adverse effect on growth and productivity. Such a policy creates a more stable environment within which entrepreneurs are able to plan ahead more effectively (ibid.: 256).

He also argues that with the then existing wage- and price-fixing institutions, only really high levels of unemployment would ensure a stable price level, a policy which has little to commend it. He prefers policies which try to persuade those responsible for setting wages and prices to show moderation. Robin was sceptical of the effectiveness of an application of monetary policy in the slump (ibid.: 262). On page 263, ‘faith by magic’, an early anticipation of rational expectations, is discussed—if businessmen believe the government is committed to full employment, the government may not need to do anything as businessmen moderate changes in short-run expectations and become confident about long-term prospects. Robin was implicitly criticising much of the conventional wisdom of the time, not least that associated with Dennis Robertson’s criticisms of Keynesianism and its policy stances.

The Trade Cycle was published four years after one of Robin’s most influential articles, ‘The Saving Function and the Problem of Trend and Cycle’, was published in R.E. Studs in 1954-1955. Already he was considering the nature of the interaction between trend and cycle, starting from Roy Harrod’s 1939 classic article and taking in Goodwin’s emerging, indeed emerged, writings on their entwining (see Harcourt 2015), Duesenberry’s 1949 classic on the consumption function, Modigliani’s theory of lifetime saving behaviour (Modigliani 1949; Modigliani and Brumberg 1954; Modigliani and Ando 1957), and John Hicks’s ‘contribution’ to the theory of the cycle (1950).

Goodwin had concentrated on the investment function as the means by which trend and cycle could be merged in an entwined relationship. Robin reached similar conclusions to Goodwin but concentrated instead on the saving function. The starting point, he argued, is the neglect in Duesenberry’s book of ongoing productivity growth when he dated the onsets of the ratchet effect as the previous highest levels of income attained, rather than the previous lowest levels of unemployment. This is the sort of insight that the best economists provide and then spell out the implications, in Robin’s case, for the relationship between trend and cycle. The latter has become the base on which our understanding of the processes at work in capitalism is built.

Subject to the limitations of ‘a relatively high level of abstraction’ (Matthews 1959: 95), Robin concludes:

The hypothesis that the proportion of income saved depends on the amount of unemployment leads to the inclusion in the saving function of a negative term that grows pari passu with population and productivity [causing] growth to take place in the type of system that would in its absence fluctuate about a stationary level. The growth that takes place will be at the natural rate ... The shape of the saving function may, therefore, be regarded. as the explanation of why the increase in saving does not prevent income from growing at the natural rate. The average level of activity over the cycle.. .will lead to an average ratio of saving to income equal to the average ratio of investment to income required by the natural rate ... The natural rate will.. .depend in part on the average level of activity, so that anything that.raises [the latter] will also raise the growth rate.

Robin thus provided a more convincing explanation of an endogenous link between the natural rate and the investment to income ratio than Kaldor’s imposition of the required ratio onto his macroeconomic ‘Keynesian’ theory of distribution and its assumption of full employment being necessary to achieve long-period growth (see Kaldor 1955-1956; Harcourt 1963).

Here, Robin also departs from the usual interpretation then of Harrod’s theory that gw ga, ge, and gn do not interrelate whereas in fact technical progress is embodied in the capital stock by new accumulation, as Salter (1960, 1965) showed and the late John Cornwall devoted his life’s work to developing (see Harcourt and Monadjemi 1999).

In 1961 Robin published in Economica an article entitled ‘Liquidity Preference and the Multiplier’. Using Keynes’s liquidity preference theory of the rate of interest, Robin wrote one of the earliest and best accounts of what is happening in the banking and financial sectors of the economy when the Kahn-Meade-Keynes multiplier is working itself out in the real sector. He also showed how self-finance and ‘other imperfections in the capital market’ (ibid.: 37) could be incorporated into the analysis. All the assumptions are set out clearly and concisely and the implications and limitations of the analysis are similarly treated. The article illustrates Robin’s ability to integrate into his approach appropriate analysis from often seemingly opposing stances. Thus he incorporates ideas from both Robertson’s loanable funds theory of the rate of interest and his lagged form of the consumption function with its accompanying use of process analysis. Keynes’s theory is set out in terms of a utility function analysis incorporating indifference curves between bonds and money (but Robin points out that the theory takes in choices at the margin between all possible financial assets, expressing Keynes’s essential insight that the levels of rates of interest are the means by which an uneasy truce is brought about between bulls and bears in financial markets). Robertson’s theory is argued to be partial because it is solely concerned with flows. As it ignores stocks, it overlooks that what is being analysed are markets where stocks dominate flows and expectations about future yields dominate what Robertson regarded as the fundamental determinants—productivity and thrift.

