Mark Blaug

Economic theory in retrospect

Introduction: Has economic theory progressed?

Cambridge University Press, 5^th Edition, 1996, 1-8.

Introduction

This is a critical study of the theories of the past: it concentrates on the theoretical analysis of leading economists, neglecting their lives, their own intellectual development, their precursors, and their propagators. Criticism implies standards of judgement, and my standards are those of modern economic theory. This would hardly be worth saying were it not for the fact that some writers on the history of economic thought have held out the prospect of judging past theory in its own terms. Literally speaking, this is an impossible accomplishment for it implies that we can erase from our minds knowledge of modern economics. What they have meant to say, however, is that ideas should be weighed sympathetically in the context of their times, lest the history of economic thought degenerate into a boring exercise in omniscience. The danger of arrogance toward the writers of the past is certainly a real one - but so is ancestor worship. Indeed, there are always two sorts of dangers in evaluating the work of earlier writers: on the one hand, to see only their mistakes and defects without appreciating the limitations both of the analysis they inherited and of the historical circumstances in which they wrote; and, on the other hand, to expand their merits in the eagerness to discover an idea in advance of their own times, and frequently their own intentions. To put it somewhat differently: there is the anthropomorphic sin of judging older writers by the canons of modern theory, but there is also what Samuelson once called ‘the sophisticated-anthropomorphic sin of not recognising the equivalent content in older writers; because they do not use the terminology and symbols of the present’. For an example of the former, take Pigou’s reaction when asked to review a work on the Theories of Value before Adam Smith: ‘These antiquarian researches have no great attraction for one who finds it difficult enough to read what is now thought on economic problems, without spending time in studying confessedly inadequate solutions that were offered centuries ago.’ For an example of the latter, take the opening page of any doctoral dissertation on the works of a neglected forerunner."

“The conflict between those who regard earlier economic doctrine as simply ‘the wrong opinions of dead men’ and those who view it as the repository of a series of prescient insights goes deeper than economics. It is a fundamental division of attitude toward intellectual history as such. With a little training in German philosophy it is possible to represent the conflict in terms of two polar opposites: absolutism

and relativism. The relativist regards every single theory put forward in the past as a more or less faithful reflection of contemporary conditions, each theory being in principle equally justified in its own context; the absolutist has eyes only for the strictly intellectual development of the subject, regarded as a steady progression from error to truth. Relativists cannot rank the theories of different periods in terms of better or worse; absolutists cannot help but do so. Now, of course, few commentators have ever held either of these positions in such an extreme form, but almost every historian of economic thought can be placed near one or the other pole of what is in fact a continuum of attitudes to the theories of the past.

Either of the two positions is capable of further subdivision. One version of the relativist position, for example, is that the ideas of economists are nothing more than the rationalisation of class or group interests, or, to go one step further, the motivated pleadings of people with a political axe to grind. This is the doctrine of ‘ideology’ or ‘false consciousness’ which in its Marxist form is forever equating ideological bias with apologetic intent, though the two are by no means equivalent. The first edition of E. Roll’s History of Economic Thought (1939) perfectly exemplifies this approach, although in later editions the author goes no further than claiming that changes in economic institutions are ‘major influences’ a question begging phrase! - on economic thinking. Relativism is driven to extremes in W. Stark’s History in its Relation to Social Development (1944) which views theories as little more than a mirror reflection of the contemporary world: we are asked to believe, to open the book at random, that Ricardo was justified in advocating a labour theory of value in 1817 because fixed capital was little used at the time, but when he qualified the theory three years later he simply ‘yielded to the victorious march of mechanisation’. A singularly untenable version of the relativist interpretation is to be found in L. Rogin’s study The Meaning and Validity of Economic Theory (1956). In appraising the validity of an economic theory, relativists are always likely to ignore considerations of internal coherence and explanatory scope and to fix attention solely on congruence with the historical and political environment. But Rogin goes further and argues that the objective meaning of a particular economic theory lies in its practical policy recommendations; what is worse, he seems to mean by this, not the logical implications of a theory for the policy problems of its time, but rather the policy implications as they appear to a twentieth-century economist writing under the influence of the Great Depression. The trouble with the entire thesis is that economic theories are seldom devised to reach specific policy conclusions: time and time again, economists have recommended diametrically opposed policies while appealing to the same theory for authority.

