Nicholas Kaldor and Mainstream Economics

Professor Lord Kaldor, who died at Papworth hospital near Cambridge on 30 September 1986, aged 78, was one of the most distinguished economists of the 20th century who will be recorded in the history of economic thought as a brilliant theoretician and applied economist, surpassed in originality only by Keynes and Harrod among British economists this century. He was a dominant influence in economic debates on the world stage for over fifty years, and hardly a branch of economics escaped his pen. At the London School of Economics (LSE) in the 1930s, while still in his twenties, he emerged as one of the country’s leading economic theoreticians making fundamental contributions to controversies in the theory of the firm and in capital theory; to trade cycle theory and welfare economics, and to Keynesian economics by ‘generalizing’ Keynes’s General Theory, which nearly 50 years later led Sir John Hicks to remark, ‘I think that your (1939) paper was the culmination of the Keynesian revolution in theory. You ought to have had more honour for it.’1 His reputation was such that in 1938, and still only thirty, he was offered a Chair by the prestigious University of Lausanne — the home of Walras and Pareto — which he reluctantly declined. Keynes thought extremely highly of him. In a letter to Jesus College, Cambridge in 1943 suggesting Kaldor as an Economics Fellow, Keynes wrote, ‘I put him very high among the younger economists in the country … He is of the calibre which would justify the immediate election to a Readership … He is a brilliant talker and one of the most attractive people about the place.’2

My friendship with Nicholas Kaldor, close and unfailing, lasted over a full fifty years. We first met in the autumn of 1937 when he was a junior presence at the London School of Economics, I a post-doctoral fellow pursuing Keynes and Keynesian economics at Cambridge University. I was frequently in London, came to know both the Kaldors and joined Nicky at the seminars of Hayek and Robbins.

Kaldor is famous as a leading member of that great generation of Cambridge economists — Joan Robinson, Piero Sraffa, Richard Kahn and himself — who together attempted the construction of an alternative paradigm to the neo-Classical/neo-Keynesian synthesis. For Joan Robinson and Richard Kahn, the engagement with Keynes’ economics was a continuous one and indeed part of the permanent revolution in economic thinking going on in Cambridge in the 1920s. Piero Sraffa started one phase of that revolution — the one in value theory — but his commitment to the Keynesian revolution was very slight. Indeed one could say that the two Cambridge revolutions, one started by Sraffa in 1926 and the other by Keynes in 1936 (although known earlier in Cambridge), were in contradiction with each other by the 1960s. The second Sraffa revolution begun in 1960 was a fulfilment of the one previously launched in 1926 but it had a value and capital theory that sat uncomfortably with the Keynesian revolution. Despite his intimate acquaintance with and encouragement of Sraffa’s work, Keynes in his General Theory took in his value theory a pre-Sraffa Marshallian line as far as he took any. Joan Robinson who with her exceptional rigour had worked out the consequences of the first two revolutions, Sraffa 1926 and Keynes 1936, bore the brunt of reconciling the second Sraffa revolution with the Keynesian revolution (See essays in Eatwell and Milgate, 1983).

Nicholas Kaldor is rightly famous for his many theoretical and empirical researches in the fields of microeconomics and macroeconomics. Among many other accomplishments, he independently discovered the von Neumann time-phased system, in which there are no primary (non-producible) factors of production and in which goods as outputs are produced out of themselves as inputs. This remarkable 1937 contribution is little known,1 much less known for example than Kaldor’s 1940 intuitive derivation of a unique limit cycle of determinate amplitude and period, that is asymptotically approached from any perturbed initial business-cycle position.

Kaldor begins his paper of 1972 with a resounding blast against what he calls ‘equilibrium economies’. The opening paragraph is worth quoting in full for its candid irreverence:

The purpose of my lecture today is to explain why, in my view, the prevailing theory of value — what I called, in a shorthand way, ‘equilibrium economics’ — is barren and irrelevant as an apparatus of thought to deal with the manner of operation of economic forces, or as an instrument for non-trivial predictions concerning the effects of economic changes, whether induced by political action or by other causes. I should go further and say that the powerful attraction of the habits of thought engendered by ‘equilibrium economics’ has become a major obstacle to the development of economics as a

science —

meaning by the term “science” a body of theorems based on assumptions that are

empirically

derived (from observations) and which embody hypotheses that are capable of verification both in regard to the assumptions and the predictions. (Kaldor, 1972, p. 1237)

The purpose of the present paper is to embed the compensation principle due to N. Kaldor into some more general framework so as to discuss conflict resolving by compensation for social as well as individual decision making. Kaldor’s intention in (1939) was to show that the classical argument for free trade does not require interpersonal comparison of utilities. In section 2 Kaldor’s compensation principle is presented in detail together with some of its criticism. In contrast to the criticism the formulation of Kaldor’s principle as given in this section does not involve utilities. This formulation then leads to a general concept of compensation presented in section 3. Compensation thereby is described by an equivalence relation on the set of alternatives. A conflict among alternatives occurs when there are (at least two) preferences that point in different directions. The conflict resolution hoped for is portrayed by the factor relation obtained by taking the conjunction of the preferences modulo the equivalence relation.

