Thoughts on the New Economic History

Letters

Vol. 4 No. 9 · 20 May 1982

From Paul Johnson

SIR: David Cannadine’s critique of the new economic history (LRB, Vol. 4, No 7) goes about its business in a style of which he clearly approves: that is, ‘with a great deal of punching, if rather less judiciousness’. His forthright praise of the ‘heroic’ picture of the industrial revolution has led him to ignore or misunderstand many of the bold new designs. Some misunderstanding is excusable: the language of new economic history is often needlessly dense and jargon-ridden, sometimes downright ungram-matical, and seldom approaches the prosaic heights of ‘good old-fashioned economic history at its best’, telling the dramatic story of the transition from sweet Auburn to Workshop of the World. Part of the reason for this is that the plot is now acknowledged to be rather more complex and much more pedestrian than was once thought. The plight of, say, the handloom weavers was sudden and vicious and fully reported in the textbooks of the ‘old’ economic history. But this easily romanticised story was romantic because it was an exception to a general pattern of gradual change and adjustment to that change. One achievement of the new economic history has been to question ideas of take-off and leading sector, of discontinuity and climacteric; it must be remembered that Britain experienced the first and the slowest of all industrial revolutions.

The density of its prose is also in part the result of encroachment by the language of economics. Dr Cannadine accepts that economic ideas must be expressed in a precise and perforce somewhat artificial language, yet he seems to resent the need to become conversant with this new tongue. He likes stories about the appearance in the late 19th century of fish and chips and New Zealand lamb, but not the analysis of consumption propensities, price differentials and demand elasticities that alone can give some explicatory basis to such stories. When a bad workman is lacking in skill, he should not rush to blame the new tools of his trade.

It is with these tools of new economic history that Dr Cannadine finds the greatest fault, objecting in particular to counterfactuals, cliometrics, and the rigidities of neoclassical economic theory. He has some sensible things to say, particularly on the futility of the sophisticated statistical manipulation of primitive and unreliable data, but he should not pretend that the be-all and end-all of new economic history lies in the application of the, tape-measure to Clio. Measurement is certainly important, and this has always been true of economic history: what is new is that the measurement is consistent, and the application of it is consistent, too. Instead of cameos of smoking chimneys and accumulating capitalists interspersed with the odd (sometimes very odd) fact or figure, we now have accounts explicitly based upon thoroughly researched statistical series. This statistical rigour cannot always provide concrete answers to such questions as ‘What happened? How did it happen? What is the dynamic interaction between the different factors of production?’, but without the details of population, employment, output, productivity, trade and growth there can be no answers at all.

Consistent statistics are an important element of the new economic history, but conclusions drawn from them depend crucially upon the questions asked, and it is in the formulation of questions that new economic history most clearly breaks with the old. Assumptions are made explicit and hypotheses are framed in ways that can be tested by the application of coherent economic models. The mysteries of en-trepreneurship are turned into theories of constrained profit-maximisation or cost-minimisation, by which success or failure can be measured. Naive assertions about the ‘necessity’ of railways for British economic development are tested by assessing the additional cost in any particular year of moving all rail passengers and freight by the next best available means of transport. The estimation of the indispensability or otherwise of any particular factor, which Dr Cannadine considers so ‘bizarre’, has been essential to the demythologising of ‘heroic’ economic history, to the fuller understanding of the intricate process of economic growth. Heroic events are not untrue, they, do provide glimpses of what happened, but they seldom serve to answer the question ‘How?’. The Jarrow hunger march should: have a place in any account of the British economy in the 1930s, but one event cannot be allowed to; dominate the study of consumption and saving, of; tax and exchange rates, of investment and the building cycle.

The deliberate conjunction of economic theories and historical data to answer clearly defined questions has given economic historians the problem of deciding how appropriate is any modern theory as an explicator of economic relationships in bygone times. As Dr Cannadine points out, neoclassical theory assumes much which for 18th and 19th-century Britain is demonstrably untrue, and though: these assumptions are not necessarily crucial, they I must induce some doubt about the results obtained. It is not true, however, that all new economic history ‘presupposes perfect competition, full utilisation of resources, rational and profit-maximising entrepreneurs, and non-interference by government’, as studies of oligopoly and cartelisation, of tariffs and subsidies, make clear.

