ECONOMIC METHODOLOGY SINCE KUHN 571 CHAPTER THIRTY- FOUR Economic Methodology since Kuhn John B. Davis 34.1 INTRODUCTION This chapter surveys contemporary thinking about economic methodology sub- sequent to Thomas Kuhn’s Structure of Scientific Revolutions (1962), in an effort to explain its recent development, current nature, and possible future. Economic methodology since Kuhn might be characterized as having been in something of a state of permanent revolution. Before Kuhn, most economists adhered to a relatively small number of enduring methodological views: Lionel Robbins’s conception of economics (Robbins, 1932), Milton Friedman’s instrumentalism (Friedman, 1953), and verificationist views that arose out of logical positivism. Of course, there were methodological debates in the years immediately before Kuhn (cf., Blaug, 1992 [1980]; Caldwell, 1982) – Terence Hutchison versus Fritz Machlup, Friedman versus Richard Lester and the critics of marginalist analysis, and Tjalling Koopmans versus Friedman and Robbins (and indeed even older debates such as the famous Methodenstreit) – but these debates did little to change thinking about economic methodology prior to 1960. Subsequent to Kuhn, however, the situation changed dramatically. Although new views prevailed for brief periods – Karl Popper’s falsificationism, Imre Lakatos’s methodology of scientific research programs, D. McCloskey’s rhetoric of economics approach – it cannot be said that these new views came even close to acquiring the lasting status and breadth of appeal that their pre-Kuhnian predecessors had possessed. Currently, no methodological views enjoy widespread acceptance either among economists or economic methodologists. At the same time, the very volume of methodological thinking about economics today significantly exceeds, in both quantity and 572 J. B. DAVIS diversity of ideas, the total of all the methodological thinking about economics that predates Kuhn (cf., Hands, 2001). What happened to bring about his change? One fact stands out. Prior to Kuhn, there were very few individuals who speci- alized in economic methodology as a relatively independent field of investigation, and economic methodology was, rather, only one area of concern for practicing economists within the general field of economics. But in the years that followed this was less and less the case, as the field of economic methodology, along with the history of economics, gradually became a distinct sub-discipline and the pro- vince of a new set of specialists – particularly after 1980, with the first appearance of books devoted explicitly to the field (Stewart, 1979; Boland, 1982; Caldwell, 1982; Blaug, 1992 [1980] ). This had two important consequences for the relation- ship between economic methodology and economics as a whole. First, since most economists were not current with the new developments in methodology – and were often aware that they weren’t – they either abandoned interest in the field or continued to half-heartedly support old pre-Kuhnian ideas. Secondly, eco- nomic methodologists, cut loose as a relatively independent sub-discipline from a more inertially evolving economics, were free to explore a whole range of new ways of thinking about economics. In fact, this new thinking took the form of reasoning in epistemological, historical, and sociological terms about economics. After Kuhn, then, economic methodologists increasingly turned away from reasoning in terms of economics to, rather, reasoning about economics in epistem- ological, historical, and sociological terms – a form of thinking that was unfamiliar and uninteresting to most economists. The high rate of change in methodological thinking since Kuhn’s book, com- bined with the considerable diversity in ideas that now characterizes the field, alone justifies our saying that the field is currently in a state of permanent revolu- tion. But the separation between economics and economic methodology since Kuhn suggests that the field of economic methodology is also in a state of permanent crisis. On the one hand, economic methodologists since Kuhn have developed a multitude of competing and intriguing new perspectives on economics with which to explain the nature and practice of economics. On the other hand, these new perspectives have been consistently devalued in virtue of their rejection or – perhaps even worse – their simple disregard and neglect by those in the field to which they apply. The crisis in the field of economic methodology, then, is a crisis of “exciting irrelevance”; that is, one that comes of an unwillingness to give up on any of a multitude of intriguing new perspectives, combined with a nag- ging sense that they may ultimately all constitute uninteresting, if not mistaken, pathways. This state of affairs has created considerable pluralism in economic meth- odology regarding the nature and direction of the field, a situation that is likely to persist for the foreseeable future. This pluralism can be seen in terms of a number of unresolved issues that have divided and seem to continue to divide methodologists: (1) whether economic methodology should be prescriptive or descriptive; (2) whether economics should be understood as a natural science or a social science; and (3) whether economic methodologists, together with his- torians of economics, should shift their focus more toward a general history and ECONOMIC METHODOLOGY SINCE KUHN 573 philosophy of science framework. These three debates are addressed in the next three sections, followed by concluding remarks on the state of contemporary economic methodology in the fifth and concluding section. 