The Decline of the “Original Institutional Economics” in the Post-World War II Period and the Perspectives of Today
This paper is closed for comments.
As is known, original, or “old”, institutional economics (OIE) ─ we also indicate it as “institutionalism” ─ played a relevant role in its first stage, and it can safely be said that it came to be, although perhaps by a slight margin, the “mainstream economics” of the time. That period ran approximately from the first important Thorstein Veblen’s contributions in 1898 — the article “Why Is Economics Not an Evolutionary Science?” — to the implementation of the New Deal in the early 1930s, where many institutionalists played a significant role.
However, notwithstanding its promising scientific and institutional affirmation, institutional economics underwent a period of marked decline, that spanned from, approximately, the mid-1930s to the late 1980s, when a new season for institutional economics was set in motion.
In order to cast some light, without any pretense of completeness, on this complex issue, we have organized the work as follows: in the first paragraph we consider the main interpretations of this phenomenon. Then, in the subsequent paragraphs we analyse a number of “endogenous” aspects which might have played a significant role in such decline: (i) the relations of institutional economics with Keynes’s macroeconomic theory; (ii) the links between theoretical and empirical analysis and the supposed lack of a clear theory; (iii) the interdisciplinary orientation.
Arturo Hermann is to be commended for his attempt to identify the main reason for the decline in disciplinary power and prestige of Institutional Economics, by which he means the Original Institutional Economics, or OIE, as I will call it in this essay. Hermann notes that there is a strong case to be made that OIE was the dominant approach among economists in the U.S. in the first three plus decades of the 20th century. Why then did OIE decline in importance in the years after WWII? Why did a deductivist, non-evolutionary theory that stands quite apart from the development of other social sciences come to dominance, while a more scientific, evolutionary, and interdisciplinary approach to study of the economy was marginalized in most of the academic world?
Hermann surveys the answers that have been given to this question. He quite correctly, in my view, identifies the following as among the most powerful answers: 1) the narrowness of approach into which Clarence Ayres and his followers forced the Veblenian branch of institutionalism; 2) a lack of agreement on “fundamentals” among OIE practitioners; 3) the success of Keynesian and neoclassical economists in their claim of being more truly scientific than the “naïve empiricist” Institutional economists.
The change in economic practice and the shift in prestige among the different practices, is a complex story as Hermann recognizes, but it is also one that he does not recount quite accurately. In this brief comment I will focus on only one of the explanations for OIE’s decline that Hermann discusses: the relationship between theoretical and empirical analysis and the failure of OI economists to articulate a satisfactory (to other economists) reconciliation of the two. On this point Hermann argues that a chief failure of the OI economists was that “their philosophical background oscillated between pragmatism and positivism, and was never clarified.” As Hermann explains it, to the extent that OI economists tended toward “positivism,” they considered statistical evidence to be the only truly scientific aspect of their work and relegated analyses of legislation and court decisions, as well as case studies to less than truly scientific status. Hermann then argues that “John Dewey’s notion of behaviourism,” which is, he says, “the pragmatist conception of behaviourism,” requires that the “experience of a person in its entirety” be analyzed.
The words “behaviourism,” “pragmatism,” and “scientific” are very tricky words and especially so when we in the early 21st century try to understand the arguments that engaged our intellectual ancestors in the late 19th and early 20th century. There were, in effect, several different discussions going on among pioneering sociologists, philosophers, emerging anthropologists, those would become psychologists and even a few economists, and each group was fighting against previously dominant ideas among their own ancestors. The ancestral groups did not entirely overlap, so that some of the ideas against which each separate group was fighting were of little or no importance to the other groups. And some of the words for the emerging thought that was held to be superior among all of the pioneering social scientists were the same even though with shades of different meaning depending upon the discussion in which they were used.
This confused and confusing intellectual history has, unfortunately, led Hermann astray. He is simply wrong in saying that a fundamental problem for OI economics in the 1930s and 40s was an unreconciled tension between “behaviourism” and “pragmatism.” When Hermann says that this was the case he is arguing that OI economists fell victim to a belief that only that behavior that could be “measured” through use of statistics could be considered “scientific” and therefore worthwhile for purposes of understanding economic processes, he is assuming that an intellectual fight in the discipline of Psychology had equal relevance in Economics. Now, it was indeed the case that the Institutional economists of the 1920s and 30s were enthusiastic about the use of statistics as a way of describing and understanding socioeconomic processes. They did not, however, regard legal and political analyses or case studies as somehow less valuable or legitimate. A quick review of the Institutionalist literature of period as is provided by Malcolm Rutherford in Chapter 2 of his 2011 book will show this to be the case. Valued OI work took many forms and as Rutherford wrote: “Being ‘scientific’ meant not being satisfied with speculative armchair theorizing based on unreal premises, including an outmoded psychology. . . . Being scientific did mean being investigative, exposing hypotheses to critical empirical examination, and bring economic thinking into line with recent developments in related disciplines such as psychology, sociology, anthropology, law, and philosophy” (Rutherford 2011, p. 26). The work cited by Rutherford as meeting this requirement will be seen to involve a wide variety of analytical approaches.