Lord Kaldor, or Nicky as most of us called him, was my teacher at the London School of Economics and later became a good friend. My debt to his teaching is evident from several of my early writings, which were based on ideas I first encountered in his published work and which I criticized or developed a little further. More fundamentally, I also learned from him to always focus on the facts (or as he called them ‘stylized facts’) of the real world and guard against being seduced by elegant theories into mistaking their simplifying assumptions for reality. Remember how he chafed against the assumption of linear homogeneous production functions underlying most economic theories, and how he tried to incorporate increasing returns to scale into growth theory.

This paper explores one aspect of the relationship between the system of production and the macroeconomic structure, namely the role of profitability in determining investment demand and the level of economic activity. Within the system of production, wages are a cost: the lower are profits per unit of production, the lower the stimulus to investment. In a Keynesian view of the macroeconomic structure, however, wages are a source of demand, hence a stimulus to profits and investment. In this view, aggregate demand provides the way out of the dilemma that high wages pose for the system of production. If demand is high enough, the level of capacity utilization will in turn be high enough to provide for the needs of both workers and capitalists. The rate of profit can be high even if the profit margin and the share of profit in output are low and the wage rate correspondingly high.

As is well known, the post-Keynesian theory of distribution was generated during the 1950s in Cambridge (Cambridgeshire). The first formal presentation was given in a seminal paper in 1956 by Kaldor. After that Kaldor utilized this theory in formalizing several growth models (Kaldor, 1957,1961; Kaldor and Mirrlees, 1962) in order to provide a solution to Harrod’s problem on the convergence of the ‘warranted’ growth rate to the ‘natural’ growth rate. After 1966 Kaldor did not return to the post-Keynesian theory of distribution except to clarify the origins of the theory (Kaldor, 1978, 1980).

The introduction of the corporate sector into the post-Keynesian theory of distribution has been worked out by Kaldor in an appendix to his 1966 paper, named, for obvious reasons ‘a neo-Pasinetti theorem’. Some reformulations, and precisions will be useful here before coming to comparisons with other approaches or formulations (especially Marris, 1964, 1971; Kahn, 1972; Wood, 1975; Tobin, 1969).

It was uncharacteristic of Nicholas Kaldor to take an ambiguous stand on any issue, but in the above it is not clear where he stood on the question of the endogeneity or the exogeneity of money, or whether he believed it to be of importance. In his later writings on money he viewed the attempt to control the path of nominal aggregate demand by controlling the path of an arbitrarily defined money supply — the fundamental policy posture of monetarism — as a ‘scourge’.2 If monetarism was a ‘scourge’ then the authorities by operating on interest rates could determine the supply of money (money supply is exogenous), but the overall impact of such policies was so adverse that it was not wise to do so. Once the price of monetarism became evident the authorities would have to accomodate the markets (money supply is endogenous).

Nicholas Kaldor gave superb testimony on monetarism to the Select Parliamentary Committee in 1980. There and elsewhere, he effectively criticized Milton Friedman’s empirical claims that Money causes Nominal Income. One of his criticisms was essentially econometric, namely that Friedman was using as evidence correlations over periods when central bank policies were accommodative. That is, the authorities were deliberately allowing money stocks to respond to income variations. Such correlations could not indicate what would have happened if the central banks had not been accommodative.1 Kaldor was surely right in making this econometric point. Of course, the fact that Friedman’s correlations don’t prove his case does not support any other proposition about the effects of non-accommodative policy.

In the investigation of how the social surplus was distributed, Adam Smith, David Ricardo, J. S. Mill and Karl Marx all viewed the rate of interest as some proportion of the rate of profits. Their analysis attempted to describe the factors which determined this proportion, and which prevented the rate of profit from falling to the rate of interest.

Professor Hahn, a distinguished theorist, has recently written: ‘The most serious challenge that the existence of money poses to the theorist is this: the best developed model of the economy cannot find room for it’ (Hahn, 1981, p. 1).