34.2 SHOULD ECONOMIC METHODOLOGY BE PRESCRIPTIVE OR DESCRIPTIVE? Thinking on this issue has gone from the idea that economic methodology should be prescriptive through the idea that it should be descriptive to contemporary views that mix both postures in different ways. The opening phase in this sequence is most evident in the logical positivism of the 1930s, which employed a highly prescriptive verificationist criterion of meaning that asserted that the only mean- ingful (and thus scientific) propositions were those that could be empirically verified. Although there was much debate over how to formulate this criterion, it was widely agreed that its purpose – to demarcate science from nonscience (often disparagingly termed “metaphysics”) – was necessary to the advancement of science. By 1960, pre-Kuhnian philosophy of science and economic methodology was almost entirely demarcationist in attempting to set out rules for identifying good science (Hands, 2001). Even those not fully in the empiricist tradition shared this general aim. In the marginalist controversy involving Lester and Machlup, regarding whether actual businesses maximize profit (cf., Mongin, 1998), Machlup argued that good economics needs to employ theoretical assumptions that are not themselves empirically verifiable, but which enable us to reject empirical results that may be at odds with our theories (Machlup, 1955). Alternatively, in connection with the same debate, Friedman argued that we make good the- oretical assumptions even when they are unrealistic if they nonetheless allow us to make reliable empirical predictions (Friedman, 1953). For both Machlup and Friedman, then, demarcating the practices of good economics was the core prin- ciple of economic methodology. However, the high point for prescriptivist economic methodology was Popper’s falsification criterion, which was perhaps even more focused on demarcating good and bad science than the verification criterion had been, since it (at least in its simplest form) makes entire theories answer to a single disconfirming test. Indeed, Popper coined the term “demarcation,” and used it to distinguish science and “pseudo-science,” examples of which, for him, included Marxism and Freudianism (Popper, 1963, p. 255). In fairness to Popper, identification of his methodological views with the falsification criterion overlooks the critical ration- alist side to his thinking and the “Socratic Popper” (Boland, 1997, p. 263). But the traditional Popperian view was nonetheless for a time thought by many to embody the highest virtues in economic methodology (e.g., Blaug, 1992 [1980]). One meas- ure of the significance of Kuhn’s book, then, was that it was seen not only as discrediting Popperianism, but also as having cast doubt on the entire prescriptivist project of formulating demarcation criteria for science (e.g., Laudan, 1983). As an historian of science, Kuhn’s goal was to examine the actual historical development of scientific theories. Thus his aim was to describe how scientific 574 J. B. DAVIS theories came to be accepted, rather than to judge which ones deserved the label “scientific.” Moreover, his particular view of the historical developmental pro- cess in science emphasized revolutionary transformations in thinking, in which one paradigm replaced another. How one paradigm replaced another, however, was left largely unexplained, in large part because Kuhn saw this process as his- torically contingent. Scientists worked within paradigms, doing “normal” science and ignoring empirical anomalies that were at odds with the reigning paradigms, until an accumulation of these somehow precipitated the movement of (usually a new generation of) scientists to a new paradigm. That this involved a “revolu- tion” not only implies that there were no rules by which competing paradigms might be judged, but also that paradigms were incommensurable. External criteria for judging the scientificity of a paradigm, then, simply did not exist. Kuhn’s views were also taken to support the idea that observation is theory-laden, thus undermining the reliance that both the logical positivists and falsificationists had placed on judging theories according to how they measured up to evidence. The major response to Kuhn in the prescriptivist tradition (indeed, one adopted by many of those who abandoned Popperianism) was Lakatos’s methodology of scientific research programs (MSRP) strategy. Rather than a single theory, Lakatos focused on networks of interconnected theories, or “research programs,” which could be characterized in terms of “hard cores” surrounded by “protective belts” (Lakatos, 1978). The former were not subject to criticism, but developments with respect to the latter could be characterized with hindsight as either progressive or degenerating. A research program is progressive if it is both theoretically progressive – new theories include new empirical content – and empirically pro- gressive – this new content is corroborated and involves the prediction of “novel” facts (Lakatos, 1970, p. 118). Lakatos thus accommodated Kuhn’s historical emphasis while yet retaining elements of a