That OI economists did think and often did write that numbers might be the best tool under many circumstances for understanding the experience of people in the U.S. economy can be illustrated by the work of Morris Copeland on money flows. (See Copeland 1952, Dawson 1996, Part II.) Copeland (following the lead, as he said, of J.R. Commons) recognized that transactions that were completed by money flows were basic provisioning actions in our modern economy and he set out to trace those money flows. As he did so he was not unaware of the full range of experiences that accompanied these transactions (Copeland, 1946). But, in an “accounting economy” dealing with accounts is a good way to understand what is happening. Here is how Copeland put it, “Because money plays a major role in organizing our economy, it has frequently been characterized as a money economy. In a significant sense it is also an accounting economy. Our system of moneyflows has become so complex that many transactors must keep detailed accounting records. Such records, and financial statements and reports derived from the, today help significantly to organize economic activity “(p. 8).
Copeland was not saying that behavior must be quantified for analysis to be scientific. He was saying two things that set his work, and that of his mentor, Wesley C. Mitchell, apart from earlier work in Economics. First that the 20th century U.S. economy was largely organized by money so money flows were important although largely ignored in earlier economic analysis. And, he was saying that we should use the records that the subjects of enquiry used in order to understand what they were doing. In stating the importance of his approach, Copeland was arguing with the classical and neoclassical ancestors of his chosen discipline.
So too was Wesley Mitchell arguing with economists when he wrote the passage that Hermann quotes from Mitchell’s Presidential Address to the American Economic Association. What Mitchell was arguing was not that quantitative data was required for scientific exploration; he was quite specifically arguing that Alfred Marshall had conceived of “economic behavior as controlled by two opposing sets of motives, the motives that impel us toward consumption and the motives that repel us from labor and waiting” (Mitchell, 1925, p. 25.) Mitchell went on to say that he doubted that it would be possible to make more precise statements about these motives, just as it doubted that it was going to be impossible to measure “pleasure” or “the strength of desire.” He went on to say that use of these terms “are something that the theorist adds to the data” and that “In the present state of knowledge of human nature, such interpretations smack more of metaphysics than of science” (1925, p. 25).
I can see why Hermann equates this position with the kind of positivism of mechanistic behaviorism advocated by J.B. Watson, but it is nonetheless a mistake to do so. Watson had rejected the “functional psychology” that was part of the social science context in which Veblen, Dewey, Mitchell, Copeland and other OI economists were developing their ideas and arguments. (For more on this see Rucker 1969, Chapter 3 entitled “Psychology: Functionalism and Behaviorism.”) It may also be tempting to equate what Mitchell wrote in his Presidential address with later “positivistic” forms of argument about the irrelevance of assumptions so long as predictions are shown to be correct. But that too would be a mistake. Mitchell sought realism in assumptions, in evidence, and in conclusions.
So, contrary to what Hermann argues, my reading of literature does not lead me to the conclusion that the Institutional economists fell victim to a “positivist attitude” by considering statistical analysis to be the only truly scientific form of evidence. The debate over the use of “measurable” observations was of greater significance in some other disciplines and, arguably, in some other forms of economic analysis where the important variables did not emerge in numerical form from the human actions deed important. Neither Mitchell nor Copeland had to, nor did they want to, “impute” values; they wanted to use the values recorded by the subjects of their studies.