For Harry G. Johnson, perhaps the most famous Canadian economist, Nicholas Kaldor and John Hicks were ‘illiterate monetary policy amateurs’ (1978, p. 126). It is well known that Milton Friedman (1970) in his response to Kaldor’s ‘New Monetarism’, called him a Johnny-come-lately. In much more diplomatic terms, this was also James Tobin’s more recent assessment of Kaldor’s involvement in the monetarist debates (1983, p. 36). On the other hand, British authors, such as A. P. Thirlwall (1983, p. 43) and Grahame Thompson (1981, p. 68), have claimed that Kaldor’s memoranda on money presented the most effective repudiation of key monetarist assumptions.

The Keynesian business cycle follows a straight forward scenario. Begin in an expansion with a rising volume of transactions. Under tight money, interest rates rise. If they rise sharply enough investment is eventually reduced. If this depressing effect is strong enough, a recession is induced by the corresponding fall in aggregate demand; unemployment and excess capacity increase, the transaction demand for money decreases, and the interest rate falls. If the latter influence is sharp enough investment may overcome the depressing influence of excess capacity and low profits. Recovery sets in and the stage is set for a repetition of the story. A dramatic example of such interactions occurred in the early 1980s when the price level was brought under control with tight money policy; interest rates reached unprecedented levels and investment in some sectors came to a virtual standstill. When interest rates eventually fell, the impact on new housing starts and investment was immediate and substantial.

The model described here reflects concepts and methods developed in several different traditions of the analysis of capitalist economies.

Classical economics conceived of capitalism as an inherently expansive system which was ultimately regulated by its level of profitability. This approach reached its highest development in the works of Marx and Schumpeter, with their portrayal of a system driven by its inner mechanisms along erratic and periodically unstable paths of accumulation (Bleaney, 1976, ch. 6; Garegnani, 1978, pp. 183–5; Shaikh, 1984, section II). In what follows, I will refer to this overall perspective as the classical tradition.

In recent times nonlinear macrodynamic models with cyclical behavior have been revived, continuing a tradition starting with Kalecki (1937a, 1937b), Kaldor (1940), Hicks (1950), and Goodwin (1948, 1951). Most of these approaches, including Kaldor’s seminal contribution of 1940, have predominantly focused on real economic activities and have neglected the role of money and financial markets. Recently, there have been several papers which, partly arising from Keynesian theory and influenced by Minsky’s writings (Minsky, 1975; 1982), have attempted to integrate monetary and financial variables in macrodynamic models (cf. Taylor and O’Connell, 1985; Foley, 1987; Day and Shafer, 1985; Woodford, 1988). Our paper seeks to contribute to this line of research.

This paper is one in a series directed toward the construction, with certain modern building blocks, of a non-monetary theory of employment fluctuation in market economies. The closed-economy model here parallels the open-economy model in Phelps (1988). The objective is a plausible theory that helps to account for some or all of the long swings in economic activity over recent decades. In fact, this non-monetary theory has grown out of one of the models used by Fitoussi and Phelps (1988) to account for the 1980s depression over much of the world, a slump that demand-driven models are hard-pressed to explain.

Are there strong tendencies in laissez-faire capitalism toward steady growth at full employment? Are the causes of fluctuations in output and employment to be found outside the economic system or are they intrinsic to the system? The answer to these questions is fundamental to economics and more importantly, to almost all economic and political decision making.

Attempts to connect monetary phenomena and business cycles often have been the province of economists who also accepted the Walrasian full employment paradigm. Monetary explanations of business cycles preceded Keynes (cf. Zarnowitz, 1985), and in the post-Keynesian era Monetarists of different varieties have also advanced them. Old style Monetarists (e.g., Poole, 1978) have tried to integrate money into a disequilibrium account of fluctuations. By assuming a demand for money proportional to nominal income, lags in price adjustment, and an exogenous money supply, they depict movements of the money supply as a main source of fluctuations. In this scheme an increase in the money supply, with prices constant, translates into an increase in real demand. This causes an increase in real output, even if markets are initially at the point of Walrasian general equilibrium. However, when perceptions catch up with reality and nominal prices adjust, so will real output. Decreases in the money supply cause fluctuations in the opposite direction. New style Monetarists reject the disequilibrium elements of this story and insist on rational expectations. The assumptions of market clearing and rational expectations are integrated into a monetary theory of cycles by introducing new assumptions about the availability of information. Lucas (1981), for example, suggests that changes in the exogenous money supply cause changes in nominal prices in ways which are only partially understood — i.e., with Friedman’s notorious ‘long and variable lags.’