demarcationist approach. However, this combination was not to be sustained, primarily because few agreed that Lakatos’s account of progressivity was adequate, especially in connection with his emphasis on novel facts. Indeed, after Kuhn the entire demarcation project was increasingly looked upon more skeptically. While Kuhn’s own views about scientific revolutions were also subjected to criticism, nonetheless left standing by the 1970s was a new emphasis on descriptive, historical accounts of science. It was on this basis that the sociology of scientific knowledge literature developed as the first genuinely post-Kuhnian approach. The literature on the sociology of scientific knowledge has gone through a number of stages, but rather than attempt to describe these in any detail (although cf., Hands, 1997; 2001, pp. 175ff.) I rely here on a brief account of the original Edinburgh Strong Programme to show the fundamental change in direction that the sociology of scientific knowledge brings to the philosophy of science and economic methodology. In the first place, the Strong Programme “is concerned with knowledge, including scientific knowledge, purely as a natural phenom- enon” (Bloor, 1991 [1976], p. 5). By this, Bloor meant that we should study the causal processes underlying scientists’ adoption of their beliefs rather than focus upon the epistemological foundations of those beliefs, as was done in traditional philosophy of science. This involved examining scientists’ actual behavior and, in ECONOMIC METHODOLOGY SINCE KUHN 575 particular, explaining the role that social interests played in the determination of their beliefs. Thus scientific theories were not thought to be adopted on logical grounds but, rather, because historical social forces supported the emergence of certain ideas. For example, were a group of scientists to find that governmental resources existed for certain types of research, they would be expected to find it in their interest to pursue that type of research. This does not imply that the content of science reduces to underlying social forces, but it does imply that the direction of development in science cannot be separated from an understanding of the historical social forces operating upon scientists. But then, as with Kuhn, the question was not whether there were rules that prescribed how to do good science but, rather, what were the practices and social contexts of the scientists involved. Thus, in much of the later literature that followed a sociology of scientific know- ledge approach, detailed studies of scientists’ laboratories and practices were a dominant focus (e.g., Collins, 1992 [1985]). But very little in the way of prescript- ive comment on the nature of science in general could emerge from such studies. Indeed, even the emphasis on social interests became too blunt a tool for most research that used this approach, since to argue that social interests were always at play suggested that there might be an appropriate or good sort of methodo- logy based on the analysis of social forces that might be universally applied in the study of science. At the same time, the emphasis on detailed studies of scientists’ laboratories and practices helped to bring out that there was much that was not transcontextual about science. Harry Collins investigated scientists’ prac- tice of replicating experiments to establish the legitimacy of their results, and found that not only was replication not systematically practiced, but that the terms on which it was practiced varied considerably from one scientific context to another (Collins, 1992 [1985]). Thus even scientists themselves eschewed the sorts of legitimation procedures that might enable us to prescribe rules for good science. However, more recently there has been something of a backlash among sociologists of science against the strongest anti-prescriptivist views, leading to approaches that aim to be nonevaluative as a whole, while nonetheless ruling out views thought to be relativist and/or idealist. Thus, Collins has been charged with “methodological idealism” and with failing to adequately expose “the social element in knowledge,” because his methods “invite us to make unchecked suppositions about the scope and role of social factors without providing any controls on them” (Barnes, Bloor, and Henry, 1996, p. 15). His error, it is argued, is in proceeding “as if the natural world, and our experience of it, played no significant role in the production of knowledge” (ibid., p. 13). But what seems rather more reasonable is to say that there is a relative stability and autonomy to perception that helps to explain scientists’ success in achieving consensus when creating instruments for socially accepted methods of observing the world. Thus, in the attempt to understand science, it is not enough to simply describe what scientists do. “Good” description of science involves giving proper weight to the role that observation plays in science. It follows that “good” science may itself be identified according to the role that observation – even when socially determined – plays in the formation of scientific theories. 