Nott only do I think Hermann wrong in thinking that OIE was weakened by adoption of a “positivist attitude,” I also think that he is wrong in saying that this alleged adoption created a conflict with a ”pragmatic conception of behaviourism” in OIE. Behaviourism as an analytical approach was, as already noted, important in the development of modern Psychology but actually had very little importance in the development of OIE. However, Pragmatism was and continues to be an important foundation for OIE. What the Pragmatic tradition, as established by C.S. Pierce, W. W. James, J. Dewey and others, means is that individuals and the groups of people who are the basic unit of OI analysis are thought to adapt “means to ends that cumulatively change as the [life] process goes on” (Veblen, 1906, 74-75). Pragmatism is not, as is often thought, simply a matter of expediency, nor is it in any sense a form of analysis that sets it apart from “positivism.” Rather it a fundamental assumption that the ends or objectives of human action change as the means of achieving them change. This process is cumulative and interactive. The importance of this for OI economists was that it meant there was no one thing—such as maximization of utility—that could be both taken as an unchanging goal and be given any specific meaning. The acceptance of the Pragmatists’ view of humans and their evolution through time and space also meant that individuals took and changed their tastes and preferences as part of an ongoing social process that was shaped by the means of achieving what was desired/needed.
This proposition represented a fundamental challenge to economic thought. The introduction of active human agents who were changed by their environments even as they changed those environments could not be reconciled with the “globules of desire” that, in the famous Veblen passage, “oscillate “under the impulse of stimuli that shift about the area, but leave him intact” (Veblen, 1898, p. 389.) The problem was that the “principles of economics” that Alfred Marshall had so meticulously set forth were among those aspects of 19th century thought that were hard to test using newly available statistics and were hard to reconcile with the economy as the OI economists were describing it. Mitchell thought, as he said in his Presidential Address, that economists would change their questions and would not simply try to prove the truth of those principles. But that is not how it went down.
Ragnar Frisch, whom Hermann quotes to illustrate the weakness of the neoclassical rejoinder to Mitchell’s Presidential Address, gave the most direct statement about being unwilling to abandon the “principles” that were inherited and were held to be both true and of greatest importance for economic analysis. The background is this: Henry Schultz, an agricultural economist who, as the US Department of Agriculture began collecting data on prices and purchases of goods such as sugar, realized that the observed price and purchase combinations, if translated into Marshallian terms, were points of intersection between a theoretical and unknown demand curve with a theoretical and unknown supply curve. In order to construct a demand schedule you had to hold the supply schedule fixed or vice versa. The question was how to adjust for the many other things that, as textbooks briefly remind students in the beginning, must be held constant for the common demand-supply curves to look as they do.
It was at this point that Ragnar Frisch and his co-author Frederick Waugh in effect created modern econometric protocol and set neo-classical economics on its modern and somewhat awkward path. (For much more on this see J. Morgan, 2016 and the essays contained in that volume.) In a manner almost identical to the example that Hermann provides of how the truth is known about the tides, Frisch and Waugh declared that an empirically determined relation is “true” if it approximates fairly well a certain well-defined theoretical relationship, assumed to represent the nature of the phenomenon studied. There does not seem to be any other way of giving a meaning to the expression “a true relationship.” (Frisch and Waugh, 1933:38).
Marshall’s supply and demand schedules were thus declared that to which statistical evidence had to bend.
The declaration of Frisch and Waugh should be taken as much more than a solution to a technical problem for statisticians. It was also a statement that Marshall’s principles should be taken as definitive. By itself this would not have been sufficient to lead to the decline of OIE, an approach that involved a fundamental rejection of Marshallian economics as set forth in his Principles. Many other things were involved as well.
My view is that OIE declined in part because of the support and the prestige that Frisch and Waugh and their colleagues brought to the new field of econometrics and to their adoption of Marshall’s principles as fundamentally true. I have also written elsewhere about the narrowness of approach adopted by many followers of Ayres as a major reason for the decline of OIE. (See Mayhew 2008) And, as Hermann notes in his final remarks the failure of OIE to offer deterministic results and its gradualist approach to socioeconomic change have contributed to that failure. So too has the reformist role that Keynesianism (of an earlier sort) and then Post-Keynesianism have offered to academics eager to help create better economies. The story of the decline, and indeed, the failure of reformist and gradualist social science in the post-WWII era, remains to be fully told. I am pretty sure that when it is told and told well that the failure of OIE will not be explained as Hermann has in the essay under review.
Finally, I will note that one of the continuing strengths of OIE, a strength not usually so regarded in the mainstream of economics, is the continuing interdisciplinarity of work done by OI economists. Hermann, takes the importance of Psychology to the 1920s OI economists to indicate that that is still the field of greatest interdisciplinary importance. Not so. Having abandoned the notion of a narrowly rational economic man/person, OI economists have moved on and found opportunities to work with legal scholars, sociologists, anthropologists, political scientists and many others. Psychology is where the action was back in the first two or three decades of the 20th century. For those interested in economic issues and policies that is no longer the case but OI economists have not, as a result, turned inward. That interdisciplinarity flourishes can be confirmed by looking at the affiliations of those who have published in the leading OIE journal, The Journal of Economic Issues, over the last several decades.