There is a growing awareness among new economic historians that many of the analytical tools they borrow from the economist’s work-box are crude and blunt, and give a heavy-handed appearance even to the most skilful practitioner’s output. Economic history can play a part in sharpening and refining these tools, especially by questioning assumptions about tastes and preferences, about motivation and response to incentives, and about conformity across classes, cultures and centuries. This will not require any reduction in methodological rigour or statistical accuracy, though it will need a fuller appreciation of historical detail, a greater feeling for subject and period, than most of the new economic history has shown to date. We should not sacrifice the advances of the last twenty years just to return to ‘exciting’ stories and ‘heroic’ events. Economic history in the 1980s needs a little more Clio, perhaps, but no more heroes.

Paul Johnson
Nuffield College, Oxford

Vol. 4 No. 11 · 17 June 1982

From David Cannadine

SIR: Paul Johnson’s reply to my thoughts on the new economic history (LRB, Vol. 4, No 9) is a fine example of grudging and unintended agreement masquerading unconvincingly as belligerent and wayward disapproval. En passant, he admits that ‘the language of the new economic history is often needlessly dense and jargon-ridden, sometimes downright ungrammatical’; that I have ‘some sensible things to say’ about ‘the futility of the sophisticated statistical manipulation of primitive and unreliable data’; that ‘statistical rigour cannot always provide concrete answers’ to a variety of important questions; that ‘neoclassical theory assumes much which for 18th and 19th-century Britain is demonstrably untrue’; and that the subject needs ‘a fuller appreciation of historical detail, a greater feeling for subject and period, than most of the new economic history shown to date’. I am, of course, delighted that my punches should have encountered so judicious a target – although this does leave me wondering what is left of the subject’s supposed ‘bold new designs’ which I have, apparently, ignored or misunderstood.

Having conceded so much of my case, it is hardly surprising that Johnson makes heavy weather of defending the remnants of his scuppered citadel. ‘The mysteries of entrepreneurship,’ he writes, ‘are turned into theories of constrained profit-maximisation or cost-minimisation.’ But no such treatment is to be found in the first of Floud and McCloskey’s volumes. He seems to think that I ‘like’ stories of fish and chips, when I was merely trying to convey the flavour of the book in question and to compare it favourably with another work. He claims that I dislike ‘the analysis of consumption propensities, price differentials and demand elasticities’, when I expressed no opinion on these concepts one way or the other – except to suggest that Supple’s quantitative and Fraser’s impressionistic approaches to the study of late 19th-century consumer demand were complementary rather than mutually exclusive. ‘When a bad workman,’ Johnson continues, presumably with me in mind, ‘is lacking in skill, he should not rush to blame the skills of his trade.’ Agreed. But the real cause of Johnson’s anger emerges later in his letter, when he concedes, with understandable coyness, that there is also ‘a growing awareness among new economic historians that many of the analytical tools they borrow from the economist’s workbox are crude and blunt’ (my italics). How curious it is to be attacked for holding a view which one’s critics seem to share!

The remainder of Johnson’s remarks shows a regrettable ignorance of the historiography of economic history, which leads him to understate the accomplishments of the old-fashioned economic historians he so cavalierly derides, and so to overstate the originality of the new economic historians he so unconvincingly defends. ‘One achievement,’ he writes, ‘of the new economic history has been to question ideas of take-off and leading sector, of discontinuity and climacteric’ But this claim to novelty (and monopoly) is untenable: these ideas were being questioned by old-fashioned economic historians from the time they were first formulated. Likewise, he argues that the counterfactual has been ‘essential’ to the ‘demythologising’ of ‘heroic’ economic history. But this shows equal disregard for much pioneer work by old-fashioned economic historians, and displays a naive belief that overturning one mythology will not lead to the enthronement of another. (In any case, since his colleagues are so anxious to insist that nothing is indispensable, how can the counterfactual be ‘essential’ for anything?) Finally, he suggest that events such as the Jarrow hunger march should not be allowed to ‘dominate the study of consumption and saving, of tax and exchange rates, of investment and the building cycle’. But when have they ever done so: where are these unnamed studies of ‘consumption and saving, of tax and exchange rates, of investment and the building cycle’, dominated by the Jarrow hunger march?

As described and defended by Johnson, the new economic history emerges as a subject regrettably ignorant of the scholarly tradition to which it belongs, over-ambitious in the novelty and substance of its claims, better able to tilt at defunct windmills than to make constructive arguments – and therefore unsurprisingly vulnerable when over-extended into a two-volume textbook. Less time in the microchip factory and more in the fish shop might well prove salutary.