576 J. B. DAVIS This argument in turn raises the question of whether realism constitutes a necessary component of scientific methodology. To answer this question in the affirmative would reintroduce a prescriptivist methodological theme, in the form of the argument that science ought to be evaluated according to how well it represents what really exists. Sociologists of science known as “constructivists” argue along Collins’s lines that the real exerts no fundamental influence on sci- ence. They do not deny the existence of the material world “out there,” since it resists what we and scientists attempt to do, but they argue that “for these resistances to make sense, they have to be interpreted [and] [t]he very moment you interpret them, you enter the real of the social world” (Knorr-Cetina, in Callebaut, 1993, p. 185). Against this, however, it has also been argued that we are better off abandoning the representational idiom and employing a perform- ative image of science, in which scientists and nature are engaged in a “dialectic of resistance and accommodation” that combines an “interactively stabilized” human and material agency (Pickering, 1995). The idea here is that nature and the world play an even larger role in determining scientific thinking than is normally allowed by those who emphasize interpretation and the influence of the social world permit. This allows for a “pragmatic realism” that requires that we under- stand science in terms of how knowledge relates to the world, without employ- ing traditional realist notions such as correspondence. Again, description is moderated by views of what constitutes “good” science. Although these arguments have been made within the sociology of science community, they have not gone unnoticed by methodologists of economics. An example of the ambivalence over whether methodology should be descriptive or prescriptive from the latter is Daniel Hausman’s account of the methodology of economics (Reuten, 1996). To begin with, Hausman doubts that a science – perhaps especially economics – that operated according to the prescriptivist standards of “either Popper’s or Lakatos’ methodology could exist” (Hausman, 1992, p. 204). He then characterizes economic methodology as a deductive a priori method (in the tradition of J. S. Mill and Robbins), and states that “[a]lthough the methodological rules of the method a priori . . . cannot be defended, I shall nevertheless defend the existing practices of theory assessment among economists” (ibid., p. 206). That defense proceeds by recognizing questionable methodological commitments in neoclassical economics, followed by assertions such as that “economists have good reason to be committed to them,” they do “not do much harm,” and that there “is a good deal of truth to them” (ibid., p. 210). At the same time, however, he concludes that while proceeding on an a priori basis is “not unjustifiably dogmatic, there is a serious risk” to setting aside empirical anomalies and being unwilling to “consider alternatives” (ibid., pp. 210–11). We can understand these remarks in terms of the kind of methodological approach that Hausman recommends. By building on Mill’s a priori method, Hausman frees economics of the responsibility of directly addressing empir- ical evidence. This weakens the prescriptivist side of his account, since theory assessment acquires special force in connection with whether theories measure up to evidence, while at the same time giving weight to Hausman’s project of ECONOMIC METHODOLOGY SINCE KUHN 577 describing the current practice of economics. But in setting out a new methodo- logical framework in terms of an inexact deductive method, Hausman has still generated criteria for theory assessment. Thus he cannot but doubt at times that neoclassical economics lives up to the standards of “good” economics. In effect, his dilemma is not unlike that faced by those writing in the field of the sociology of science. When they emphasize how social practices among scientists influ- ence observation, they de-emphasize the empirical. But faced with the task of explaining what counts as reasonable sociology of science, they too reintroduce normative criteria for the sciences that they examine. I turn in the conclusion below to how these developments may be linked to the gap between economists and economic methodologists. 34.3 SHOULD ECONOMICS BE UNDERSTOOD AS A NATURAL OR A SOCIAL SCIENCE? Before Kuhn, this question was better phrased as follows: Do all sciences operate on the same general principles, or do the social sciences operate on distinct prin- ciples? In the 1950s, the majority view among philosophers of science and those interested in economic methodology was the former one. Logical empiricism, the dominant philosophy of science, required that all scientific theories – whatever their subject matter – be translatable into empirical observation language. Scien- tific inquiry was structured according to the hypothetico-deductive method: from a general hypothesis and statements of initial conditions, predictive statements were deduced that could then be evaluated according to empirical evidence. On this view, scientific theories that were successful did not actually explain but, rather, simply described our empirical observations. Scientific “explanation,” then, involved applying the covering law or deductive–nomological (D–N) model, in which a particular observed event was accounted for by being subsumed under a general law together with a set of initial conditions (Hempel and Oppenheim, 1948). This general conception emphasized the existence of scientific laws, fram- ing these laws in terms