Many will agree that one of Kaldor’s most outstanding theoretical contributions was his theory of growth and distribution, which he illustrated by means of models for the years 1957–62.1 His interest in the matter did not end with this period, however, even though his subsequent research was not simply a continuation of his earlier work but revealed a change of view. Even in the fields where his work had already made him famous, Kaldor, as a true scientist, was dissatisfied with the results he had obtained hitherto and continued to extend the scope of his research.

One of Bert Brecht’s Geschichten vom Herrn K. goes like this. A man who had not seen Herrn K. for a long time greeted him with: ‘You haven’t changed at all!’ ‘O’ said Herr K. and grew pale.

In retrospect, Kaldor saw the fundamental shortcomings of the post-Keynesian growth and distribution models on which he concentrated his analytical work in the 1950s and the early 1960s rooted in the fact that all these models were one-sector models.1 Instead he advocated a two-sectoral model as the basis for gaining a thorough understanding of the nature of the growth and distribution process in a developed capitalist economy. Already the young Kaldor (1938) in a pioneering paper had pointed our attention to the facts that complementarity between equipment and labor is characteristic for modern technique and that full employment not only presupposes a certain level of real income (effective demand) but also a certain composition of production between consumption and capital goods because of the specificity of most equipment.

Three solutions have been proposed to the problem of reconciling Harrod’s ‘warranted’ rate of growth with the ‘natural’ rate of growth of the effective labor force so as to maintain full employment in long period equilibrium.1

All during his life, Nicholas Kaldor touched and investigated an impressive number of areas within economic analysis. Every economist knows his path breaking papers on speculation, non-linear models of the business cycle, his alternative theory of distribution, and so many other topics on taxation and economic and monetary policy. However, growth and development theories had been a recurrent theme for him all throughout his life. Around a basic core analysis, Nicholas Kaldor continuously revised his precise views about the factors limiting growth, whereas his hypotheses have been challenged. Still more, the breaking down of previous growth trends in the 1970s and the uncertain prospects about a recovery in the 1990s bring new questions into the cumulative causation model.

The late Lord Nicholas Kaldor made seminal and pivotal contributions to our understanding of the sources and character of trend and cycle in advanced capitalist countries. And yet we still lack a firm basis for evaluating the relative usefulness or explanatory power of Kaldor’s ideas. This paper attempts to provide one foundation for such an evaluation, constructing and evaluating the properties of a macroeconometric model of the US economy which embodies the central features of Kaldor’s macro system.

Disembodied technical progress is a factor of growth that appears in all kinds of growth models. As a matter of fact, in neo-Classical growth models it is the only type of technical progress since it lays in the foundation even of the different, sometimes rather complicated types of embodied technical progress.

Nicholas Kaldor is widely (and I think correctly) seen by economists as the most prominent and persuasive supporter of the proposal for an international Commodity Reserve Currency system (for short, ‘CRC’) in the second (post-German) generation of its advocates. He was very much aware of the work of the pioneers in the first generation — Benjamin Graham and Jan Goudriaan — and would not have countenanced any claim that he had founded a CRC movement. But he put CRC in the context of economic development problems, and made clear the essential fact that its logical form is not CRC monometallism but CRC/gold bimetallism.

It is a privilege for me to have this opportunity of paying tribute to Lord Kaldor for his immense contributions to the ideals of international cooperation. The very same generosity of spirit that had led him to associate himself with Sir William Beveridge in working on plans for full employment for Britain prompted him also to look far beyond the problems of Britain to the international economic objectives of the United Nations. He was indeed a United Nations man. Not only did he work for the United Nations in many different capacities, but his theoretical writings on world trade, development and the international monetary system as well as the technical assistance missions that he undertook to many different countries made an enormous reputation for him throughout the world, a reputation for trenchant analysis of the world’s ills and for brilliant advocacy of strategies for dealing with them. And his work was invariably permeated by a desire to reduce inequity and inequality in the economic relations among nations as well as in the internal economy of each country, rich or poor, developed or underdeveloped.

Comparing capitalism and socialism, it is important to remember that no actual economy is purely one type or another: all are mixtures bearing traces of their national histories, international relations and political compromises. Nevertheless, analytical study is best carried out at an abstract level in terms of pure types; prominent features of actual economies will be identifiable as belonging to one system or another and the logic of these features can be traced in the abstract system of which they are a part, where they have free play and full scope. We will treat capitalism and socialism as such abstract systems, and in doing so will draw on a central theme of Kaldor’s later years, the distinction between ‘demand-constrained’ and ‘resource-constrained’ systems, developed implicitly by Kalecki, but first explicitly defined by Kornai. This distinction requires replacing the scarcity-based theory of value with a Classical approach in which manufacturing prices are largely invariant to changes in demand.