I. INTRODUCTION
It is a common argument that one of the factors contributing to the decline of institutionalism as a movement within American economics was the arrival of Keynesian ideas and policies. In the past, this was frequently presented as a matter of Keynesian economics being “welcomed with open arms by a younger generation of American economists desperate to understand the Great Depression, an event which inherited wisdom was utterly unable to explain, and for which it was equally unable to prescribe a cure” (Laidler Reference Laidler1999, p. 211).Footnote 1 As work by William Barber (Reference Barber1985) and David Laidler (Reference Laidler1999) has made clear, there is something very wrong with this story. In the 1920s there was, as Laidler puts it, “a vigorous, diverse, and distinctly American literature dealing with monetary economics and the business cycle,” a literature that had a central concern with the operation of the monetary system, gave great attention to the accelerator relationship, and contained “widespread faith in the stabilizing powers of counter-cyclical public-works expenditures” (Laidler Reference Laidler1999, pp. 211-12). Contributions by institutionalists such as Wesley C. Mitchell, J. M. Clark, and others were an important part of this literature.
The experience of the Great Depression led some institutionalists to place a greater emphasis on expenditure policies. As early as 1933, Mordecai Ezekiel was estimating that about twelve million people out of the forty million previously employed in the “industrial sector” were unemployed and that unemployment relief, to be effective, “must start with a large scale and vigorous program of construction and public works” (Ezekiel Papers Box 1, Folder: Agriculture, US Dept of 1932-33). From the early 1930s J. M. Clark was working on his own version of the multiplier,Footnote 2 and published on the cumulative effects of changes in aggregate spending in 1935 in research conducted for the National Planning Board (Clark Reference Clark1935a, Reference Clark1935b). Others within the New Deal administration, particularly Laughlin Currie, were also developing and promoting a consistent program of Government expenditures from 1935 onward. Currie's ideas found support from institutionalists such as Ezekiel and Isador Lubin. The Keynesian Revolution in America, then, had many of its roots in the work of New Deal economists, both institutionalist and otherwise (Barber Reference Barber1996, Laidler Reference Laidler1999).
It is also worth mentioning the institutionalist background of the person who became the leading American Keynesian, Alvin Hansen. Hansen completed his Ph.D. dissertation “Cycles of Prosperity and Depression, 1902-1908” in 1918 at Wisconsin, his supervisors including R. T. Ely and J. R. Commons.Footnote 3 Although Hansen did not associate himself with the institutionalist movement, and initially developed more of an interest in European cycle theories,Footnote 4 he maintained a close concern with the problems of unemployment, unemployment compensation, social security, and held to a social reform program with obvious Wisconsin roots. When Hansen took to Keynesian economics he married a Keynesian “full employment program” with a “comprehensive system of social security,” adequate public education including adult education and research, adequate public health facilities and services, adequate provision of housing including public housing, urban redevelopment and renewal, river valley development along the lines of the TVA, conservation of agricultural and forest resources, and improvement of transportation facilities (“A Full Employment Program,” Hansen Papers, 3.42, Research Notes, Box 2).
The relationship between institutionalism and what became Keynesian economics is, however, a complex one. Parts of what was institutionalism were taken over into the American Keynesianism of the 1940s, and there were certainly aspects of that Keynesianism that institutionalists could and did approve of. This however, should not be taken to mean that institutionalists simply adopted Keynesian ideas wholesale, because they most certainly did not. Hodgson's recent statement that “institutionalists such as Mitchell, Commons, Clark, Copeland, and Ayres were very sympathetic to Keynes's theories and policies” (Hodgson Reference Hodgson2004, p. 385) represents a significant overstatement. What will be argued below is that the institutionalist response to Keynes was a varied one,Footnote 5 and involved some substantial criticisms of Keynesian economics on methodological, theoretical, and empirical grounds. These lines of criticism grew out of the several institutionalist research agendas on business cycles and unemployment that were active in the 1930s. Keynesianism did not simply arrive and sweep the board, but met challenges and criticism, and not just from more orthodox economists,Footnote 6 but also from the more empirical and reform-minded institutionalist contingent.