Copeland, M.A. (1952). A Study of Moneyflows in the United States. New York: National Bureau of Economic Research, Inc.
Dawson, J.C., editor (1996). Flow of Funds Analysis: A Handbook for Practitioners. Armonk, NY and London: M.E. Sharpe.
Mayhew, A. (2008). Clarence Ayres, technology, pragmatism and progress. Cambridge Journal of Economics 34:1:213-222.
Mitchell, W.C. (1925). Quantitative Analysis in Economic Theory. Presidential address delivered at the 37th Annual Meeting of the American Economic Association, Dec. 1924 and reprinted in Mitchell, W.C. (1950) The Backward Art of Spending Money. New York: Augustus M. Kelley, Inc.
Rucker, Darnell. (1969). The Chicago Pragmatists. Minneapolis: The University of Minnesota Press.
Rutherford, Malcolm. (2011. The Institutionalist Movement in American Economics, 1918-1947: Science and Social Control. New York: Cambridge University Press.
Veblen, T.B. (1898). Why is Economics Not An Evolutionary Science? The Quarterly Journal of Economics.
In providing a first answer to Anne Mayhew’s comments, I would say that I agree with some of her remarks. For other aspects, I think that my ideas have been misunderstood, so I hope that conveying them in a more condensed way would help clarify these complex issues.
I agree that (i) the terms “behaviourism,” “pragmatism”, ”positivism” embody different meanings; (ii) there are different strands within the original institutional economics (OIE); (iii) and that, in this respect, “qualitative analysis” received pioneering contributions by early institutionalists (including Veblen and Commons), and that Wesley Clair Mitchell was a profound and holistic thinker. However, it is also true that OIE, as being at that time a very young discipline, did not formulate a fully fledged analysis of what should be considered “scientific evidence” in economics and other social sciences.
In this sense, it seems no surprise that “their [of the early institutionalists] philosophical background oscillated between pragmatism (intended as a holistic and humanistic conception of persons in their individual and collective unfolding) and positivism (intended as a conception that considers “scientific” only measurable phenomena), and was never sufficiently clarified”.
This does not mean, however, that there was among early institutionalists a clear-cut and irreconcilable conflict between the two concepts or that these scholars openly adhered to some positivistic doctrine. Quite the contrary, that period sees, as I have tried to point out, the presence of a variety of (continually evolving) positions among institutionalists, with hardly none of them being identifiable as utterly “positivist” or “pragmatist”. This came about also because pragmatism too was a newly born discipline and was articulated in different strands. In this respect, I should clarify better that I am chiefly referring to the John Dewey’s and William James’s conceptions of pragmatism, which adequately consider the holistic nature of the persons in their individual and collective expressions.
The same discourse applies to the fields of sociology and social psychology and to the complex bonds that these disciplines established with the OIE (we can mention, for instance, the interdisciplinary and pluralistic atmosphere that characterized the University of Chicago and several other Universities in the early years of institutionalism). These aspects are treated, among others, by Rutherford, M. (2011), The Institutional Movement in American Economics, 1918-1947, Cambridge (Engl. and Mass.): Cambridge University Press.
All that said, it seems fairly evident that one of the key issue at stake was that of identifying the meaning of “scientific evidence” in economics and other social sciences. And that, in this respect, there is a marked difference between the positivist and pragmatist approach to the study of economics. In relation to the points raised by Mayhew, what I have simply tried to put forward is that:
(I) A distinctive trait of the OIE is a holistic conception of the persons in their individual and collective expressions.
(II) This requires an interdisciplinary approach in order to better analyse the multifarious ties between persons and their economic, social and cultural contexts.
(III) For this reason, the OIE approach strongly demands a conception of “scientific evidence” that ─ by going beyond the narrow positivistic claim that only “measurable phenomena” can have “scientific validation” ─ would fully consider the whole set of qualitative and unmeasurable phenomena.
(IV) There were between early institutionalist economists various (and evolving) opinions on the “pragmatist” and “positivist” conception of economics; however, these opinions were most often implicit and unfledged ─ also because they tended to be intermingled with other issues ─ and hence did not steer a thorough and specific debate on these aspects.