In its moderate versions, the relativist interpretation can yield a really valuable fusion of the history of economic thought with the history of political and moral philosophy against the background of economic and political history. One of the best examples of this broad approach is W. C. Mitchell’s lecture notes on Types of Economic Theory (1949), which deliberately plays down ‘the passing on of ideas from one to another and the development of these ideas by successive generations’ as ‘an intellectual stunt’. The same viewpoint is upheld in A. Gray’s delightful intro-

ductory survey, The Development of Economic Doctrine (1931): ‘Economic science, if it be a science, differs from other sciences in this, that there is no inevitable advance from less to greater certainty; there is no ruthless tracking down of truth which, once unbared, shall be truth to all times to the complete confusion of any contrary doctrine.’ A glance at the latter portions of Mitchell’s or Gray’s book, dealing with the period after 1870, shows immediately what is wrong with the argument. Economics only became an academic subject in the 1880s, and thereafter, for the first time perhaps, ‘the passing on of ideas from one to another’ did dominate the development of the subject. No relativist has been able to carry the institutional or historical interpretation convincingly beyond the classical era that ended around 1870; and so, like Mitchell and Gray, they either neglect the modern period or, like Roll and Rogin, shift grounds in their treatment of economic ideas after 1870.

Speaking generally, it is absurd to think that economic and social history alone can furnish the key to intellectual variations in a discipline like economics. Many relativists claim only that economists write always sub specie temporis and that a knowledge of the prevailing historical context ‘illuminates’ the theories of the past. This is obviously true, but one wonders why it is necessary to argue this so insistently unless it is subtly designed to make us forget that ideas have a momentum of their own. As Jacob Viner observed, relativism frequently amounts to a kind of white-washing with historical necessity:

The economic historians seem to derive from their valid doctrine, that if sufficient information were available the prevalence in any period of a particular theory could be explained in the light of the circumstances then prevailing, the curious corollary that they can also be justified by appeal to these special circumstances. There are some obvious obstacles to acceptance of this point of view. It would lead to the conclusion that no age, except apparently the present one, is capable of serious doctrinal error. It overlooks the fact that one of the historical circumstances which has been undergoing an evolution has been the capacity of economic analysis. (Studies in the Theory of International Trade, p. 110)

No assumptions about economic behaviour are absolutely true and no theoretical conclusions are valid for all times and places, but would anyone seriously deny that in the matter of techniques and analytical construct there has been progress in economics? Adam Smith, for example, had a firm grasp of the way in which the market mechanism is capable of coordinating the independent decisions of buyers and sellers, but anything so fundamental as the functional relationship of demand and price escaped him. It never occurred to him that it was possible to demonstrate precisely in what sense a decentralised economy produces optimum results and when Walras and Pareto worked out the logic of Smith’s convictions a hundred years later, their demonstration of the optimal properties of a competitive regime bore no resemblance to Adam Smith’s views about the workings of ‘the invisible hand’. Thoughts such as these produce the absolutist who, looking down from present heights at the errors of the ancients, cannot help but conclude that truth is largely concentrated in the marginal increment to economic knowledge.

It is very likely that absolutists are created by reading the works of too many

relativists. It is difficult nowadays to appreciate the freshness of Cannan’s iconoclastic approach in his famous book The History of the Theories of Production and Distribution (1893) - a veritable catalogue of the elementary blunders of great economists - to a generation nurtured on the relativist texts of Blanqui, Roscher, Ingram and Cossa. Nevertheless, the recognition that economic theory has indeed progressed should not be allowed to obscure the highly uneven rate of improvement which has typified the history of analytical progress in economics. General insights into the pure logic of the price system make their appearance embedded in a particular theoretical framework associated with conditions and problems peculiar to the times. As the body of ideas gives way under criticism, much of what is still valuable gets discarded in an enthusiasm over the latest novelty. As a result, the history of economics is not so much the chronicle of a continuous accumulation of theoretical achievements as the story of exaggerated intellectual revolutions in which truths already known are neglected in favour of new revelations. Indeed, sometimes it seems as if economics has been propelled forward by a sense of symmetry which demands that every new theory should always be the exact reverse of the old.

