Theories of International Trade

Adam Klug, Michael Bordo, Michael Bordo

• Publisher: Routledge
• Year: 2006
• ISBN:
  9781134307050
• Edition: 1
• Topic: Business
• Subtopic: Business General
• Index: Business

Before his untimely death in 2000, the brilliant young Israeli economic historian Klug conducted a thorough survey into the different theories of international trade. The results of this are now available here for the first time with an introduction from Warren Young and Michael Bordo.

Utilizing the inter-temporal open economy model as a case study, Theories of International Trade illuminates the phenomenon of recurrence and the problem of recurring doctrines in economic thought and analysis. This compelling book will be of interest to scholars in the history of economic thought, and to international economists in general.

1 Introduction

I came to believe that almost all differences of opinion when analysed, were differences of method ... He who can throw most light upon the subject of method, will do most to forward that alliance among the most advanced intellects and characters of the age, which has been the object of all my endeavours.

J.S. Mill, Earlier Letters (1963: 78–9)

In his magisterial History of Economic Analysis, Schumpeter stated that: ‘Filiation of Scientific Ideas–the process by which men’s efforts to understand economic phenomena, produce, improve and pull down analytic structures... has met with more inhibitions in our field than it has in almost all others’ (Schumpeter 1954: 6). One of the characteristics of this slow rate of progress was the fact that, according to Schumpeter: ‘Much more than in physics have results been lost on the way or remained in abeyance for centuries. We shall meet with some instances that are little short of appalling.’

It is for this reason that Schumpeter regarded the study of the history of economic thought as a relevant activity for the economist:

Stimulating suggestions and useful if disconcerting lessons are much more likely to come to the economist who studies the history of his science than to the physicist who can, in general, rely on the fact that almost nothing worth while has been lost in the work of his predecessors.

Schumpeter did not in fact develop most of his work on this premise. He pointed out many fascinating examples of forgotten and obscure anticipation of important ideas, such as Sir Thomas More on oligopoly or Launhardt’s original proof of Hotelling’s theorem on marginal cost pricing (Schumpeter 1954: 308, 948). Nevertheless, Schumpeter did not catalogue and classify this phenomenon exactly, nor did he attempt to arrive at a systematic understanding of it. I therefore intend to make a study of this phenomenon in economics, which I call the ‘prevalence of recurring doctrines’. I shall be concerned later in this book to dispel any doubts the reader may have concerning the importance and pervasiveness of this phenomenon in the history of economics, but with the exception of one chapter, I shall not deal with this problem at a high level of generality. The economic field has simply become so vast that one must concentrate on a limited, esoteric aspect in order to exhibit the detail required in a single work. In addition, when working in this area one wants not only to interpret the evolution of economic ideas but to make some kind of contribution to the sum of our historical knowledge. Instead, I have focused on intertemporal theories of the balance of payments and exchange rates. These doctrines are of considerable potential and practical importance since they deal with world issues of pressing concern. On the other hand, they are quite typical of the formal nature of much recent macroeconomics, and therefore many conclusions apply to more than just this narrow subject. It turns out, however, that the earlier versions of these doctrines which appeared in the 1920s and 1930s have been almost completely forgotten.

I would now like to give the reader a feeling for the ostensible novelty of work in international macroeconomics that took place in the 1980s, and for the even more surprising manifestations of similar ideas in the 1920s, 1930s and 1940s.

The ‘new’ intertemporal international macroeconomics

In the 1980s, a ‘new’ approach came to dominate theoretical research on the open economy. This literature appeared to represent a complete break with the past of the subject. In fact, I shall devote much of this book to arguing that much of the ‘new’ approach of the 1980s had a history and even a prehistory. Yet to anyone brought up on the material which postdates the General Theory, the work of the 1980s will appear strikingly novel. The reason for this is the insistence that all explanations of fluctuations of the current account or exchange rates should be consistent with optimizing behaviour on the part of individual economic agents, where this behaviour is of the type to be found in microeconomic theory. At first blush this appears to be nothing more than an application of what has been termed the ‘microfoundations of macroeconomics’ to the problems of an open economy. In fact, as I shall be at pains to point out later, what was involved was not merely the ‘opening up’ of a particular class of macroeconomic models, but a fundamental change in the way economists thought about the concepts which were the object of their analyses. The phenomenon, whose behaviour is itself to be explained by the theory, the balance of payments itself, had a different significance under this approach. As will become apparent later in the exposition, the balance of payments is now held to record acts of intertemporal trade between coun- tries, i.e. exchanges of present for future commodities. Since rational farsighted individuals, behaving in the manner dictated by capital theory, would wish to continue such trades over time, there is, according to this approach, no theoretical reason for policy-makers to worry about the existence of current account deficits. Thus it has been stated that:

As for the question of whether one should worry about the current account situation, it is clear that there is Pareto optimal intergenerational allocation of resources when an optimizing private sector is left to itself in a competitive framework with perfect foresight, provided that there are tradable assets. The role of government is simply to transfer income across generations according to social norms.