II. INSTITUTIONALIST RESEARCH AGENDAS OF THE 1930s
Within institutionalist research on depressions and unemployment in the 1930s, there are a number of strands visible. The first of these is the program of empirical research on business cycles conducted primarily by Wesley Mitchell, Arthur Burns, and others such as Simon Kuznets and F. C. Mills at the National Bureau of Economic Research. This research built on Mitchell's earlier studies of business cycles (Mitchell Reference Mitchell1913, Reference Mitchell1927). In broad terms Mitchell saw business cycles as involving the functioning of an advanced “money economy” and deriving from the interaction of business decision making based on profit expectations, the banking and financial system, and the leads and lags in the adjustment of prices and wages.Footnote 7 Mitchell also thought of the business cycle as a complex multi-causal phenomenon, and even with certain lines of causation changing over the course of the cycle. Mitchell was certainly familiar with Keynes's work and the work of other monetary theorists such as Cassel and Wicksell, and saw Keynes's Treatise on Money as a work that sought to make money a “larger element in general economics,” and in that sense marking an important advance over the “real exchange” focus of earlier economists (Mitchell Reference Mitchell1969, vol. 2, p. 826), but these developments in theory did not deflect Mitchell from his own empirical approach.
By the later 1930s the business cycle project at the NBER had expanded to include a vast array of studies of national income and its composition, cycles in particular industries and lines of activity, the structure of prices over the cycle, inventories over the cycle, employment and incomes over the cycle, cycles in stock prices, bond yields, and much more.Footnote 8 Out of this developed the “NBER method” of specific and reference cycles. The NBER approach was not to work with highly aggregated data, and their investigations discovered a considerable amount of variation between cycles, leading to a great deal of caution in the making of generalizations or in assuming the stability of certain relationships. All of this empirical work was intended to lead to a final theoretical volume to be called The Rhythm of Business Activity, but this volume was never produced. In terms of policy, Mitchell saw the empirical work of the NBER as linked to improved decision-making in both the private and public sectors, and holding out the prospect of stabilization via improved forecasting, indicative national planning, and counter-cyclical public works programs (Laidler Reference Laidler1999, pp. 221-22; see also Alchon Reference Alchon1985, pp. 71-90).
One aspect of this research program deserves special emphasis in the present context: F. C. Mills's work on prices, begun in the 1920s (1927, 1929) and continued into the 1930s. Mills's research found growing productivity in manufacturing industry over the 1920s combined with large increases in profits and increasing price inflexibility. In his 1936 book he explicitly linked this price inflexibility to problems in “the wide and prompt diffusion of purchasing power” (Mills Reference Mills1936, p. 440). Of course, Mills was not the only person concerned with the relative inflexibility of industrial prices; Gardiner Means published his first work on the subject in 1935, and continued throughout his career to stress the significance of administered pricing. During the Depression, farm prices had fallen much more than industrial prices, leading to a general concern with “price balance,” but Mills's work linking productivity gains and price inflexibility to a rise in the share of profits was widely cited by other institutionalists and, as will be seen below, often used as part of an underconsumptionist argument.
A more theoretical strand of institutionalist work, although still relying heavily on the empirical work on business cycles done by the NBER, was that produced by J. M. Clark. Clark's Strategic Factors in Business Cycles (1934) was an attempt “to chart a course between the extremes of theoretical study which gave causes too far and too simple on the one hand, and inductive studies which revealed ‘so many factors at work, so completely interrelated, that we are likely to come to the conclusion that everything is both cause and effect’ and present too many complications to be of practical use” (Shute Reference Shute1997, p. 92). The “short cycle” is described by Clark in terms that have their origin in Mitchell's Reference Mitchell1913 volume, but with an emphasis on the accelerator mechanism, now expanded to include not merely capital goods but durable goods more generally. Clark also points to a number of other strategic factors including “movements of credit,” and cyclical shifts in the proportionate distribution of income between different groups. These shifts, together with diverse habits of consumption and saving between the groups involved, lead to “instability in the proportions of the national income saved and consumed over the course of the cycle” (Clark Reference Clark1934, p. 198). Findings such as those by Mills, Clark discussed in terms of longer term trends, with increased price inflexibility making the recovery process more difficult. At the same time, Clark was developing his version of the multiplier, and his Economics of Planning Public Works (1935b) is an extended examination of the promise and problems of public works programs as a means of countering depression and unemployment.
Clark certainly favored a public works program to help the recovery, but in Clark's view the exact effects of any such program would also depend on the method of financing, what is happening to public expenditures more generally, and on the effect on private capital formation. Clark expressed himself in favor of public deficit spending for purposes of stimulating private sector investment, but he explicitly warned against the dangers of “unlimited” deficit spending, particularly its likelihood to crowd out private investment or have adverse effects on business confidence by creating a fear of inflation: “A fair conclusion seems to be that public-deficit spending cannot bear the burden of lifting production from the level of a serious depression without going so far as to bring about . . . deferring effects on private business” (Clark Reference Clark1935a, p. 19; Reference Clark1935b, pp. 110-12).
