Changing Aims in Economics.

To the non-specialist, economic methodology is a confusing field. It is full of unfamiliar terms - hermeneutics, praxeology, sophisticated falsificationism - to list only a few. It examines issues that seem irrelevant or simply settled to practicing economists: the philosophical and intellectual foundations of the discipline; the epistemological status of economic "knowledge;" the scope and limits of economic inquiry; the propriety of various methods, approaches and techniques; the aptness of alternative units of analysis; the roles of subjective judgments, intuition and values; alternative criteria for appraising the validity of hypotheses; the role of theory; and professional norms for policy advising. Accordingly, before reviewing Hutchison's latest contribution, it is worth our while to briefly summarize - I do not pretend that this is the whole story - the evolution and current state of methodological thought in economics.

Early classical political economists - Cantillon, Petty and Adam Smith, for example - followed a rough and ready approach that mixed induction and deduction, analysis and synthesis, and in which empirical evidence was largely historical or anecdotal, and typically presented in literary form. The later classicals - e.g., Ricardo, Senior, James and John Mill - adopted Utilitarianism as their philosophic foundation. Believing the whole equals the sum of the parts (i.e., in "atomism"), these later classicals were able to introduce higher levels of abstraction, and increasingly more deductive reasoning into their investigations. Moreover, Utilitarianism provided an explicit criteria for guiding policy: the greatest good of the greatest number. The effects of alternative policy proposals, it was supposed, could be examined for their extent, propinquity, intensity, certainty, duration, etc. Then judgment could be passed as to which policy seemed superior. Note that the appeal was not to experience - how could one have experience with policies not yet implemented - but to reason and judgment.

Classical political economy and its Utilitarian foundations dominated economic thought in England but were challenged on the continent by German Historicism - grounded upon the Idealistic philosophies of Kant and Hegel - and, more importantly for our purposes, Positivism - founded by Auguste Comte.

The original positivism of Comte was purely empirical. Comte maintained that correlations among observations are all that we can "know." Causation and explanation are metaphysical illusions; theories mere metaphysical speculations. It was as though we can only "know" what, where and when; but never how or why.

Critics of Comte's vision - among them John Stuart Mill - were quick to point out that in practice it is impossible to make all observations, and that some insight, hunch, intuition or theory - call what one may - would have to guide the selection of what observations one should make. Furthermore, observations and correlations would have to be "public," that is available for inter-subjective verification. Finally, the human desire to "understand" demands explanation in the sense of stating logical connections among observations that render them systematic, dependable and orderly. Nevertheless, establishing validity would henceforth require an appeal to experience, not to "values," intuition or judgment. Indeed, the goal of the Positivists was to make knowledge universal, value free (i.e., wertfrei), objective and not contingent upon the temporal, spatial or cultural situation of the observer.

In time, Positivism evolved, in response to these and other criticisms, into what was called the Logical Positivism of the Vienna Circle (of philosophers and scientists: e.g., Carnap, Schlick and Wittgenstein). Theories and hypotheses - derived from whatever source - were to be "tested" against data. The more quantifiable the data, the better. (After all, what could be more transpersonal and less biased than mathematics and statistics?) Verification against empirical data became the criteria for warranting that knowledge is "scientific."

The doctrine of verification drew opposition from Karl Popper. Seeking to verify theories, Popper argued, readily degenerates into a pursuit of evidence that will confirm them. With a little selection and imagination, evidence could be made to confirm virtually any hypothesis. Confirmationism might then validate theories or hypotheses that were in competition with one another or even mutually exclusive of one another. The better, the truly scientific, strategy is to seek to falsify theories and hypotheses. Any statement that cannot be (even conceivably) falsified was demarcated as lying outside the realm of science. Statements that remain unfalsified, after passing through many stringent tests, are to be tentatively accepted as scientific knowledge. Therefore, an appeal to intersubjective experience remained the criteria for warranting that knowledge is "scientifically" valid.

Popper's falsificationist reformation of positivism served - in one form or another, for example, the Operationalism of Samuelson or the Instrumentalism of Friedman - as the philosophic foundation of main-stream (i.e., neoclassical and neo-Keynesian) economics for a generation or so. Most practicing economists still think of themselves - when in a self-conscious frame of mind - as some species of reformed positivists. Ultimately the test of validity in economics is corroboration by, or conformity to, the facts: just as it is in the "hard sciences" of chemistry or physics.

Yet by the 1960s, doubts were arising among philosophers and economic methodologists about this self-satisfying image of practice among economists. For one thing, much of the more prestigious work in academic economics seems to have little or no empirical reference at all. Highly abstract theories, not meant to be tested against facts are granted special dispensations. Such work is appraised according to non-empirical criteria, such as logical consistency or mathematical elegance.