(Razin 1982: 391)

This, of course, follows from the fact that ‘in these models, a large current account deficit is not necessarily a “problem”, it is a reaction of the market, given the availability of international financial markets, to an underlying disturbance to the economy’ (Stockman 1988: 54).

It is no exaggeration to say that these ideas appear to constitute a complete change with regard to the analysis of balance of payments or exchange rates which was prevalent throughout the history of economics. As documented extensively by Flanders (1989) there was from about 1800 until at least the early 1960s only one underlying model of the open economy–the most simple model of the monetary approach to the balance of payments (for example Dornbusch 1973: Sec. 1). This model has, of course, been drastically altered in many ways, such as elaborating the money supply process to take account of fractional reserve banking (Taussig and the Harvard School–see Flanders 1989: Ch. 7), introducing price rigidity (the early Keynesians), adding a nontraded goods sector (Hawtrey 1928, among many others), or allowing portfolio capital movements of various types. Yet, none of these elaborations had anything new to say about the determinants of the underlying behaviour which ultimately cause the adjustments which take place in the external sector. No notice was ever taken of the idea, expressed by the great poet William Blake, that a theory should be determined at some fundamental ‘molecular’ level: ‘Art and science cannot exist but in minutely organized particulars.’ No attempt was apparently made to open up the black box called ‘the balance of payments adjustment mechanism’ and look at the cogs and wheels inside. Furthermore, an equilibrium condition, namely current account balance, was always imposed on the external sector in these accounts, although, as stated above, there is nothing in individual optimizing behaviour which justifies this procedure.

Yet the adoption of a reductionist research programme–so called because of the emphasis on individual behaviour, and the consequent emphasis on intertemporal preferences as determinants of behaviour in the external sector–is not the only major change of emphasis which has occurred. The attitude towards government intervention is also a basic change, in particular because discretionary policy in the external sector is an old tradition which predates the Keynesian Revolution. It is true that those of the classical economists who saw themselves as successors of Ricardo and the Currency School in particular believed that under ideal conditions the mechanism of adjustment was automatic and imposed no costs on the economy. To the extent that frictions occurred, they believed that these could be overcome by a suitable design of the monetary system which, once in place, would allow events to take their natural and desirable course with no intervention by the authorities. As described in Fetter’s great work of scholarship (Fetter 1965) this attempt at creating an automatic currency system, embodied in Peel’s Act of 1844, was unsuccessful. As a result of this failure, the doctrines of Bagehot’s Lombard Street (1872) were eventually accepted, and the monetary authorities have since then practised a degree of discretionary policy in the external sector, in order to restore trade or current account equilibrium. The paradigmatic example of such intervention is the use, by central banks, of their discount rates to control payments imbalances during the classical gold standard period (1880–1914). Later, the scope for intervention was extended. Conversely, under the impact of the Great Depression, it was suggested that governments could also use external instruments to affect employment (an early example being Joan Robinson 1936), and finally in the 1960s it was argued that internal and external balance should be pursued simultaneously, and that the appropriate instruments (i.e. fiscal and monetary policy), would vary according to circumstances. (A classic reference is Mundell 1962.)