of empirical evidence such as could be collected through experimentation, and seeing science as an objective enterprise that was inde- pendent of the sort of socio-historical forces that operated in most other domains of human life. After Kuhn, these ideas came increasingly under attack. Although Popper’s Logik der Forschung originally appeared in 1934, its later English translation (Popper, 1959) gave timely weight to Kuhn’s claims that science did not develop inductively through the production of general laws. Popper pointed out that universal statements can never be confirmed by singular statements, but can always be falsified by them. Rather than aiming at producing laws, then, science aimed at producing reasonable conjectures that might stand up to potentially falsifying empirical tests. Consequently, science was, rather, a process of continu- ally abandoning general propositions that did not stand up to empirical evid- ence. One reason why this was important to economic methodologists was that, while economics seemed to include a variety of laws (Gresham’s Law, the law of 578 J. B. DAVIS demand, the law of diminishing returns, Okun’s Law, and so on), these laws were generally imprecise by comparison with, say, the laws of physics. Popper’s thinking was thus influential among economic methodologists who thought that he offered an approach that was more suitable to economics. But if economics as a social science was not nomological in the fashion of physics, did this not also imply that empirical evidence had a different status in economics? One of the most important ideas in Kuhn’s book was the claim that empirical observation is theory-laden. Logical empiricism assumed that experiment and observation were theoretically presuppositionless. Kuhn argued, however, that scientists “see” things in terms of the paradigms in which they operate, and that this meant that science needed to be explained in terms of sociological and his- torical perspectives that accounted for paradigm emergence and development. But this made the progress of natural science depend upon categories that were far removed from its traditional understanding – making a place, for example, for concepts such as interpretation, which have been argued to be the basis for treating social science as distinct from natural science (e.g., Winch, 1990 [1958]). Effectively, this also reversed the question with which we began this section. Rather than asking whether all sciences operate on the same general principles, or whether the social sciences operate on distinct principles, the question now became one of whether all sciences operate on the same general principles or whether the natural sciences operate on distinct principles. A number of new initiatives in economic methodology gave an affirmative answer to the first part of this new question and a negative answer to the second part, in the process rewriting the methodology of economics in the language of interpretation. Perhaps most radically, McCloskey argued that the methodology of economics was the practice of rhetoric and persuasion, thus substituting the classical idea of a skill or an art for the modernist, epistemological idea of science as a body of knowledge (McCloskey, 1985). McCloskey was influenced by the American philosopher Richard Rorty (Davis, 1990), whose neo-pragmatist philo- sophy critiqued another key idea behind the traditional logical empiricist view of science, namely that – philosophically speaking – science “mirrored” nature (Rorty, 1979). Rorty understood interpretation in reflexive terms as entering the hermeneutic circle, and viewed science and philosophy as an historical social practice that lacked certainty-producing epistemic foundations. This was con- sonant with parallel developments in the sociology of science (discussed above) that described how science emerged in concrete communities whose character reflected scientists’ interests. In the strongest versions of this latter literature, science was a socially constructed phenomenon that placed the real world behind a barrier to interpretation. Yet the momentum behind these developments did not go unresisted. Economic methodologists since Kuhn have also defended the view that economics should be seen as a science on the order of the natural sciences. Perhaps the strongest posi- tion of this sort has been advanced by the philosopher of economics Alexander Rosenberg, who has argued that economics cannot become a science until it abandons its long-standing reliance on “folk psychology” (Rosenberg, 1992). Folk ECONOMIC METHODOLOGY SINCE KUHN 579 psychology explains human action as the effect of our desires working in combina- tion with our beliefs, such as is involved in the analysis of utility maximization. More broadly, folk psychology involves the domain of intentionality or the subject of human mental life. For Rosenberg, mental entities have no place in scientific explanation, because they cannot be explained in the rigorous material terms that are characteristic of natural science. However, it ought not to be thought of as an insurmountable difficulty for economics to address this problem. Human mental states can be re-described and identified as neural or brain states – a view some- times called “eliminative materialism.” Were economists to abandon their charac- terization of mental states in intentional terms, and treat them