The Duke Economists’ Papers Project has a grab-bag of papers from the distinguished economic historian Earl J. Hamilton. A soul braver than myself might some day try to create order out of that chaos, but I was able to stumble upon the following early “remarks” by future Nobel-prize economist John R. Hicks, though lacking all context save the date. Perhaps a Hicks expert or an historian of economic history can identify where these remarks were given (or perhaps eventually published?). These remarks sound much like Schumpeter’s recipe for a good economist writ large to economic research. I can only say, “Hear, hear!”
________________________
RESEARCH IN ECONOMIC HISTORY
John R. Hicks
January 11, 1947
The following remarks about the desirability of encouraging research in economic history are written from the standpoint of the general economist, who is not primarily a historian. He is not interested in economic history as history, but he is interested in furthering the development of economic science in general. He is looking for the general principles governing economic behaviour, and his particular interest is the application of those principles to the modern world.
As compared with the situation in the natural sciences, the economist’s object of study is essentially a historical process, spread out in time. In practice his main preoccupation is with the advancing edge of that process (the present), and it is right and proper that this should be so, since the present is more likely than the past to have a bearing on the future, control over which is the ultimate practical object. But this preoccupation can easily go too far. The past, no less than the present, is part of the material available for study and out of which generalisations can be built up. Generalisations based upon the present alone, or the present and recent past alone, are necessarily insecure; no doubt all economic generalisations are insecure, but these are more insecure than they need be.
The relevance of economic history to economic science has greatly increased of late, in view of the recent tendencies to bring economic theory to earth and achieve a more effective marriage between theory and statistics. Econometric work based upon very short time series is statistically unsatisfactory, and cannot be used as a basis for prediction with any high degree or probability. There is thus a tendency on the part of economic statisticians to push further back into the past as a means of increasing the amount of analysable material. But such additional material cannot be securely used unless its reliability is evaluated by people who are accustomed to use historical evidence—collaboration between the trained statistician and the trained historian (a very awkward collaboration with our present academic background) is going to be urgently needed at the next stage of development of economics. Further, it is not only the material which needs checking—the use which is made of it needs checking too. As we push backwards into history, institutions change; the whole background, economic, semi-economic and non-economic, changes. One of the commonest sources of error in economic reasoning is a failure to recognise that an institutional change has made a profound difference to the working of some particular “mechanism” or standardised response pattern. We notice this most often in a failure to “keep up to date”—the “out of date” economist is he who has failed to realize that a change in institution had modified or even completely destroyed some of the reaction patterns which may have been valid enough when he was young. The opposite error has hitherto been of less importance, but there are indications that it is now becoming serious; although it will never have the practical importance of that just described, it may be a serious impediment to scientific progress. To read the events of the past against an institutional background which is not theirs, is just as wrong as to read the accounts of the present against a background which is not theirs. Unless the background is in good shape, historical statistical data cannot be used; they can only be misused.
The above is not only an argument, as might appear at first sight, for better training of economists and statisticians in economic history; it is also an argument for research in economic history. For the sorts of questions which economists and statisticians are beginning to ask of the historical material are different from the questions which the historians have been asking. The historical background which is needed is not there, to be had for the asking, in the textbooks—or the classics—of economic history; to a large extent, it is yet to be discovered by new work.
I have here one example mainly in mind, though I am sure it is not the only example—not by a long way. The “Keynesian revolution” has thrown a powerful new light on contemporary economics; just how far the light extends is an arguable matter, but that it extends some distance can hardly be questioned. Now it would be of great help in our evaluation of the current uses of the Keynesian hypotheses if we could tell how far back in history they go on being useful. If it can be shown that they are useful in the interpretation of the economic history of the nineteenth or even eighteenth centuries, it would strengthen their position as a “General Theory”; if on the other hand, it becomes apparent that we have to force the historical material to get it into a Keynesian mould, we should get an indication of the dependence of the theory on a particular institutional (and perhaps psychological) set-up, and this would hardly fail to affect our attitude towards the theory and even our use of it vis a vis the problems of to-day.
I pass on to a much wider consideration. The ascertainment of economic principles or generalisations is only a step towards the understanding of events; one may say that the object of all economic inquiry—the penultimate object, perhaps, short of the ultimate object of increasing our control over the future—is to give an intelligible and analytical account of economic and economico-social processes, both the completed processes of the past and the uncompleted processes of the present. Now in some important ways the processes of the present are more difficult to study; they are more difficult because the sheer mass of material drives us to excessive specialisation, and also because their lack of completion deprives us in another way of the advantage of seeing the processes as a whole. In historical work it is at least in principle easier to take a synoptic view; and one cannot help feeling that if a rather larger proportion of economic research was devoted to historical problems it would help to maintain better standards of “all-roundness” in the sector—undoubtedly the more important sector from a practical point of view—which is concerned with the problems of the contemporary world.
This, in my view, is the case for encouraging research in economic history But I am well aware of the main difficulty which stands in the way of such research, if it is to be the kind of research which really meets the ends which I have set down. The number of people who have the equipment. to do the work—equipment in history and economics and probably statistics as well—is at present extremely limited. Work of this sort needs a bigger equipment than more specialised work, and therefore involves a longer preparation. At present there is little incentive to undergo this long preparation, and even for those people who have strong personal inclinations for it, there are strong incentives to turn aside on the way. In all the universities of the British Isles (I speak of what I know) there are at present only five chairs of economic history—two in London, one each at Oxford, Cambridge and Manchester. Apart from these, the subject offers very few openings indeed. Thus if greater encouragement were offered, one could not expect that supply would adjust itself to demand at once; it would take time before the number of suitable people could be much increased. If however one looks round at the people who have been diverted into teaching or research in “straight” economics or “straight” history, one can not doubt that the potential supply of first-rate economic historians is quite considerable; it would take time to show itself, but it would show itself in time.
Source: Duke University. Rubenstein Library. Earl J. Hamilton papers. Box 2, Folder “Correspondence—Misc. 1930’s-1950’s and n.d.”
