Some years ago, George J. Stigler reminded the community of historians of economic thought that a great thinker of the last century, Adolphe Quetelet, had made a real methodological breakthrough in the social sciences by opening the door to quantification. Stigler himself tried to implement this method in the history of economic thought.

This paper is a sequel to De Vroey (1998d) and (1998e). In these papers, I pondered upon the relationship between the Marshallian and the Walrasian research programs and defended the view that a divide should be drawn between them, contrary to the opinion of the majority of economists who see no need for such. I argued that these research programs differ on two scores. First, they are based on different conceptions of equilibrium. Second, they differ on the way in which they broach the issue of the working of the decentralized economy. The hallmark of the Marshallian approach is that it proceeds in two steps. The working of particular markets is analyzed in a first stage whereas the issue of their coordination is assigned to the second stage of the inquiry. Contrarily, the hallmark of the Walrasian approach is to immediately start the analysis at the level of the economy as a whole. Put differently, in the Marshallian approach, partial equilibrium analysis is seen as a first preliminary step to general equilibrium analysis, whereas in the Walrasian approach one immediately proceeds with the latter. The aim of the present paper is to show the incidence of this twofold difference on the interpretation of Keynesian theory.

This issue of JHET marks the end of one era and the beginning of another. Don Walker, the only editor that JHET has ever had, has stepped aside after 9 years on the job. Under Don's guidance, JHET developed into a highly respected outlet for scholarship in the history of economic thought, to the point where the supply of high quality submissions prompted the move to quarterly publication in 1988. His labors on behalf of the Journal stretched well beyond the standard editorial function: he handled or oversaw all aspects of the physical production of the Journal, the mailing of the Journal, and numerous other messy details that the present editor has the luxury of passing off to Carfax (that, too, largely owing to Don's efforts).

In March 1776 when Etienne Bonnot, abbe de Condillac, published Commerce and Government, he was sixty-one years old, the same age as Francois Quesnay when he first published on economics. According to Jacqueline Hecht he had been a habitué of Quesnay's entresol (1958, p. 252), and Nicholas Baudeau has referred to Condillac's close friendship with Quesnay (1776/1903, p. 443). A book aimed to present a complete account of political economy from one of France's most distinguished philosophers, a member of the Academie franchise (“the immortals”) and the Royal Academy of Berlin, was bound to arouse great interest.

With the coming of marginal utility theory, the economists of the late nineteenth century were left with a theory of exchange values, but not with a theory of value. For example, William Stanley Jevons suggested that the concept of value be dropped from economics, leaving behind only vectors of exchange ratios (Jevons 1879). Not a few general equilibrium economists today hold much the same view. However, then, as today, there were seemingly fundamental economic questions that required a more general concept of value. Whenever economists turned to describing the movements in the economy over time or comparing the economies of different countries, they inevitably wanted to make statements about changes or differences in the real value of goods.

In recent years the reputation of Herbert Spencer (1820-1903) has undergone a transformation. Indeed, until the mid-1970s the answer to the rhetorical question, “Who reads Spencer now?” would surely have been “hardly anyone.” However, since then sociologist Jonathan Turner, intellectual historian Robert Bannister, political scientist Robert Perrin and dozens of economists and humanists have contributed to reconsideration of his role in nineteenth-century thought. They have also reinterpreted the corpus of his work and suggested new ways to utilize his contributions as a social theorist. Most important for our purposes is their effort to demonstrate that his stances on policy issues were not as mean-spirited as his critics have claimed. Indeed, Bannister, for example, forcefully argues that neither Spencer nor his American contemporary, William Graham Summer, were as consistently in favor of laissez faire or as harsh in their attitudes toward the underprivileged and the lower classes as scholarly convention once had it.

Our concern is John Stuart Mill's methodological pronouncements, his actual practice, and the relationship between them. We argue that verification played a key role in Mill's method, both in principle and in practice. Our starting point is the celebrated declaration regarding verification in the essay On the Definition of Political Economy; and on the Method of Investigation Proper to It (1836/ 1967; hereafter Essay): “By the method à priori we mean … reasoning from an assumed hypothesis; which … is the essence of all science which admits of general reasoning at all. To verify the hypothesis itself à posteriori, that is, to examine whether the facts of any actual case are in accordance with it, is no part of the business of science at all, but of the application of science” (Mill 1836/1967, p. 325). The apparent position that the basic economic theory is impervious to predictive failure emerges also in a sharp criticism of the à posteriori method:

In order that this essay may be clear to the reader, some understandings about terminology must be established. First, if we speak of “theories that were formulated in the past,” ordinarily we mean to leave open the question of whether they are still valid, and if we speak of “past theories,” ordinarily we imply that they have been discarded and that implies that they were defective and were supplanted by current theories. Nevertheless, in this essay the latter term, used for the sake of brevity, will have the same meaning as the former expression. If a past theory is considered to be invalid, that will be staled explicitly.