David Cannadine
Christ’s College, Cambridge

Vol. 4 No. 12 · 1 July 1982

From Paul Johnson

SIR: David Cannadine’s response to my comments on new economic history (LRB, Vol. 4, No 11) is yet another example of his injudicious punching. So eager is he to claim a victory on points that he has failed to see just where and why the fight is taking place. Because I find so many faults with the work of new economic historians, Dr Cannadine believes I concede his case and am left defending a ‘scuppered citadel’. If he had understood my letter, he would have realised my purpose was not to defend the works of wayward cliometricians, but to identify the methodological strength of new economic history.

The strength lies in the clear expression of ideas and underlying premises: to reiterate, ‘it is in the formulation of questions that new economic history most clearly breaks with the old. Assumptions are made explicit and hypotheses are framed in ways that can be tested by the application of coherent economic models.’ In other words, pictures and stories are not taken at their face value, the ‘facts’ are not allowed to speak for themselves, but, instead, they are consciously moulded into the analytical framework provided by the historian. So much is true of some old economic history: what the new does is boldly display this analytical framework, in the belief that only by revealing the inner workings of the subject can sense be made of the questions asked and answers gained. Instead of hiding their assumptions about ‘the really important historical questions’ like ‘What is the dynamic interaction between different factors of production?’ in a maze of ‘heroic’ and ‘exciting’ stories, these new practitioners make explicit the models of dynamic interaction they appeal to, and the elements of interpretation, extrapolation or guesswork they resort to in order to reconstruct missing links in the chain of causation. If the models or guesses are inappropriate and the links malformed, they deserve (and receive) detailed and trenchant criticism from those who possess the skills necessary to understand the subtlety of the original argument.

Dr Cannadine, of course, does not indulge in this skilful criticism of the specific form the links in the chain take: rather, he doubts that the chain leads anywhere. New economic history, he says, is ‘better able to tilt at defunct windmills than to make constructive arguments’. The reason for this is the ‘ominously ahistorical nature’ of an enterprise wedded to neoclassical theory. The problem with framing testable hypotheses is that the models used to test them are crass. Agreed, many economic models are so simple as to be impotent (and neoclassical models do not have a monopoly here), and this is why I suggested that economic historians should be much more questioning of the models they use. If new economic history is at times emasculated, not by its methodological awareness, but by the imperfection of the analytical tools it borrows from the economists’ work-box, then it needs to hone these tools to a finer point, not dismiss the methodology and return to the cosy and vague eclecticism of old economic history. But the weakness of the models is often overstated. Even simple demand theory, for instance, tells us a great deal about market behavious. Downward-sloping demand curves might be a starting-point for any assessment of changes in consumption patterns in the late 19th century, and would encourage a discussion of price and income effects. Did the chip shops spread across the country because the price of their goods fell, because real income rose, or because of a fundamental change in tastes? The questions are explicit, and the alternative hypotheses are testable with a simple economic theory that even historians have heard of. Yet there is not a hint of such analysis in the work of W. Hamish Fraser, whom Dr Cannadine strangely associates with Clapham, Ashton and Postan as a practitioner of ‘good old-fashioned economic history at its best’ (a rather cavalier approach to the historiography, surely?). The great weakness of so much old economic history is that it has no sense of the underlying processes that condition economic activity, and makes little attempt to gain an understanding. It is content to tell stories rather than ask questions.

Dr Cannadine seems happy to listen to these stories rather than explore the epistemological sophistication of new economic history. He finds the subject unexciting because he cannot or will not raise his sights above the level of chip-shop counters and smoking chimneys to recognise the intellectual challenges of this new approach to economic history. It is a subject for thinkers, not raconteurs.

Paul Johnson
Nuffield College, Oxford

Vol. 4 No. 13 · 15 July 1982

From David Cannadine

SIR: How thoughtful Paul Johnson is (LRB, Vol. 4, No 12). He thinks the old economic history is sometimes bad, sometimes good. I agree. He thinks the new economic history is sometimes good, sometimes bad. I agree. He may even find I said as much in my review. Meanwhile, he wishes to ‘explore the epistemological sophistication of the new economic history’. Good luck to him.

David Cannadine
Christ’s College, Cambridge