In the first half of the nineteenth century, economics itself was regarded as an investigation of ‘the nature and causes of the wealth of nations’ (Smith), ‘the laws which regulate the distribution of the produce of the earth’ (Ricardo), and ‘the laws of motion of capitalism’ (Marx). After 1870, however, economics came to be regarded as a science that analysed ‘human behaviour as a relationship between given ends and scarce means which have alternative uses’ - an apt definition formulated in 1932 by Robbins, which, if taken strictly, would deny that much of what had gone before was economics. After two centuries of being concerned with the growth of resources and the rise of wants, economics after 1870 became largely a study of the principles that govern the efficient allocation of resources when both resources and wants are given. Classical economic theory was as much macro as micro-economics; neo-classical theory was little more than microeconomics; macro-economics came back into its own with Keynes and for a decade or so virtually replaced microeconomics. It is doubtful whether such dramatic shifts in the focus of attention can be explained solely in terms of intellectual forces - as absolutists are inclined to argue. In the final analysis, even pure economic theory is framed for the purpose of throwing light upon the actual workings of the economic system. A change of emphasis as drastic as the marginal revolution or the Keynesian Revolution must surely have been associated with changes in the institutional structure of society and with the emergence of new practical problems?

One possibility is that such shifts in emphasis within economics are due to changes in philosophical attitudes or dominant modes of reasoning. It was in opposition to this relativist interpretation that Schumpeter insisted upon the strictly autonomous nature of scientific economics. Although the political preferences and philosophical value judgements of economists impinge upon the development of economics, he declared, they leave it fundamentally unaffected: ‘economic analysis has not been shaped at any time by the philosophical opinions that economists happen to have’. This piece of dogmatic ‘positivism’, put forward in the introduce-

tion to his erudite History of Economic Analysis (1954), is not in fact sustained in the body of the text, half of which is given over to narrative history, political theory and philosophical climates of opinion, presumably because of their relevance to economic theory. Upon close inspection it turns out that Schumpeter did not mean that economic analysis is logically independent of philosophy but rather that the philosophical beliefs of economists are not relevant to the validity of the economic hypotheses they advance. The latter point is only too well taken. Witness the numerous pseudo-explanations that treat the history of economic thought in terms of a struggle between contending philosophical principles: ‘individualism’ versus ‘universalism’ - O. Spann, The History of Economics (1930); the biological view of the economic system as an organism versus the mechanical view of the system as a machine - E. Heimann, History of Economic Doctrines (1945); or, for that matter, value-free versus value-impregnated social science in G. Myrdal’s brilliant Political Element in the Development of Economic Thought (1953), which ridicules the effort to free economics from value judgements and, by implication, deprecates every analytical insight that is found to be associated with philosophical or political preconceptions.

Indeed, why stop at philosophical or political bias? W. Weisskopf in The Psychology of Economics (1955) gets the great economists to lie down on the couch, discovering, for example, a new significance in Petty’s famous remark that ‘land is the mother and labour the father of value’. For Ricardo and Malthus, he observes, the fecundity of the human female and the niggardliness of Mother Earth are the roots of all economic ills, while the only source of value is the ‘male’ factor of labour. But this is just how we would expect people to think in a patriarchal civilisation, he concludes triumphantly. Certainly it is conceivable that a knowledge of the psychological quirks of great economists might throw some light on their theories, but to infer the theories from the psychological association of words is to ignore the systematic logical character and empirical content of economic analysis.