If public expenditures are so handled as to tend to bring about a condition in which the volume of production and employment will become dependent on a perpetual continuation of such expenditures, then it will defeat the end of revival as that is commonly conceived, and will tend to bring us that much nearer the point at which the task of producing goods and maintaining the livelihood of the population could not be successfully handled by private business as now constituted (Clark Reference Clark1935b, p. 112).
It is important to understand that, in Clark's hands, neither the accelerator nor the multiplier concepts were ever applied in a mechanistic fashion, but always with caution, and a concern for the other factors that might play a role in determining the final outcome.
Somewhat similar views can be found in the work of other institutionalists such as Sumner Slichter. In 1932 Slichter strongly supported the public works programs and deficit spending suggested by Senator Wagner (Barber Reference Barber1985, p. 225), a position he continued to maintain in later work dealing with the depression (Slichter Reference Slichter1934, Reference Slichter1937). Like Clark, Slichter saw such deficit spending on public works as being a part of a “market supporting program” designed to “lift business out of the depression” (Slichter Reference Slichter1934, p. 175). Nevertheless, he was concerned that deficits be neither too large (as that may create business uncertainly) nor continued too long (as that may produce a fear of inflation), and was also concerned about possible adverse effects on relative prices (Slichter Reference Slichter1934, pp. 179-82). Slichter also emphasized the possible stabilizing role of unemployment reserves, and the importance of maintaining industrial research so that new technologies and investment opportunities may be created to produce a revival in private investment (Slichter Reference Slichter1934, p. 183; Reference Slichter1937).
A different strand of institutionalist thinking took to an explicitly underconsumptionist position, derived either from the work of John A. Hobson or from W. T. Foster and W. Catchings,Footnote 9 and often making significant reference to the work by F. C. Mills mentioned above. Hobson's work had attracted considerable attention in the United States, and he taught at the Robert Brooking Graduate School (headed by Walton Hamilton) in the 1924/25 academic year. Students at Brookings, who included Mordecai Ezekiel and Isador Lubin, were given significant exposure to Hobson's ideas (Rutherford Reference Rutherford and Pheby1994, Reference Rutherford2003a).
A good example of this line of institutionalist thinking can be found in the work of Rexford Tugwell, a member of Roosevelt's “Brains Trust.” Tugwell's position grew out of what was originally a highly optimistic view of the possibilities opened by new technology, holding out the promise of improved living standards for all (Tugwell Reference Tugwell1927). A problem in the realization of this promise was a tendency to oversaving on the part of corporations (Tugwell Reference Tugwell1931, Reference Tugwell1932, Reference Tugwell1935). Tugwell made explicit use of Mills's empirical work, arguing that firms with high overheads tended to try to build up reserves, and so did not reduce prices as costs fell. The 1920s were thus seen as a period of technological change but with little reduction in selling prices, raising profits and corporate savings, and a failure to enhance consumer purchasing power in proportion to productive potential. Tugwell, along with Mordecai Ezekiel and Gardiner Means, formed a group supporting a “structural” or planning approach to the depression. The planners supported a joint program of recovery and reform and favored some type of direct government intervention in order to increase purchasing power, allocate investment, and establish minimum standards. Many of the planners were in the Department of Agriculture and their ideas were most clearly displayed in the policies of the Agricultural Adjustment Administration (Barber Reference Barber1996, Hawley Reference Hawley1966). Despite their emphasis on planning, it should be remembered that a key ingredient of their programs was to enhance the stream of consumer purchasing power.Footnote 10
Quite closely related in spirit was the series of books produced by the Brookings Institution by Edwin Nourse, Harold Moulton,Footnote 11 and others between 1934 and 1935: America's Capacity to Produce (Nourse and Associates Reference Nourse1934), America's Capacity to Consume (Levin, Moulton, and Warburton Reference Leven, Moulton and Warburton1934), The Formation of Capital (Moulton Reference Moulton1935a), and Income and Economic Progress (Mouton Reference Moulton1935b). Nourse has stated that J. A. Hobson was “the intellectual daddy of what we did at Brookings on the Price and Income Books” (Knapp Reference Knapp1979, pp. 470-71). The main thrust of this study was summarized by A. F. Burns in the following way:
The central argument of the work is simple. The chronic retardant of economic progress is our method of distributing incomes. As national income has risen expenditures on consumption goods have failed to keep pace….The proper remedy for our economic difficulties is to increase the purchasing power of the masses, and this may be best achieved by passing on to consumers the benefits of technical progress thru price reductions (Burns Reference Burns1936, p. 477).