It is well known that the General Theory contains only one diagram, this occurring in chapter 14 criticising the orthodox theory of the rate of interest. Attached to the diagram is a footnote containing an acknowledgement: “This diagram was suggested to me by Mr. R.F. Harrod” (CW VII, p. 180). The impression conveyed by this remark is that Harrod drew the diagram and offered it to Keynes, who then accepted and used it. Further enquiry into the diagram's origins reveals the existence of an initial version which grew out of correspondence between Keynes and Harrod in August-September 1935 (CW XIII, pp. 526-63). Two of Harrod's writings–his (1951) Keynes biography and a letter to Hawtrey in 1951–imply that he was the author of the initial version, and that this was essentially the same as the published version. And, in seeming corroboration of Harrod's comments, there is the virtual equivalence of the initial version given in The Collected Writings of John Maynard Keynes (CW XIII, p. 557) and the final version printed in the General Theory (CW VII, p. 180J. Taken individually and collectively, these items of evidence all point in the same direction. They support the reasonable inference that, apart from trivial changes in lettering, Harrod was the sole author and creator of the initial and final versions of the diagram, and that Keynes was merely the recipient.

Albert Aftalion introduced Austrian marginalism into France and used it not only to tackle the issue of imputation (1911) but also to build his own theory of business cycles (1913) and of money (1927). Aftalion did not, however, receive the recognition he deserved for his contributions. The reasons are twofold. First, his use of modern analytical tools did not fit well with the dominant bias for the historicist approach that prevailed in France at the time, especially before 1914. Second, the fact that his contributions were available only in French did not help in spreading widely his ideas outside his native country.

Just over a century ago two noteworthy books in economics were published by women, contrary to the widespread modern impression that women's participation in economics is a recent phenomenon. Charlotte Perkins Gilman published Women and Economics (1898a, cf. Gilman 1898b), a landmark work of feminist and institutionalist economics (see M. A. Dimand 1995b). Rosa Luxemburg published The Industrial Development of Poland (1898) based on her 1897 Zurich doctoral dissertation. This distinguished contribution showed promise of the great abilities displayed in her magnum opus, The Accumulation of Capital (1913), a study that deserves recognition for its importance in macrodynamics and the theory of economic growth as well as in Marxian economics (see Joan Robinson's introduction to the 1951 translation). These two authors stand as giants in the history of economics–or would do so, if historians of economics were paying attention–but they were very far from being the only women contributing to economics in the 1890s.

William Petty (1623–87)—medical doctor, Oxford professor of anatomy, Gresham College professor, surveyor of Ireland for the Cromwellian Protectorate, Fellow of the College of Physicians, founding member of The Royal Society, and inventor of political arithmetic—is now most particularly remembered for his contributions to the formation of political economy. When Petty left Ireland for the last time in 1685 and returned to London, he took with him a very large collection of unpublished papers and correspondence. These largely remained in the property of his descendants—together with some important later additions—until 1993, when they were sold to The British Library. Petty's collected, published economic writings were edited and republished a century ago, together with some other materials (Hull 1899). In the 1920s Petty's descendant, the sixth Marquis of Lansdowne (H.E.W. Petty-Fitzmaurice), published two collections of manuscripts from what by then had become known as “The Petty Papers” archive (Lansdowne 1927, 1928). All these volumes have subsequently been republished in facsimile a number of times.