But more fundamental was an assault from philosophers, sociologists and historians of science who questioned the positivist story, no matter how reformed, on both historical and logical grounds. The likes of Kuhn, Lakatos, Polanyi, Toulmin, Laudan, and Feyerabend first debunked, as an historical matter, the myth that even the premier sciences - physics and chemistry - had evolved through conjectures and refutations. Similarly, these and other philosophers of science contested even the logical possibility of falsification. Just what was to be falsified: initial conditions, the quality of the data, auxiliary hypotheses, whole systems of theories? Are there any uniformly reliable methods, approaches or criteria: any certainties at all upon which a science might be built? Or does "anything go?" Is economics simply what "economists do?" If so, just who are, and who are not, "economists?" Is anybody's economics demonstrably better than anyone else's?

At present, the discipline of economics is without generally accepted answers to questions such as these. Ours is a fragmented enterprise that lacks a firm philosophical foundation. It is into this chaos that Hutchison injects Changing Aims. The volume presents an expanded version of his Hennipman Lecture, delivered in Amsterdam in 1990, in which Hutchison expounds four closely related themes. First, from the time of Sir William Petty to that of Maynard Keynes, the dominant aim of political economy and economics was to guide policy-making, public and private. Successive generations of economists sought to base these policies on "more reliable, intellectually disciplined foundations or methods, with the objective of achieving less unsuccessful and less sectionally biased policies" [p. 1].

Second, shortly after the second world war, a shift in the primary objective of the discipline gradually gathered momentum. Ever more "academic" values and criteria began to supplant the traditional goals of policy relevance and enlightenment. Increasingly, extensive efforts were expended on analyzing counter-realistic "models," by means of highly abstract, deductive techniques that are based upon minimal empirical foundations, or upon no empirical foundations at all. The aim of such work is more to demonstrate virtuosity in esoteric mathematics than to derive results relevant to policy making in the real world. Academic economics, at least, is "increasingly concerned with technique to the exclusion of the subject matter itself" (according to G.N.D. Worswick, as quoted on p. 21). One result of this shift in aims is the growing irrelevance of much university education in economics for applied work in government and business (where the overwhelming majority of professional economists actually ply their trade [pp. 20, 117]). Hutchison labels this new regime of excessive and arbitrary abstractions "formalism."

He argues that formalism is the creation of a minority of economists,(1) mainly those academics teaching in doctoral programs. (However, here he badly underestimates the allure of formalism for academic economists interred at undergraduate institutions. Their relentless and uncritical urge to ape the methods, topics and reward systems of colleagues located at more prestigious institutions is a major force sustaining the inertia of this shift in disciplinary objectives.)

Hutchison documents some results of this transformation in aims on the attitudes of the next generation of economists by citing [pp. 24-28] the 1990 survey of 212 graduate students in economics at six prestigious American universities by Colander and Klamer. Most of those students considered economics a "game" in which depth of knowledge has a low pay-off. More than two thirds believed that "a thorough knowledge of the economy was unimportant," while two-fifths considered a knowledge of the literature to be unimportant. Hutchison considers these students quite rational. What leads to success in graduate school and in academic careers is facility with techniques. Mastering a technique that can be applied to numerous areas will yield numerous articles. Mastering the facts about a specific area of the real economy may lead to only one article. Displaying mathematical or technical virtuosity is more important than policy relevance. As one student in the survey put it, "policy is sort of for simpletons" [p. 28]. In any event, real world knowledge is a hindrance, because such knowledge makes it harder to accept assumptions that render problems mathematically tractable. Models are chosen according to how likely they are to yield a publishable article, and not on the basis of how illuminating they are.

Hutchison's third theme is that the shift toward the objectives of formalism was accompanied by a diminishing consensus concerning the methodological foundations and principles of the discipline. For one thing, discussions by "methodologists" add more to obscurity and confusion than they do to clarity. As an example, Hutchison points to the ambiguities surrounding the term "theory." Among economists this term has come to cover deductive analysis, tautologies, theorems and taxonomies as well as empirically testable propositions. Therefore it is not clear what sort of understanding a knowledge of "theory" is supposed to yield: of real world relationships, processes and behavior, or of abstract processes within purely hypothetical models. Confusion over what sort of understanding we have has led to disastrous policy advising in the past, as for example, when Austrian economists successfully dissuaded European governments from responding vigorously to the inflations, deflations and unemployment of the inter-war period.

Similarly, the banal observation that some degree of abstraction is inevitable in all scientific investigation of complex processes, has been tacitly perverted into the maxim that virtually any degree of abstraction is justifiable. Reference to the well established method of "successive approximations" is vacuous when no satisfactory second approximations are forthcoming. The smoke and mirrors of methodological ambiguity, special pleading and incoherence are covering the withdrawal from grappling with messy reality.