The first volume of the series, America's Capacity to Produce, was an attempt by Nourse to estimate the productive capacity of the United States between 1900-1930. Nourse estimated a capacity to produce significantly in excess of actual production but found no evidence of a build-up of excess capacity immediately prior to the depression.Footnote 12 The second volume of the series, America's Capacity to Consume, was written primarily by Moulton, and sought to assemble “a number of estimates relating to the volume of national income, its distribution by size among families, and the relation between income and savings” (Kuznets Reference Kuznets1936, p. 301). Moulton's study argued that consumption had not kept up with potential output due largely to a growing inequality in its distribution. Too much was going into savings rather than consumption. The third volume dealt with capital formation, and presented capital formation as depending on consumer expenditure rather than on the volume of saving. Excess savings, had, according to Moulton gone into the bidding up of stock prices and other asset values. The final volume, Income and Economic Progress, included a summary and a discussion of policy:
Our study of the productive process led us to a negative conclusion—no limiting factor or serious impediment to a full utilization of our productive capacity could there be discovered. Our investigation of the distribution of income, on the other hand, revealed a maladjustment of basic significance. Our capacity to produce consumer goods has been chronically in excess of the amount which consumers are able, or willing, to take off the markets; and this situation is attributable to the increasing proportion of the total income which is diverted to savings channels. The result is a chronic inability . . . to find market outlets adequate to absorb our full productive capacity (Moulton Reference Moulton1935b, p. 16).
Moulton's solution to this problem ran in terms of a “price policy” to ensure that productivity advances were passed on to the mass of consumers through declining prices, but no real mechanism for doing this was suggested other than one of persuading large corporations to adopt a low price policy.Footnote 13 As a part of this, Moulton argued that his price policy would not reduce profits as firms could achieve lower costs with increased output. Moulton did consider other possible solutions, including redistribution of income, and public expenditures, but found them too limited in effect. He was in favor of public works programs to counter the depression, but was clearly not in favor of continued deficit financing of public expenditure. Government is not a “Santa Claus” and accumulated indebtedness ultimately involves higher taxes or a breakdown of government credit (Moulton Reference Moulton1935b, pp. 99-100).
The Brookings studies were widely discussed among institutionalists, and the first two books of the series had a particular impact. On one side, the Brookings studies elicited critical responses both from Arthur Burns and Simon Kuznets of the NBER. Kuznets complained that Moulton's analysis of capital formation was one sided, his empirical evidence sketchy, and his policy conclusions arbitrary and inconsistent (Kuznets Reference Kuznets1936). Burns produced a very lengthy and detailed critique of the methods used, statistics produced, and inferences made (Burns Reference Burns1936). Burns expressed much skepticism concerning the Brooking diagnosis of the problem as one of chronic oversaving, although he conceded a lack of adequate research on consumption and saving to that date. He also expressed bewilderment that the policy proposal did not seem to follow from the diagnosis. If maldistribution of income and oversaving are the problem then redistribution would be the solution, not a price policy that it is claimed will benefit all consumers and profit-makers alike.
On the other side, more positive responses came from some of those involved in the National Resources Board (later the National Resources Committee/NRC). Several commentators on the work on production and consumption undertaken by the Industrial Committee of the NRC have pointed to the impact of the Brookings studies Capacity to Produce and Capacity to Consume in providing a significant part of the inspiration for the Committee's work on industrial capacity and consumption patterns (Lee Reference Lee1990, Stapleford Reference Stapleford2006). Gardiner Means supervised extensive studies on industrial capacity and Hildegarde Kneeland supervised studies on consumption (including a large scale survey of family expenditure). This work culminated in Means's The Structure of the American Economy (1939), which he saw as a step towards developing a system inter-industry planning (Lee Reference Lee1990, Lee and Samuels Reference Lee and Samuels1992). More directly, Mordecai Ezekiel used the Brookings studies on production and consumption (as well as points from Gardiner Means) in his books $2,500 A Year (1936) and Jobs For All (1939a), which contained versions of his “Industrial Expansion Plan,” modeled on aspects of the AAA.Footnote 14 This plan was designed to produce a coordinated increase in output and final demand, and raise real incomes. Ezekiel wrote that these books represented an attempt to provide a mechanism to carry “into action the general economic recommendations which Dr. Moulton had proposed in his book Income and Economic Progress” (Ezekiel to Leverett Lyon, January 11, 1939, Ezekiel Papers, Box 11; Ezekiel Reference Ezekiel1936, pp. 272-76). In the hands of Means and Ezekiel, the problems pointed to in the Brookings studies indicated the need for some form of planning rather than a price policy to be adopted voluntarily by corporations.