behaviorally as brain states, economics might then function as a natural science, and perhaps qualify as a science. Another way of reconfiguring economics as a natural science – also related to the issue of how we understand the mind – derives from the work of Herbert Simon. Simon is best known for developing the idea that human rationality involves “satisficing” behavior and a bounded rationality in which individuals make use of limited information and a limited computational capacity. But Simon’s thinking about bounded rationality and human beings’ reliance on “rules of thumb” and “heuristics” in decision-making also served as the basis for his later thinking about how to model artificial intelligence systems (Simon, 1957 [1945]; cf., Sent, 1997). Computers, he demonstrated, could be designed to employ the same “rules of thumb” and “heuristics” as are generally observed in human decision-making. But this implied that if computers are information-processing systems, then the human mind could also be conceived as essentially an information-processing system. Simon’s original work in this regard long antedated Kuhn’s, but it gained new life with the rise of cognitive science in the 1980s. Cognitive science – the general view that human beings are information- processing systems – has found many applications, but in a philosophy of science context was instrumental to the development of a “naturalized” epistemology inspired by natural science (e.g., Goldman, 1986). The concept of “naturalized epistemology” originated in connection with Willard Van Orman Quine’s postwar philosophic views (e.g., Quine, 1969). Quine had earlier argued that scientific theories are “underdetermined” by the evidence offered in their support (Quine, 1980 [1951]). How, then, were we to explain the attachment to scientific views that we observe among scientists? Rather than asking ourselves traditional epistemological questions regarding how theories are justified, Quine proposed that we engage in a purely descriptive account of science, and – in order to carry out this descriptive inquiry – rely on what the sciences themselves offer. More specifically, the science of psychology – indeed, behaviorist psychology – provides a method of conducting this inquiry as well as a naturalized alternative for investigating how scientific views are adopted and maintained in general. Quine’s views have been quite influential in philosophy of science, but not as much so in economic methodology, most likely on account of his tendency toward behaviorist psychology. Nonethe- less, the basic idea that economic methodology might employ the principles 580 J. B. DAVIS of natural science is entertained by some economic methodologists in con- nection with accounts of economics that are couched in terms of evolutionary science. Evolutionary views in economics go back to Thorstein Veblen, and have recently enjoyed a considerable revival in connection with the idea that the economy undergoes a process of Darwinian evolution, although there is also considerable debate regarding whether biological metaphors are appropriate in economics (e.g., Nelson and Winter, 1982; Hodgson, 1993; Vromen, 2001). Relatedly, evolu- tionary epistemology relies on the idea a that science such as economics develops much like the process of Darwinian evolution. Although there is a large literature from this perspective in philosophy of science, much of it is not familiar to most economic methodologists. Nonetheless, some have considered this approach as a strategy of analysis via acquaintance with the later evolutionary views of Popper (1972), whose work in general has been of sustained interest to methodo- logists. Popper’s evolutionary epistemology follows from his earlier thinking about scientists practicing a falsificationist methodology. The basic idea is that, in a selection process that constantly puts up new theories in much the same way as new organisms appear, falsified propositions fail. Theories whose propositions withstand a falsificationist selection process are adapted to their environment. Although this notion of adaptation is like the neo-pragmatist ideas that Rorty advanced in his own critique of traditional epistemology, here of course the critique comes from natural science. The fact that related arguments regarding the process of theory change can come from two such very different sources is strong evidence that economic methodology is ambivalent regarding whether economics is a natural science or social science. Finally, there are methodological views that treat economics as a natural sci- ence, but make important concessions to concepts and ideas long associated with the interpretivist view of social science. One is the work of Harold Kincaid, which combines naturalism and holism (Kincaid, 1996). Kincaid believes that the standards of science are those that were historically established by natural sci- ence, but that many of the concepts and concerns of social science do not appear in natural science. For example, Karl Marx is said to have operated with a natural science methodology, especially in developing what has since become known as Marxian sociology, but his concerns with the relationships between classes and individuals required formulation of new conceptual tools. Holism is a macrosociological notion that helps in this regard, particularly when confronting theory strategies such as reductionism and supervenience. Reductionism is the idea that properties or descriptions in one domain of theory can be fully trans- lated into or reduced to properties or descriptions in another domain of theory; in contrast, supervenience concerns how one domain of properties or descrip- tions is irreducible to, or supervenes on, a second domain. Interestingly, the concept of supervenience can be meaningfully applied to disputes in natural science, such as arise in connection with understanding the relationship between biology and chemistry. This all gives further evidence of the pluralistic views in economic methodology over whether the field’s roots are in natural or social science. [...]