Economic Dynamics, Trade and Growth

Roy Harrod’s pioneering establishment of growth theory in the mainstream of modern economic analysis was one of the most significant legacies of the Keynesian revolution. He was close to Keynes from 1922 onwards and he saw the galley proofs of The General Theory in 1935. He was at the same time working independently on The Trade Cycle which he published in 1936, a few months after Keynes’s book. This contained the nucleus of his new dynamic economics which emerged as ‘An Essay in Dynamic Theory’ in 1939. In the Preface to The Trade Cycle Harrod writes, ‘Mr. Keynes, in his recent volume, The General Theory of Unemployment, Interest and Prices¹, has developed certain important ideas concerning the relations between the demand for capital goods, the propensity of the community to save, and its general level of activity and income. Full use is made of these.’ (Harrod, 1936, p. vii). In this 1936 Preface he also writes of his intention to establish dynamic analysis, ‘An attempt has been made in this essay to adopt a procedure in relation to the factor of growth similar to that of static analysis, to seek, namely, for the moving equilibrium of a steady rate of growth ... The consequences of this attempt are embodied in my theory of the “dynamic determinants”’. (ibid., pp. viii–ix).

Even a cursory reading of Harrod’s pieces on dynamics and his readers’ early comments reveals that they were at cross purposes, the former urging the need for a new approach to dynamics and the latter interpreting Harrod’s theory as a model of steady growth. A study of the literature on Harrod’s dynamics is needed to uncover any recognisable trends in interpretation and explain their origins. I am particularly interested in Harrod’s notion of dynamics and his insistence on the difference between his own approach and the ‘time-lag theories of the cycle’, and I will therefore concentrate on those criticisms (or the lack of them) that elucidate how the peculiarities of Harrod’s method and conception were (or were not) perceived and integrated by the profession. I am less interested in the analytical aspects of the mechanisms Harrod proposed to explain cycles and growth, and thus the details of debates on the conditions for stability, the classification of technical progress, and the several models aiming at representing Harrod’s ideas are excluded from this survey. (This literature has been covered by other authors, and references will be given where appropriate).

Every scholar of economic growth would acknowledge that Harrod’s contribution to economic dynamics is fundamental. However, when one observes the literature on the topic, the ambiguity with which his work has been received is evident. In fact, once original pieces of analysis are identified in his abundant and informal writings, they have been immediately criticized, denied any empirical relevance, and reduced to a very specific theoretical content. This Note will argue that Harrod’s critics have been not sufficiently perceptive of his originality, since his approach has been felt to be ‘discomforting’¹. Yet, if Harrod’s analysis is read more accurately, an interesting concept of economic dynamics emerges.

In all his writings on dynamics, Harrod was eager to stress the points of departure of his own conception from the alternative approaches to the study of economic change. At an early stage he disapproved of the psychological theories of the cycle; in The Trade Cycle he criticized the ‘time-lag theories of the cycle’, particularly Robertson’s lagged relationship between income and saving; he rejected Lundberg’s method of period analysis, and subsequently refuted Hicks’s definition of dynamics as requiring variables to be dated. Finally, from 1939 onwards, he concentrated his efforts in trying to emphasize the difference between his approach and the new orthodoxy in dynamics represented by Ragnar Frisch’s notion, which was quickly gaining widespread acceptance among economists¹. The range of Harrod’s arguments was as broad as the spectrum of his critical targets: he countered the merits of instantaneous vs. period analysis; he compared lags to ‘frictions’ and claimed that they are of secondary importance for a proper explanation of the cycle and indeed misdirect attention from its ‘true causes’; he stressed the irrelevance of lags to the analysis of equilibrium movement, and proposed a definition of dynamics based on the presence of a rate of growth among the unknowns to be determined in the fundamental equation.

Harrod’s analysis of economic dynamics does not consist of a compact and coherent set of statements. It is instead scattered over many writings belonging to different periods of his intellectual life. However, its impact on economists has been great, given its contribution to the birth of macro-dynamics.

This essay discusses and analyses how the balance of payments impinges on the growth performance of countries. This is important because mainstream growth theory still largely ignores the balance of payments. In classical growth theory, the balance of payments was assumed to look after itself through internal or external price adjustment, thereby severing any possible link between the state of the balance of payments and the use or accumulation of resources for economic growth. Harrod’s growth model (Harrod, 1939) was a closed economy model, and so was the neoclassical growth model (e.g. Solow, 1956) with the added objection that the demand side of the economy was completely ignored. Savings determine investment and aggregate demand equals aggregate supply. ‘New’ growth theory, or endogenous growth theory (see Romer, 1986; Lucas, 1988) is also supply-orientated — in which there are no demand constraints, either internal or external. Many of the ‘new’ growth models are closed economy models, and in those which are not, the focus is on growth and trade, not growth and the balance of payments. In the history of economic thought, the only school to have emphasized the importance of foreign exchange and a strong balance of payments for economic growth were the Mercantilists.

Of Sir Roy Harrod’s many contributions to economics, two stand out as being of particular importance. First, there was the development of the concept of the foreign trade multiplier in his International Economics, which was published originally in 1933, before the General Theory, but after Kahn’s (1931) formulation of the closed economy multiplier. Secondly, Harrod attempted to express the insights of the General Theory in dynamic form in a series of articles and books commencing in 1939. In the General Theory, Keynes (1936 p. 245) had taken ‘as given ... the existing quality and quantity of available equipment, [and] the existing technique’. As Eltis (1987) points out, Harrod’s observation that net investment implied that the capital stock would be increasing came as somewhat of a surprise to Keynes and the Circus. While Keynes’s approach, which ignored the effect of this increment to the capital stock and concentrated on the effect of investment on the level of income through the multiplier, was adequate for analysis of the short run, it was clearly deficient for the long run.

In his paper ‘Roy Harrod on the price of gold’ that appeared in the Harrod Festschrift, Harry Johnson (1970, p. 260) referred to Harrod’s ‘unique powers of intuitive insight into complex problems of political economy in the broad sense … and the relevance of his arguments … and his contributions to our understanding of the issues involved’ in the international monetary situation between 1953 and 1970. As early as 1953, in his report for the IMF on the problem of the ‘imbalance of international payments’, and up to the 1970s, Harrod had implicitly advocated a rise in the dollar price of gold in order to ‘correct existing imbalances’; a view which at the time was ‘unpopular’, but one which was later justified and shown to be intuitively correct (Harrod, 1953; Johnson, 1970, pp. 266, 276–278).