III. INSTITUTIONALISTS AND KEYNESIAN ECONOMICS
In 1935 the U.S. Supreme Court found aspects of the National Recovery Act (NRA) unconstitutional, and a year later made a similar finding with respect to certain provisions of the Agricultural Adjustment Act. Virtually no economists, including the planners, approved of the way the NRA had worked in practice.Footnote 15 Those in favor of planning argued for a reformulated NRA, and Ezekiel continued to present versions of his Industrial Expansion Plan through to 1939, but over the next few years the New Deal effectively moved into a second phase based predominantly on expansionary fiscal policy and more vigorous antitrust enforcement. This move was also promoted by the experience of the recession of 1937, caused, it was thought, by a sharp reduction in government expenditures (Barber Reference Barber1996, pp. 108-12), and the congruence of the continuing work of Laughlin Currie with the publication of Keynes's General Theory of Employment Interest and Money in 1936. Ezekiel has described his own transition to Keynesian ideas as follows:
Both Currie and we [Ezekiel and Louis Bean] were working with elements of what later become known as the Keynesian theory. But although Keynes’ book had been published in 1936, I myself didn't really catch up with it and absorb it fully until about 1939 or 1940. So, we were groping in the same direction that Keynes was working in England. Currie, as a matter of fact, who was the financial economist, came the nearer of any American economist to preceding Keynes in the discovery of the effect of the possibility of under-employment equilibrium and the way in which government programs worked to offset it. Of course, at the beginning of the New Deal, it was felt that the pump could be primed by public works to get activity going and then, after awhile, that could be withdrawn and they would keep on going by themselves. But when they were withdrawn in 1936-1937, that led to a new recession. And that led to a[n] awareness that however pump-priming worked, it didn't work as a primer. Something had to be kept up if conditions were such as to call for it. Later, Keynes’ analysis gave a very clear and eloquent exposition of just how it worked (Reminiscences of Mordecai Ezekiel Reference Ezekiel1957, on page 96 in the Columbia University Oral History Research Office Collection).
Also, sometime between 1936 and 1938, Alvin Hansen made his transition to a Keynesian position (Barber Reference Barber1987). Hansen's version of Keynesianism connected the idea of unemployment equilibrium to a stagnationist (or “mature economy”) thesis: that is, that insufficient investment opportunities were keeping investment below the level of full employment savings. The government would thus have to play a more permanent role in maintaining aggregate demand, and not merely act to “prime the pump” for private investment. Of course, the Keynesian program could also be connected to a policy of progressive taxation and redistribution through programs such as social security, designed to increase the buying power of the poor. This would reduce the need for deficit financing of public expenditures. This version of the Keynesian program was perhaps most evident in work by Currie, but Hansen also favored tax and social policies involving redistribution (Barber Reference Barber1987).
The Currie-Hansen program had been pretty much worked out by 1940, and it did gain institutionalist adherents. Isador Lubin gave his support and, according to Barber, Ezekiel had fully transitioned to a similar position by 1941 (Barber Reference Barber1996, p. 128). In the late 1930s and early 1940s Ezekiel did considerable amounts of empirical work on consumption, savings, and investment, with many references to America's Capacity to Consume, but with increasing reference to Keynes (Ezekiel Reference Ezekiel1937, Reference Ezekiel1942). This transition to Keynesian ideas can also be seen very nicely in Ezekiel's 1939 paper “Keynes versus Chamberlin” (Ezekiel Reference Ezekiel1939b), which presents a clearly Keynesian analysis of full employment saving exceeding investment, although with an added concern with monopolistic competition, and still with one reference to his Industrial Expansion Plan.