Today's methodological discussions also ratify the fragmentation of the discipline. Fundamentally contradictory views are being inculcated into the students of different schools of economists, rendering their training a process more reminiscent of indoctrination than of education. Appeals for methodological consensus are met with complaints against "intolerance" and with pleading for "pluralism" or assertions that "anything goes." The ultrapermissiveness demanded by the student radicals of the 1960s has made its way into the methodologies of economists.

A common theme among the complainers,(2) pleaders and asserters is the demise of "modernism," and its epistemological and methodological counter-part, "positivism." As Hutchison [pp. 49-50] aptly puts it:

"Positivism" is employed as a vast, fuzzily defined, or undefined all purpose Aunt Sally, or target of abuse, possessing the essential hold-all characteristic that, on the one hand, all sorts of writers and writings can be swept into it; while, on the other, every kind of error or evil, from capitalistic exploitation to Auschwitz, can - with calculated vagueness - be attributed to its alleged supporters.(3)

He illustrates his case with reference to several current writers on economic methodology: Caldwell, Hollis and Nell, and New Conversationalists such as McCloskey and Klamer.

Caldwell's resurrection of the a priorism or praxeology of Mises is little more than a defense of "'apodictic certainties' ... against intolerant 'positivist' criticisms" [p. 50]. Furthermore, the meaning of Caldwell's call for "methodological pluralism" is obscure, but seems to reject, in principle, choices among competing economic theories.(4)

Hollis and Nell come from the opposite end of the political spectrum in that they wish to substitute the certainties of Marx - instead of Mises - for the empirically based observations of "positivistic," neoclassical economics. In view of the unrelenting faith in their dogma that Marxists at my own university retain, even in the face of the utter collapse of the Soviet and East European economies and governments, I must agree with Hutchison's conclusion [p. 53] that:

Certainly, neo-classical theory suffers from great deficiencies and inadequacies, but these pale into insignificance beside the proven, fundamental monstrosities of Marxian economics in practice, commended to us by Hollis and Nell for its anti "positivist" methodological foundations.

On the other hand, New Conversationalists, such as McCloskey, manage to attribute to "positivism" such diverse evils as the Vietnam War, male chauvinism, McCarthyism and the Holocaust,(5) New Conversationalists (and other "methodologists" for that matter) manage to intimidate economists, who, because of their overly narrow educations, are susceptible to glib references to terms and namedropping from such fields as literary and artistic criticism or philosophy. (In response, I am moved to suggest that we modify Joan Robinson's aphorism a bit: "The best reason to study philosophy - or literary criticism, artistic criticism, etc. - is to avoid being fooled by dilettantes, name-droppers and term-borrowers who selectively mine those disciplines for their own persuasive purposes.")(6)

One particular target of all anti-positivists is the venerable "positive-normative" distinction. Maintaining that this distinction remains valid and useful does not entail a claim that any particular kind of economics is wertfrei, nor does it prohibit statements that are ethically based. Rather it is to attempt, as best we humans can, to honestly clarify what kinds of statements we are making. Similarly, it is worth our effort to try to distinguish between definitions or tautological statements in economics, on the one hand, and empirical statements, on the other hand.

The fourth theme that Hutchison addresses is the$importance of prediction or forecasting to the non-academic and non-formalist economist. Hutchison argues that for most "business" and "government" economists, forecasting or predicting is their primary raison d'etre. To be useful, these forecasts need not be as quantitatively precise as those of physicists, chemists or astronomers. Nor need they be arrived at through inductive processes. For example, economists may predict that attempts to "fix" the price of gasoline at other than its equilibrium level will lead to queues and to black markets, or to surpluses and storage costs, as the case may be. Such a prediction is useful, and may foster less unsuccessful policy-making or at least fewer surprises.

Prediction in every day human affairs clearly is possible, otherwise we humans simply could not interact as social beings. What is required of predictions by professional economists is that they are, on the average, less inaccurate and unreliable than those "everyone has to make about other people's actions and reactions" [p. 79]. Whether the predictions of economists are more or less inaccurate than those of other disciplines is beside the point. As long as our predictions are "significantly less inaccurate and unreliable than [those that] would be forthcoming without their input of systematic, more or less disciplined economic knowledge" [p. 81], economists have an important role in fostering less unsuccessful policy-making, in both the public and the private sectors.(7)

Useful predictions for real-world policy-making much more often provide the basis for damage limitation or catastrophe mitigation than for optima and maxima [p. 82].

For just as it is important to avoid exaggerated claims of prescience, it is important to avoid the nihilistic contretemps that we cannot usefully predict at all. (See Hutchison's pp. 37-38, 68-9, 82, 86-87, 77-81, 97-98, 157-58, 160-62, 165-66, for detailed discussions of these matters.)