It may be granted that, even in its purest form, economic theory has implications for policy and in that sense makes political propaganda of one kind or another. This element of propaganda is inherent in the subject and, even when a thinker studiously maintains a sense of Olympian detachment, philosophical and political preferences enter at the very beginning of the analysis in the formation of, as Schumpeter would have it, his or her ‘vision’: the preanalytical act of selecting certain features of reality for examination. The problem is not that of denying the presence of propaganda but that of separating the scientific ideas from the ideology in which they are invariably embedded and to submit these ideas to scientific tests of validity. Moreover, propaganda is not the same thing as lying: to say that Karl Marx wanted to discredit capitalism and began with preconceptions about its defects is not to imply that his analysis is for that reason worthless. Political prejudices may even assist scientific analysis: a critic of capitalism is likely to pay more attention to the real blemishes of the system and it is surely no accident, for example, that Marx’s comments on business cycles were fifty years ahead of his time.

The task of the historian of economic thought is to show how definite preconcep-

tions lead to definite kinds of analysis and then to ask whether the analysis stands up when it is freed from its ideological foundation. It is doubtful whether Ricardo would have developed his theory of international trade without a strong animus against the landed classes; but this theory survives the removal of his prejudices. When it came to proving that landlords would have no interest in making agricultural improvements, however, ideological bias prevented him from arriving at the correct result, correct, that is, in terms of his own assumptions. The history of economic thought is full of such examples, and nothing is gained by laying down flat generalisations about the relationship between the value judgements of individual economists and the quality of their theoretical work. Propaganda and ideology are always there, but so is the discipline exerted by rules of scientific procedure built into economics by generations of practitioners: economics is forever catching up with the biases of yesterday.

The problem that gave rise to economics in the first place, the ‘mystery’ that fascinated Adam Smith as much as it does a modern economist, is that of market exchange: there is a sense of order in the economic universe, and this order is not imposed from above but is somehow the outcome of the exchange transactions between individuals, each seeking to maximise his or her own gain. The history of economic thought, therefore, is nothing but the history of our efforts to understand the workings of an economy based on market transactions. But whereas received doctrine has always been concerned with the analysis of market economies, the structure of these economies has changed significantly over time and, in each generation, different concepts and methods of analysis have been employed to throw light on these changes. It is impossible to employ the findings of one method of analysis - appropriate to a particular economic environment - to pass judgement on the findings of another method appropriate to a different setting: one model cannot be used to judge another. Are we then driven into the arms of relativism? Surely, there are universal standards that can be applied to all theories?

Science, we have been told often enough, consists of the endless process of trying to falsify hypotheses. In that sense, the body of acceptable economic knowledge at any moment comprises all the theories that have not yet been falsified. But how are economic theories falsified? The great difficulty of testing economic theories, whether ancient or modern, is not so much the impossibility of making controlled experiments and thus disproving theories once and for all but rather that, lacking suitable laboratory conditions, economists (and for that matter all social scientists) cannot agree on definite empirical criteria for falsifying a hypothesis. Worse than that, they frequently disagree about the fundamental character of a theory. For example, was the neo-classical theory of perfect competition advanced as a hypothesis about how firms and households actually behave, or was it intended to furnish ideal standards for judging whether they behaved as they should? If the former is the correct interpretation, congruence with observed market behaviour is indeed the test of the validity of the theory, but if the latter, the fact that no existing market structure corresponds to the conditions laid down in the theory is a challenge to economic policy. It may be, of course, that the theory of perfect competition is both a ‘positive’ and a ‘normative’ theory, depending on the purpose for which it is used.

Positive theories about the social order cannot, in the nature of the case, be conclusively falsified by a single adverse result. An element of judgement inevitably enters into their evaluation and it is precisely for this reason that relativists and absolutists can continue to argue about the validity of the doctrine of comparative cost or the relevance of the labour theory of value. Normative theories, on the other hand, can never be evaluated solely by empirical tests. To make matters even more confusing, there are many examples in economics of theories which appear to be neither positive nor normative but merely taxonomic, providing an elaborate set of pigeonholes into which economic phenomena can be classified - Walrasian general equilibrium theory is a perfect example. Must we ruthlessly eliminate all such theories in the interest of ‘the principle of falsifiability’?

The history of economic thought is a proving ground for answering such questions. How much economics is simply taxonomy travelling in disguise? How have economists reacted to normative theories? What positive theories have been falsified by comparing their predictions with the real world? The answers lie in what economics has been: the practice of past generations still shapes what economics now is.