The larger history of the shift in view within the NRC from Means's structural planning to a Keynesian style fiscal policy has been told elsewhere (Lee Reference Lee1990), but it is certainly the case that both Lubin and Ezekiel were a part of that shift. Leon Keyserling, a former student of Tugwell's who became closely associated with Senator Wagner and a leading proponent of the 1945 Full Employment Bill, also adopted a set of ideas very similar to Currie's that he combined with some of Tugwell's planning perspective. Keyserling would, however, downplay the specifically Keynesian contribution to his advocacy of public expenditure and redistribution, and it is noticeable that Keyserling complained that many Keynesians neglected the distributional aspects of Keynes teachings (Keyserling Reference Keyserling1972).Footnote 16 For Keyserling, the problem was the maldistribution of income and savings, and this maldistributionist and redistributive interpretation of Keynesian economics was also adopted by Clarence Ayres, who became an advocate of guaranteed minimum incomes (Ayres Reference Ayres1938, Reference Ayres and Theobald1966). Thus, a number of younger institutionalists, and particularly those most closely involved with the policy discussions in Washington, did make a transition to Keynesian views, although the inspiration for this shift came as much or more from Currie than from Keynes himself.
More critical discussions of Keynes and Keynesian economics (particularly the Keynesianism of Alvin Hansen), however, came from J. M. Clark, Sumner Slichter, A. F. Burns, Edwin Nourse, Harold Moulton, and Morris Copeland. Gardiner Means also remained a critic of Keynesian analysis, continuing to emphasize the role of administered pricing and developing his own monetary theory of aggregate demand (Lee and Samuels Reference Lee and Samuels1992).Footnote 17 Some relevant material can also be found in the debates over the funding of Social Security involving E. E. Witte, other Wisconsin institutionalists, and the “Keynesians.”Footnote 18
Within this institutionalist literature on Keynesian economics there are a number of common themes. There is a concern about dogmatism, about the too-mechanical use of simplified models, about the factors omitted from the model, about the empirical foundations of some key aspects of the model, about the thesis of “economic maturity,” and the wisdom of continued deficit financing of government expenditure. Of those mentioned above, J. M. Clark was probably the most generously disposed towards Keynes and Keynesian economics, but even Clark expressed substantial reservations. Clark willingly conceded that “certain central problems cannot be successfully handled without the use (which does not imply exclusive reliance) of the income-flow method of analysis of which Keynes's studies are the most prominent form” (Clark Reference Clark1942, p. 9). Yet in the next breath he states “Keynes offers a revised Ricardianism, of similar power and exposed to similar dangers, including that of undue dogmatism of the part of disciples” (Clark Reference Clark1942, p. 9). A clear reference to the “Ricardian vice” of deriving policy recommendations from highly simplified models. Clark even wrote to Keynes using similar language:
It has seemed to me that what I call the “income-flow analysis”, of which yours is the most noted presentation, has done something which has not been done in comparable degree since Ricardo and Marx: namely, constructed a coherent logical theoretical system or formula, having the quality of a mechanism, growing directly out of current conditions and problems which are of paramount importance and furnishing a key for working out definitive answers in terms of policy. On this a “school” has grown up. All that has tremendous power; and is also exposed to the dangers of too-indiscriminating application . . . I am myself enough of an “institutionalist” (whatever that may mean) to have more than a lurking distrust of formulas and equations! But not enough of an institutionalist to ignore their importance: merely to want to think all around them and reckon with the imponderables that modify their action; and the other factors which no single formula can comprehend—for instance, the long-run incidence of continued large deficit spending (Clark to Keynes, July 24, 1941, J. M. Clark Papers, Box 2).Footnote 19
Clark also complained of the tendency of Keynesians to wish to intervene at any point short of “something called full employment,” representing a position which may not be achievable through the policies advocated, their “obliviousness” to the importance of wage and price adjustments, their “philosophy of unlimited deficit spending as the one tested and reliable way to secure full employment,” and their “dogma” that such spending will not tend to be dissipated in wage or price increases until full employment is achieved. These ideas, for Clark, were “unsupported by reason” and fly “in the face of experience” (Clark Reference Clark1942, p. 9). In a letter to Hansen he argued that “indefinite deficit spending is not an enduringly workable solution for chronic partial stagnation of an economic system like our own” (Clark to Hansen, February 13, 1939, Hansen Papers, 3.10, Box 1).