(V) The most clearly expounded ideas on the subject were Mitchell’s articles containing his Presidential Address to the American Economic Association and the subsequent appendix-rejoinder: “Quantitative Analysis in Economic Theory”, American Economic Review, 15 (March): 1-12, 1925, and “The Present Status and Future Prospects of Quantitative Economics”, American Economic Review Supplement, 18 (March): 39-41, 1928.
(VI) On that account, his position is wonderfully clear. He remarks that qualitative aspects are relevant for economics but, in order to render their analysis amenable to scientific treatment, it becomes necessary (or at least highly advisable) to link qualitative aspects to quantitative proxies. This approach is expressed in the following passages (I refrain from reporting other significant quotations in this respect) of his 1928 article, “Qualitative work itself will gain in power, scope, and interest as we make use of wider, more accurate, and more reliable measurements….quantitative work cannot dispense with distinctions of quality….with more abundant and more reliable data, more powerful methods, and more liberal assistance, the men now entering upon careers of research may go far toward establishing economics as a quantitative science.”, (Mitchell 1928, p.36, 37 of the Transaction Edition The Backward Art of Spending Money and Other Essays, 1999). The same approach is applied to psychology where, while acknowledging the importance of a qualitative appraisal of human motives, he shortly after states that “psychologists are rapidly moving toward an objective conception and a quantitative treatment of their problems. Their emphasis upon stimulus and response sequences, upon conditioned reflexes; their eager efforts to develop performance tests, their attempts to build up a technique of experiment, favor the spread of the conception that all of the social sciences have a common aim─the understanding of human behavior; a common method─the quantitative analysis of behavior records, and a common aspiration─to devise ways of experimenting upon behavior.”, (Mitchell 1928, p.26 of the quoted Transaction edition, 1999).
(VII) The previous remarks, while pointing out a possible limitation of Mitchell’s conception of empirical analysis (addressed in the next point), fully acknowledge the complexity of his approach and its relevance to the study of economic phenomena. In this regard I noted that “The rationale underlying Mitchell’s position was, however, at that time, quite innovative: in fact, it was based on the purpose of getting more precise data in order to go beyond a theoretical speculation not in contact with facts. This was particularly the case for the analysis of business cycles, where he clearly recognized the specificity and the common aspects of the various cycles. In this sense, I believe that Mitchell’s position on the importance of empirical analysis is quite appropriate, with its limitation resting on considering as reliable empirical evidence only that based on quantitative aspects.”
(VIII) The previous sentence leads to my thesis, which consists of three prongs:
(A) Many socio-economic phenomena have, along with some measurable dimensions, a set of qualitative aspects that require a qualitative interpretation (which also involves the interpretation of quantitative data). For instance, in the analysis of a particular market, we certainly need quantitative data on supply, demand and prices. But, in order to get a thorough understanding of the phenomenon, we also need a wide range of qualitative information on the characteristics of the institutional, social, cultural and psychological features that combine to define such market structure.
(B) There are various phenomena for which there are no measurable dimensions but can be nonetheless scientifically addressed. For instance, how can we “demonstrate”, say, that Martha plays the piano better than Lisa, that we love our friends, that Peter is more easy-going than James? In this regard we think that, although in these matters there is no direct demonstration as in the case of, say, identifying the fastest runners, a more “qualitative oriented” demonstration is possible. For instance, arts criticism has elaborated many criteria for assessing artistic creations, and humanistic psychology has devised many criteria for understanding the qualitative aspects of feelings (which, it is important to remark, have nothing to do with the “a priori” identification of “universal laws”). Needless to say, these assessments will always be more tentative and open to question than, say, the speed of runners. However, this does not imply that these findings are ‘less scientific’, but only that the issues addressed are more complex.”
(C) As a way of synthesis, how can we assess whether (and in what degree) a collective and evolutionary context (with its culture, institutions, norms, organizations and policy action) is conducive to economic and social progress? Of course, we can identify quantitative proxies for many phenomena, but this does not eliminate the necessity ─ in order to avoid the danger of simplification and reductionism ─ of evaluating the qualitative and specific aspects of the phenomena considered. Such appraisals would involve a process of social valuation that lies at the heart of policy action and it is likely to influence also the interpretation of quantitative data.
In conclusion, Mitchell’s conception that qualitative phenomena can be scientifically investigated only through the identification of measurable proxies weakens his right stress on the importance of empirical analysis. And I think that even today there is a need to address more closely these issues and to make more clear that the full unfolding of the distinctive elements of the OIE ─ for instance, institutions, norms, evolution, culture and social valuing ─ requires a broader concept of empirical evidence: one that would include both measurable and unmeasurable aspects of socio-economic phenomena.