I find myself in substantial agreement with most of what Hutchison has to say here. Economics, in my view, should be a search for fruit - in the form of less unsuccessful policy formation - as well as for light: i.e., knowledge for its own sake. Given this primary aim of ameliorating the human condition, certain methodological principles follow. We cannot hope to alter "what is" into "what we desire to be" merely by wishful thinking; we must begin by knowing "what is" and, in some meaningful sense, how "what is" got that way. This requires, in addition to a knowledge of theory, and mathematical and statistical techniques, an intimate knowledge of messy institutional and historical facts. Policy advising, either public or private, places a premium upon prediction and forecasting, though such activities must be practiced with self-skepticism and their results reported with humility and honest tentativeness.

Nevertheless, one glaring deficiency mars this otherwise excellent work. With justice, Hutchison decries the failure of those who attack positivism to clearly define and delimit just what they mean by that term. Yet, Hutchison himself fails to clarify what he means by "positivism." (At a minimum he could point us to writings where positivism is adequately described.)(8)

Still, some elements of his vision of positivism may be inferred. Positivistic economics rests upon empirically testable propositions. Correspondence to evidence is what distinguishes valid from invalid speculations. The appeal to (inter-subjectively verifiable or falsifiable)(9) experience is intended to protect intellectual debate and policy formation from dogmaticism. Positivistic methods may incorporate mathematics and statistics, quantitative and qualitative data, institutional and historical facts, and inductive as well as deductive reasoning. Attempts to maintain the admittedly tenuous positive-normative dichotomy are a form of self-discipline, intended to remind us to critically examine our statements for bias and to clarify just what sort of statements we are making. The aim of scientific economics is to foster less unsuccessful private and public policy-making. Pareto is identified (on p. 109) as "the economist to whom the much-misused term 'positivist' might least invalidly and misleadingly be applied."(10)

Despite this deficiency, I heartily recommend this work to the practitioners of our discipline. It persuasively informs government and business economists that they have a useful contribution to make toward ameliorating the human condition. It places in perspective the vague, nihilistic and often career-serving drivel of self-styled "methodologists" and "New Conversationalists," especially now that many of these specialists in economic methodology do not themselves "do" the economics that they presume to analyze and criticize.

William Guthrie Appalachian State University 

1. Hutchison recognizes the difficulties involved in defining precisely who is a professional economist [pp. 67-70, 156-57]. Nevertheless, within the United States no more than 20 percent of economists are "academics," and of these only a tiny fraction teach in the graduate programs of the more prestigious universities. Nevertheless, this tiny fraction has successfully imposed its own narrow, academic outlook on the education of all professional economists.

2. Hutchison misses yet another lactic employed by today's "anti-positivists," the whining of self-identified "victims," who claim, without substantiation, that they have been abused by intolerance.

3. The reference to Auschwitz is found in McCloskey's "Why I Am No Longer a Positivist," Review of Social Economy, 47(3), Fall 1989, 225-38.

4. Hutchison emphatically advocates pluralism in the sense of tolerance for "historical, institutional, or mathematical methods or approaches" [p. 50]. His particular distaste for the implications of Austrian abstract-a prioristic-deductive theory for policy arises from its disastrous advice that governments and central banks ought not to intervene during the inter-war inflations and deflations [p. 75].

5. While this may seem bizarre, a carefully reading of the cited article in note 3 has persuaded me that Hutchison's interpretation of it is valid, with the exception of his reference to Vietnam. (There McCloskey's reference was to McNamara's "whiz kids," who kept assuring us that every quantitative measure they had proved that we were winning the war. Those whiz kids knew how to count; they just didn't know what counted.) McCloskey's understanding of positivisms - he refers to several including, "early positivism," and "Chicago's version of positivism" - seems highly idiosyncratic. In particular, George Stigler is painted as a sort of positivistic devil, complete with pointed tail and horns!

6. The vague, glib and all-inclusive manner in which such terms as "modernism" and "post-modernism" are tossed about are cases in point.

7. Paradoxically, our capacity to predict may be improving, while our actual predictions become less successful. The forces of ever more rapid technological change, increasing global interdependence, increasing discretionary spending and continuing cultural change may make the structure of life less predictable and more unstable.

8. McCloskey at least, in the article previously cited, has told us what he means by "positivism." However, near the end of that article (on his pp. 236-37) he asks, how economics would be different without it [i.e., positivism]." It strikes me that his answer describes something not different in practice from what Hutchison seems to consider positivistic economics!

9. Hutchison does not discuss the issues of verification, confirmation, corroboration, naive falsification or sophisticated falsification that inevitably surround his appeals for empirical validity.

10. Examining Pareto's methodological approach is beyond the scope of this review. Readers are referred to Pareto's MethodologicaI Approach to Economics, Chapel Hill: University of North Carolina Press, 1968, by Vincent J. Tarascio.