Slichter also expressed concerns involving the possible inflationary effects of the Employment Act combined with the increased power of unions, and he strongly rejected Hansen's stagnationist thesis, emphasizing instead the possibilities for private investment opened up by new technological advances (Slichter Reference Slichter1961, p. xiv; Leeson Reference Leeson1997, p. 447n).Footnote 20
Mitchell and Burns had greater difficulties with Keynesian economics. Mitchell himself did not write on the General Theory, but he did discuss it in some of his seminars at Columbia. Keynes's work is presented as an attempt to integrate monetary theory and general economic theory, although largely within the “orthodox tradition” (Mitchell Reference Mitchell1969, vol. 2, p. 826). Mitchell saw a need for such an integration, but also a need to “work on a deeper level than do Keynes and his disciples” (Mitchell Reference Mitchell1969, vol. 2, p. 826). Mitchell's response to Keynesian enthusiasm on the part of students made it clear that he was not overly impressed (Interview with Eli Ginsberg, March 17, 2000). It was Arthur Burns who mounted the principal NBER criticism of Keynesian economics, although the correspondence from Burns to Mitchell concerning this critique makes it clear Burns thought he had Mitchell's general agreement with his position (Burns to Mitchell, December 18, 1936, Mitchell Papers, 1998 Addition, Box 46, Folder B).Footnote 21
Some of Burns's language is strikingly similar to Clark's. He compares Keynes to Ricardo. Both had produced “bold and vigorous theoretical speculation” related to the pressing problem of the time, but both were subject to “serious error if the premises accepted for purposes of reasoning are contrary to fact” (Burns Reference Burns1946, pp. 4, 8). He was also concerned with the “religious” element he finds among Keynes's “more zealous followers.” Although Burns admits that Keynes had an important liberating effect on those brought up within orthodox theory, “he failed to teach economists that economic truth cannot be firmly reached by speculation alone” (Burns Reference Burns1947). For Burns, unless the economist examines the correspondence between speculation and actual facts, he “may be taking a grand vacation from life and its realities” (Burns to Joseph Willits, November 5, 1947, Rockefeller Foundation Archives, 1.1, 200S, Box 369, Folder 4371).
Burns also argues that Keynesians tend to assume that the consumption function is stable, that savings increase with income, and that private investment opportunities are too limited to assure full employment. Burns points out that equilibrium in a Keynesian model is at the aggregate level and does not imply that all markets or firms are in equilibrium, so that the final position may be not as predicted by the model. Furthermore, he points out that the research undertaken or being undertaken by the NBER suggests that the consumption function is far from stable,Footnote 22 that the secular trend of savings in relation to income has been downward, and that there is little clear evidence of secular stagnation as suggested by Hansen (Burns Reference Burns1946).Footnote 23 Burns also complained of Keynes's focus on the level of employment as opposed to the business cycle (Burns Reference Burns1952, see also Wells Reference Wells1994). Hansen replied to Burns, disputing the characterization of Keynesian economics as “Ricardian” in character, pointing to the empirical component in Keynesian macroeconomic work, and questioning Burns's understanding of Keynesian theory (Hansen Reference Hansen1947). Burns in turn wrote a lengthy rejoinder, backing down not at all:
I suspect that Hansen is troubled because my essay conveys the impression that the Keynesians are excessively mechanical in their thinking, that they gloss over the turbulent life that goes on within aggregates, that they give little heed to adjustment processes in our society, that they subject ceteris paribus to excessive strain, that they slight in particular the instability of the consumption function; and that while Keynes is guilty on all these counts, the Keynesians—among whom Hansen is outstanding—are guiltier still (Burns Reference Burns1947, p. 260).Footnote 24
Morris Copeland also criticized Keynesian analysis, his most considered response coming in the early 1950s, after the language of “models” and “model building” had begun to find its way into the literature. Copeland admitted the popularity of the Keynesian model, and found the causes of this popularity in the facts that it was aggregative in nature, embedded in the neoclassical tradition, did not attempt to deal with the entrepreneur or consumer on an individual level, and lent itself to empirical investigation. This last point was one to which Copeland, as a statistical economist himself, attached considerable importance. It is also the case that Copeland's own work on unemployment and expenditure policy came to utilize Keynesian terms and ideas (Copeland Reference Copeland1944, Reference Copeland1966; Rutherford Reference Rutherford2002). Nevertheless, Copeland complained about the number of factors omitted from Keynesian models, argued that the Keynesian contribution was one of “model analysis” alone, and that model analysis had serious limitations (Copeland Reference Copeland1951, Reference Copeland1952). Keynes's General Theory was too selective; “it omitted points that ought not to have been omitted” (Copeland Reference Copeland1952, p. 7). Being aggregative it omitted the detail of the cyclical process that had been such a central feature of the work of the NBER. Copeland argued the General Theory, in addition to ignoring relations with political and social theory, “added not to what we know about [macroeconomics] but our ability to express what we know in terms of model analysis” (Copeland Reference Copeland1951; Reference Copeland1952, p. 26).