And so, has there been progress in economic theory? Clearly, the answer is yes: analytical tools have been continuously improved and augmented; empirical data have been increasingly marshalled to verify economic hypotheses, metaeconomic biases have been repeatedly exposed and separated from the core of testable propositions which they enmesh; and the workings of the economic system are better understood than ever before. And yet the relativists do have a point. The development of economic thought has not taken the form of a linear progression toward present truths. While it has progressed, many have been the detours imposed by the exigencies of time and place. Therefore, whether we adopt a relativist or absolutist interpretation of the subject depends entirely on the questions that we wish to raise. If a commentator is interested in explaining why certain people held certain ideas at certain times, he must look outside the sphere of intellectual debate for a complete answer. But if s/he wants to know why some economists in the past held a labour theory of value while others believed that value is determined by utility, and this is not only at the same time and in the same country but also in different countries generations apart, s/he is forced to concentrate on the internal logic of theory, willy-nilly becoming an absolutist.

We may sharpen the contrast we are making with the aid of a distinction, borrowed from the history of philosophy, between ‘historical reconstructions’ and ‘rational reconstructions’, a distinction that is almost the same as that between relativism and absolutism. ‘Historical reconstructions’ attempt to give an account of the ideas of past thinkers in terms that these thinkers, or their disciples, would have recognised as a faithful description of what they had set out to do. ‘Rational reconstructions’, on the other hand, treat the great thinkers of the past as if they are contemporaries with whom we are exchanging views; we analyse their ideas in our terms in order to locate their mistakes and to verify our fond belief that there has been progress in the course of intellectual history.

It is easy to show that historical reconstructions are literally impossible, while ratio-

nal reconstructions are invariably anachronistic. Historical reconstructions as such are literally impossible because we cannot travel backwards in time. In order to have any view at all of a text one must wear spectacles and these spectacles, unavoidably, must be the spectacles of the present if only because we can never forget what we now know: an element of hindsight is simply unavoidable in the same way that no adult can be asked to recall his or her childhood as if adulthood had never happened. It would seem, therefore, that we are driven towards something like rational reconstructions as the natural way of thinking about the history of ideas. But the trouble with rational reconstructions is that they can easily degenerate into omniscience. If the present generation does possess knowledge of absolute truth, there is little point to intellectual history except as a form of antiquarianism; in the words of one of Kenneth Boulding’s essay titles ‘After Samuelson Who Needs Adam Smith?’

What we have here is a standard Scylla and Charybdis problem. Historical reconstructions may be literally impossible but that is no reason for not trying to come as close as possible to a genuine appreciation of the past without benefit of hindsight. In other words, historical reconstructions are rather like the efforts of natural scientists to achieve objective knowledge of nature; strictly speaking, such knowledge is unattainable but we can aim at it in the hope of approaching it ever more closely. Similarly, rational reconstructions are plainly anachronistic and, if carried to undue lengths, destructive of any historical interest whatsoever. Nevertheless, if conducted in full knowledge of their anachronism, rational reconstructions are unobjectionable.

So, both historical reconstructions and rational reconstructions are each perfectly legitimate ways of writing the history of economic thought - if kept distinctly apart. Unfortunately, what is separable in principle is almost impossible to keep separate in practice: every interpretative exercise in the history of economic thought starts out either as a historical or a rational reconstruction but, in the course of argument, these tend invariably to shade into another. What is explicitly claimed to be a reworking of, say, Ricardo in modern dress is soon claimed to be at the same time a statement of what Ricardo really would have meant to say if only he had been as analytically advanced as we now are, as if a historical reconstruction is an unexpected bonus of a rational reconstruction. Likewise, although less frequently, many an examination of what Ricardo actually said ends up being at the same time a rendition of what he should have said, thus capping a historical with a rational reconstruction.

We shall return again and again in the text that follows to this tendency to run these two approaches together when methodological clarity instead demands that they be kept apart as explicitly as possible. I am not sure even now that I have always managed to avoid this tendency myself.