Robin’s formal analysis of the approach to a new short-period equilibrium following an autonomous change in either the saving function or the investment function allows him to discuss what happens within each period on the way where temporary hoards or dishoards of money may occur and neither desired nor planned saving and investment, or their equality, are achieved until the new equilibrium has been reached. He combines the simple multiplier analysis with other factors affecting holdings of cash and examines their impact on both the size of the multiplier and that of the multiplicand. The discussion at the close of the article of imperfections and retained profits argues that the latter may bring an element of Say’s Law reasoning back into an essentially Keynesian analysis as the retention of profits serves to match a planned increase in investment expenditure.

Altogether, the article is an example of highly original and relevant thinking set within major approaches by others to the issues concerned at the time. The tone of the article reveals an author who is on top of his material but who is also matter of fact, modest, and unassuming, so typical of Robin then and later.

In 1964, with Frank Hahn, Robin published in the December 1964 issue of the Economic Journal the famous survey of the theory of economic growth, the survey article that has been the role model for survey articles ever since. Though all three sections of the survey achieved high levels of exposition and scholarship, the second section on technical progress, which was primarily due to Robin, is its jewel in the crown. Its lucidity and clarity bring out the deep economic intuition it contains.

The first section, ‘Growth without Technical Progress’, is a joint product. It starts with a methodological discussion concerning steady states, existence, different types of stability, and different classifications of dynamic analysis. It takes the Harrod-Domar model as its starting point and reference point from then on. Coupling Harrod and Domar together is not a correct reading of their respective approaches and contributions, even though their writings contain the same expressions for their growth rate concepts. Harrod was concerned with growth rates at a point in time and their relationships, the dynamic counterparts of levels at a point in time in static analysis. He wanted to know the conditions that allowed overall investment plans to be realised and so maintained because expectations had been fulfilled—hence his concept of the warranted rate of growth, gw which he distinguished from the natural rate of growth, gn, a supply concept which set out the maximum potential growth from workforce growth and technical progress. He examined the stability, or otherwise, of gw and how gw and gn may be equalised. Domar was concerned with the maintenance of full employment of labour and the capital stock over time when both the employment-creating and capacity-creating effects of investment are taken into account. Hahn and Matthews ran these different objectives together, which defined the classification of their discussions from then on.

In Section 1, the authors discuss how Harrod’s two puzzles, but especially the stability of gw, have been tackled by various authors. Having established from the Keynesian saving-investment relationship that g = s/v, they examine the contributions of various authors—Kaldor, Joan Robinson, Solow, and Swan—who tackled initial inequality by looking at market signals which affect the value of v and those which affect the value of s through changes in the distribution of income when sn > sw. They examine first one-sector, one-commodity models, and then two-sector models. They provide helpful diagrams, and sometimes simple algebra, to illustrate their excellent economic intuitions.

All this sets the stage for Section 2 where the effects of technical progress on the attainment or not of steady-state growth are exhaustively analysed. An organising principle is whether technical progress is Harrod-neutral or Hicks-neutral. The discussion sets out both the formal requirements for various results and the likelihood or not of actual economies providing the conditions for them to be achieved. The discussion includes the introduction of vintage models and the stance that rejects the use of a production function whereby the effects of deepening are distinguished from those of embodiment of technical advances through accumulation. Much of the discussion takes in the writings of Kaldor on these themes, the development of his growth models from 1957 to their culmination in Kaldor-Mirrlees in 1962.

The survey was written while both Ken Arrow and Bob Solow were on leave in Cambridge for a year. The first draft was discussed in early 1964 by Solow at the secret seminar—the seminar run by Joan Robinson, Kahn, and Kaldor in order to discuss on-the-frontier topics (but which excluded those members of the Faculty who were associated with Robertson). Arrow was present on this momentous occasion.[2]

  • [1] Vela Velupillai tells me (7 May 2015) that Dick Goodwin probably did not approve of this ‘rechristening’ of the flexible accelerator.
  • [2] One of the greatest intellectual regrets of my life was that I was not present because I caught mumps theweek before from a member of the Department of Applied Economics (DAE). He came to work whenhe was most infectious and I foolishly took it upon myself to tell him to go home as some of us had smallchildren.
< Prev   CONTENTS   Source   Next >