While the monetary approach to the balance of payments, popular in the 1970s, was sceptical about the scope for policy intervention or the real effects of devaluations (Dornbusch 1973: Sec. 1), even it did not mark a fundamental departure from the tradition sketched above. In comparing this with other approaches, Frenkel and Johnson wrote that ‘the monetary approach should, in principle, give an answer no different to that provided by a correct analysis in terms of the other accounts’ (Frenkel and Johnson 1976: 22). What occurred was a change in emphasis from the current account to the monetary flows to be found ‘below the line in the capital account’ (Frenkel and Johnson 1976: 23). Even the most sophisticated versions of the monetary approach did not depart from all of the previous literature, in the sense that ultimately the economy was required to be in a long-run equilibrium where the current account was balanced (Frenkel and Mussa 1985: 690–1). Not surprisingly then, the monetary approach, when augmented by some type of sticky wage or price mechanism in the goods sector, replicated many of the positive conclusions and policy recommendations of the ‘Neo-Keynesian’ era (see, for example, Dornbusch 1980: Ch. 9). We can conclude then that most analyses of the balance of payments, from the classical economists to the monetary approach of the 1970s, were essentially of a piece. The first reason for this is that they all used a concept of equilibrium which required trade or current account balance. This may refer to the short run, as in the elasticities approach or in income–expenditure models, or to the long-run dynamics only, as in some of the more sophisticated versions of the Mundell–Fleming model (which was the workhorse of international macroeconomics in the 1960s), but this definition of equilibrium was nevertheless always an integral part of the analysis. This, as I have stated above, is emphatically not the case with the ‘new’ intertemporal approach of the 1980s. The second reason for this apparent break with the past is that while some form of intervention, at the very least by the central bank, was advocated between the time of the Currency/Banking School controversy in the 1840s and the monetary approach in the 1970s, such discretion was trenchantly criticized in the 1980s by many writers using an optimizing approach.¹Of course, this literature does not completely rule out the need for government intervention, but to the extent that it does, it is for a completely different reason from those previously envisaged. The reason given is the lack of complete futures markets in the economy. The last characteristic of this literature is, of course, as stressed above, its uncompromising methodological individualism, a term originated by Hayek and Popper, which means in this context that economic processes should be explained by being deduced from the principles governing the behaviour of the participating individual economic actors, given that within the situation they find themselves in, they are subject to certain constraints.²It is from the application of this reductionist principle that all else follows. However, as I have stressed above, this has not necessarily meant the mere replication of the result to be found in other areas of economics with strong microfoundations but also the discovery of new insights into the determinants of the balance of payments.

The unknown antecedents

Everything I have said so far suggests that the work I have been surveying was something completely new, sui generis in the history of economics. This was certainly the view of the New Palgrave’s entry on ‘International Finance’, which states that work based on microeconomic foundations and rational economic behaviour began in the 1980s (Obstfeld 1987). It is hardly surprising that the contributor made this statement, since previous accounts of the historical development of the subject, from Viner’s classic (Viner 1937), through Kenen’s survey (Kenen 1985) of work covering the 1930s to the 1980s suggest nothing else. The main contention of this book will be that this is not so; in fact, I shall try to show that the theory of the 1980s had been substantially anticipated by previous writers, mainly in the interwar years. A number of economists wrote in the modern vein–Frank A. Fetter, Carl Iversen, Friedrich von Hayek, Pierre-Louis Reynaud, Ragnar Nurkse, Jacob Mosak and Oskar Lange. Of these, only Hayek has a strong influence on today’s economics, and only he, Lange, Nurkse and Fetter are remembered.³However, along with the others, nothing they wrote along the lines of an intertemporal theory of the balance of payments has found its way into any recent citations or apparently provided any inspiration for the thinking of today’s economists. This type of phenomenon, which, as stated by Schumpeter above, is not at all uncommon in the history of economics, has profound implications for any attempt to construct grand models of the development of the subject, or to import such schema from the history of science. It is these issues which I ultimately wish to approach by means of this particular case study.

Hegel wrote that ‘the narration of a number of philosophical opinions as they have arisen and manifested themselves in time is dry and destitute of interest’ (Hegel 1912: 307). I believe that this is as true for the history of economics as Hegel held it to be for philosophy. I therefore use a number of tools of intellectual inquiry, and consciously adopt more than one intellectual orientation, in order to make my interpretation as fertile as possible.

Exegesis and explanation

There are, in effect, two goals which a study of this kind attempts to achieve. These are a clear exegesis of the doctrines concerned, and an accurate explanation of why they were held by particular economists at a particular time. For example, when studying the famous and notoriously difficult chapter of Ricardo’s Principles, ‘On Machinery’, one should do two things: first, attempt to elucidate what Ricardo was trying to do, perhaps by means of a heuristic model, as in Barkai (1986); second, set out the machinery question in its historical context, linking it in particular to the economic problems of the time (as in Berg 1980). (To my mind, this is an exemplary work of its kind, which forcefully reminds the reader just how new and terrifying the industrial machinery was at the time of the Industrial Revolution.)

In this example, it turns out that economic theory is useful for the performance of exegesis, since it points out logical flaws or hidden assumptions in the work of an economist, while social and economic history are necessary adjuncts of explanation since they highlight particular areas of emphasis in a writer’s work which may otherwise seem anachronistic, and explain why he came to hold particular views. Thus, on the one hand, Ricardo’s arithmetical examples have been clarified, and on the other, his emphasis on particular aspects of the problem has been explained by the historical context in which he worked.