... that the history of economics had traditionally been pursued as a history of economic ideas, and this seemed to be at the expense of a fuller historical examination of the conditions and circumstances under which ideas are produced and maintained Moreover, the history of ideas approach was more the sort of thing that economists would approve of (were they to pay attention to the field), while the history. .. between economics and other sciences There were many responses to Schabas, most of which granted that it would be desirable for historians of economics to pay additional attention to the practices and methods of the history and philosophy of science But many of even Schabas’s more sympathetic respondents thought that historians and methodologists of economics ought to retain close ties to economics (even... Essentially this same issue was debated by historians of economics in connection with an examination of their relation to economics in a symposium published in History of Political Economy (Weintraub, 1992) The debate was revisited in a subsequent volume on the “Future of the history of economics” (Weintraub, 2002) Since economic methodology arose as a sub-field of the history of economics, and since... with the practice of the history of economics is itself one form of economic methodology (historiography), it will be helpful to review this debate in connection with the question of how the field of methodology should relate to economics as a whole The original debate was initiated by Margaret Schabas, who looked at the matter from the point of view of an historian of science (Schabas, 1992) Schabas... general ignorance of developments in the field? Let me suggest that although most economists could say very little about the content of recent methodological thinking, many are nonetheless aware of Kuhn and his idea that there may be such a thing as a “scientific revolution”; are aware that, since Kuhn, economic methodology has become an active domain of specialization; and are aware that, by and large, economic. .. reciprocated), and not just for practical reasons such as maintaining their 582 J B DAVIS employment One argument recalled the debate here over whether economics ought to be seen as a natural or social science If the history and philosophy of science were dominated by a concern with natural science, and yet economics was better thought of as a social science, then closer contact with economics was to be... state of affairs reverberated back upon methodologists? On the one hand, the separation between economics and methodology has helped to make the former an object of the latter As economists distance themselves from the field of economic methodology, this reinforces the relative autonomy of the field, thus giving methodologists confidence that their investigations are of value On the other hand, this motivated... economics and the economy My model for the latter is one of the most influential contributions to economics in the postwar period, namely Robert Lucas’s famous critique of empirical macroeconomics (Lucas, 1976) The Lucas critique was based on the idea that when people in the economy make choices, they take into account government policy regarding the economy and the theory on which it is based The large... economy and its theory No longer could one assume that the economy was an object of investigation that was separate from the investigator; rather, the object being investigated was now influenced by the act of investigation, and this in turn influenced the nature of the investigation Of course, the degree to which, and the manner in which, reflexivity operates in connection with the macroeconomy has since been... epistemological evaluation of economics pursued from an elevated philosophic vantage point external to economics, while the second involves descriptive examination of all the concrete methods by which economists seek to persuade one another of their arguments To carry out the latter sort of investigation, one needs to have close contact with the practice of economics Not surprisingly, then, at first economists . (Schabas, 1992). Schabas observed that the history of economics had traditionally been pursued as a his- tory of economic ideas, and this seemed to be at the expense of a fuller historical examination. treat them behaviorally as brain states, economics might then function as a natural science, and perhaps qualify as a science. Another way of reconfiguring economics as a natural science – also. of methodo- logy should relate to economics as a whole. The original debate was initiated by Margaret Schabas, who looked at the matter from the point of view of an historian of science (Schabas,