Roy Harrod (1951, p. 453) explicitly states that he “supplied a diagram purporting to reconcile the classical theory [of interest] with [Keynes's] theory, and he incorporated it in the [General Theory]-the only diagram in it!” (italics added). John M. Keynes (1936, p. 180n) himself writes: “This diagram was suggested to me by Mr. R.F. Harrod.” Furthermore, Harrod, in a letter to Ralph Hawtrey (19 June 1951) also says: “I was myself responsible for the diagram on page 180 which was an attempt to demonstrate to Maynard that his theories were not quite so out of line with the classical position as he supposed” (quoted in Warren Young 1987, p. 136). Now Rod O'Donnell (1999) claims that Harrod never supplied a diagram but only suggested how one could be drawn. Thus both Harrod and Keynes have misled us about the genesis of the only diagram in the General Theory. O'Donnell advances arguments in support of his claim which do not appear convincing.

Economics in the United States in the 1920s and 1930s was notable for the richness of its methodological and theoretical approaches. Encompassing the peak period of American institutionalism, these years also witnessed a recurrent debate over the proper scope and method of economics which was bracketed by a minor methodenstreit in the 1920s and the measurement-withouttheory dispute of the late 1940s. In retrospect it is apparent which lines of thought would dominate economic discourse in later decades. At the time, however, this future was not as clear. A late 1920s evaluation by Paul Homan of the state of contemporary economics concluded that economists “seem in our own day to be separated by more impassable barriers of thought than at any time in the past” (Homan 1928, p. 10). In looking beyond “the present impasse,” as he called it, Homan concluded that “whether economics in the future shall consist of a body of doctrines, or a body of facts scientifically ascertained, or a technique, or more or less of one and the other, is on the laps of the gods” (ibid., pp. 466-67).

The history of economics can be compared to a calm sea that once in a while happens to be shaken by heavy storms. This arises when works come out aimed at turning upside down the received interpretation of a great bygone economist's views. Professor Donald A. Walker's recent book, Walras's Market Models (1996), is likely to be among them. Its main thrust is that the view present-day economists have of Léon Walras is incorrect. The basic reason, he claims, is that to date all interpretations of Walras have been based on the last (posthumous) edition of the Eléments d' économie pure (henceforth the Elements), itself a slightly amended version of its fourth edition. To him this is a pity because Walras's most interesting theoretical ideas are to be found in its second and third editions—the embodiment of what he calls Walras's mature phase of theoretical activity—yet were abandoned by him when he revised his work for the fourth edition. The aim of Walker's book, then, is to bring to the fore the picture of what he considers to be the real Walras: an economist interested in the functioning of real-world markets and abiding by a realistic methodology who is attentive to the institutional set-up underlying his system of equations, and who is keen to provide his readers with disequilibrium models. In other words, Walker is trying to make the same claim apropos Walras as Axel Leijonhufvud (1968) did thirty years ago about Keynes when defending the view of a breach between the economics of Keynes and Keynesian economics. To Walker, modern Walrasian economics, or neo-Walrasian theory as it is more often called, is a betrayal of Walras's economics.

The continual interaction between economic change and economic policies designed to manage or guide this change seldom finds such dramatic expression as when one type of economy replaces another, e. g., when the industrial economy replaces the agrarian or is in turn supplanted by the “post-industrial” economy.

Thus, when the American economy was in the thick of its industrial revolution during the decades around the turn of this century, it was subjected to a series of government interventions the lowest common denominator of which has been summarized in the title of a book by Morton Keller, Regulating a New Economy (1996). In the United States as in all other countries, these interventions consisted of a glorious jumble of attempts to solve problems as they arose, of “bespoke jobs” in response to diverse economic interests, and of sundry ideologically-motivated efforts to move events in particular directions.

When Walras's Market Models was published, a friend wrote to me that he hoped that it would stimulate research on Walras's work. I am very pleased that his wish is coming true. A number of writers have reexamined it from new perspectives. It is, perhaps, inevitable that the quality and tone of the texts is uneven, and it is to be hoped that in the future all Walrasian research can be done carefully, in good taste, in a dispassionate fashion, and without rhetorical devices and patronizing remarks. In any event, I have presented in my book the results of 30 years of study of Walras's work, and I invite the reader to contrast what I have written with differing expositions. I have nothing to add to those results at this time, except the data which follow.

To declare my case “not proven” is irrelevant. I agree my case is not proven, but all other cases on this topic, including Ahiakpor's, are equally unproven. Arguing about proof here is a waste of time. We simply do not have enough data for deductive certainty or even for the less stringent criterion of “proof beyond reasonable doubt.” The written record is not conclusive and the protagonists are dead. What is not a waste of time, however, is discussing probabilistic inferences from incomplete evidence.