Researchers in the history of economic thought may attempt either exegesis or explanation. As pointed out by Laudan in his Progress and its Problems (Laudan 1977), intellectual history has concentrated largely on exegesis, and I do not think that the history of economics has been an exception to this rule. In this book, both an exegesis of rediscovered texts and an explanation as to why these doctrines have reappeared are attempted.

Explanations in the history of economic thought

Epistemology identifies various modes of explanation, such as causal explanation in the sciences, functional explanation in sociology and historical explanation, which is in a class of its own. It is not clear, however, what type of explanation is appropriate in the history of economic thought, since all of them appear to be relevant; economics has tried to offer scientific deductive explanations of phenomena, yet in order to explain its development, we must explain the actions of individuals in the past who are organized in a profession which has a social dimension like other professions. It therefore seems that any mode of explanation, causal (i.e. scientific), functional (i.e. sociological), or historical could constitute the correct methodological approach. Some of the other fields on the fringes of economics have adopted clear-cut methodological positions; for example, the ‘counterfactual’ methods of the New Economic History are obvious and conscious examples of causal explanation, while the ‘evolutionary theory of the firm’ invented by Nelson and Winter is an attempt at applying the functional explanations of sociology and biology to economics.⁴ In the history of economic thought, on the other hand, no dominant mode of explanation is apparent, nor has any rationale been worked out as to what mode of explanation is appropriate. In fact, while practitioners in the field have made methodological analyses of economic theory as a whole, they have left their own subject curiously immune from such scrutiny. Perhaps one has no choice but to be eclectic, given the interdisciplinary nature of the subject. My response is to use various explanatory methodologies. Each in turn may help to illuminate the problem of recurrence in general, and the recurrence of intertemporal theories of the balance of payments in particular.

Types of exegesis in the history of economic thought

It might be thought that exegesis is a less problematic concept than explanation from the cognitive point of view. This is true as long as exegesis is regarded as consisting merely of ‘explication des textes’ and placing them in the correct chronological order. There are, however, a number of ideal types of exegesis in the history of ideas, and their relevance to a particular study depends on the specific questions we are asking about a given text. In a brilliant essay, the philosopher Richard Rorty (1984) has classified some of the types or genres, as he calls them, which are to be found in the history of philosophy. They are, I think, equally relevant to the history of economics, and can be used to describe clearly and succinctly much of this thesis. The first such category is that of ‘rational reconstructions’ of the arguments of deceased scholars, in the hope of treating these thinkers as contemporaries, as colleagues with whom one can exchange views. Attempts at constructing models of pre-analytic systems of thought clearly fall into this category. Such rational reconstructions correct past mistakes and highlight anticipations of recent doctrines. I should point out that the term as used here is subtly different from its use by Popper, Lakatos and their followers.⁵For them, the object of the rational reconstruction is to recover a synthetic ‘problem situation’ faced by a researcher. This consists of his objectives and those logical interrelations between them which are manipulated by his rational actions. This type of procedure has often led to the accusation that Popperians rewrite history to suit their philosophical views. Whatever the truth of these criticisms by Feyerabend (1975: Sec. 16) and others, Rorty’s version of the concept of rational reconstruction, since it explicitly recognizes the way in which past texts are rewritten by the historian of ideas, is immune from them. Of greater importance, however, are ‘historical reconstructions’, which describe the views of the dead in their own terms. Such reconstructions permit a dialogue between, say, Malthus and Ricardo or Marshall and Wicksteed, but they must obey a constraint formulated once by the political philosopher Quentin Skinner: ‘No agent can eventually be said to have meant or done something which he could never be brought to accept as a correct description of what he had meant or done’ (Skinner 1969: 28).

Skinner stresses that such accounts must be written in such a way that they exclude the clearly absurd possibility that a scholar’s views were themselves dependent on the use of criteria of description and classification not available to the scholar himself. Models for such an account in our field would be Patinkin’s painstaking account of the development of Keynes’s thought, using as a reference point not later ‘Keynesian’ models, but Keynes’s own notes and letters, and Rymes’s reconstruction of Keynes’s lectures (see Rymes 1989). There exists a much more general genre which Rorty calls geisgeschichte. This is composed of works which reinterpret the past in terms of the present as rational reconstructions do, but do so on a broader scale.

Geisgeschichte asks why certain questions, usually tho...