Upholding the methods of institutionalism, Copeland aimed criticism at the limitations of model analysis, affirming it to be only a part of economic analysis. “Model analysis has thus far been confined to quantitative relationships and it has invariably involved a great oversimplification of the real world. Institutionalism has been concerned to emphasize the complexity of reality and the importance of qualitative as well as quantitative facts” (Copeland Reference Copeland1951, p. 57). Not only was the General Theory limited to model analysis, for Copeland, Keynesian models were lacking key factors and details; they were an “expurgated” version of Mitchell's theory of business cycles. It is quite clear that Copeland felt that any such simplified model would have limited applicability and predictive capacity (Copeland Reference Copeland1951, Reference Copeland1952).
Nourse and Moulton had particular difficulties with Keynes and Hansen on the policy front. Moulton wrote in 1949 that “Keynes's analysis has little kinship with that of the Brookings Institution. It diverges sharply at critical points, and the ultimate implications with respect to the future of private enterprise are fundamentally at variance from ours” (Moulton Reference Moulton1949, p. 128). Neither Moulton nor Nourse appear to have been impressed by Keynes, and were suspicious of solutions to the depression that substituted government expenditure for private investment. Nourse held a highly skeptical view of the original 1945 Full Employment Bill, but was happy enough with the Employment Act of 1946 to accept the position of Chairman of the Council of Economic Advisors (Knapp Reference Knapp1979, p. 201). Nourse's concept of the role of the council and his reluctance to adopt the kinds of policies being advocated by Keyserling, however, led to his eventual resignation. Nourse continually expressed concern over the building up of government indebtedness, and criticized government for “slipping back into deficits as a way of life in a period when production and employment are high” (Knapp Reference Knapp1979, p. 327). He also doubted the effectiveness of marco level policy tools on their own: “even fiscal and monetary policy cannot be adequately analyzed and formulated in isolation from the processes of the private business world—collective wage bargaining, administrative price setting, capital formation and investment” (Knapp Reference Knapp1979, p. 405). Those who believed in the certain ability of monetary and fiscal policy alone to control the level of economic activity he called “a cult of economic magicians” (Knapp Reference Knapp1979, p. 405). He also very pointedly rejected Hansen's stagnationist views, and continued to promote a “price policy” (Nourse Reference Nourse1944).
Not surprisingly, Moulton held similar views of Keynesian economics. In Capital Expansion, Employment, and Economic Stability Moulton and his co-authors argue that the idea of the American economy having reached a mature state is false. They point out that “a primary implication of the conception of economic maturity is that government must increasingly supplant private enterprise” (Edwards, Moulton, Lewis, and Magee Reference Edwards, Moulton, Lewis and Magee1940, pp. 163-64). The authors reject the thesis and its implication. Moulton's The New Philosophy of Public Debt (1943) carries the argument on to a direct criticism of what Moulton took to be the disregard of the problem of government debt by Keynesians such as Hansen. Moulton draws a distinction between the traditional view and the new conception of public debt. He describes the former as the view that “a continuously unbalanced budget and rapidly rising public debt imperil the financial stability of the nation,” and the latter as the idea that “a huge public debt is a national asset rather than a liability and that continuous deficit spending is essential to the economic prosperity of the nation” (Moulton Reference Moulton1943, p. 1). Moulton argues that those who adopt the new philosophy believe that “public finance is really only a matter of book keeping, that a rising debt has no adverse consequences, and that without a constantly increasing debt we cannot hope to have full employment and prosperity” (Moulton Reference Moulton1943, p. 3). Hansen's supporters responded with the charge that Moulton was misrepresenting Hansen's position (Wright Reference Wright1943, Reference Wright1945), a charge that drew angry denials from Moulton (Hansen Papers, 3.42, Box 7).
That many institutionalists took a much more cautious attitude towards government debt than those on the Keynesian side can also be seen clearly in the debates surrounding the financing of social security. The original Social Security Act of 1937 involved funding from payroll taxes only and included provision for a reserve fund and contemplated working towards a fully funded system of social security. The large Wisconsin contingent involved in Social Security, including Arthur Altmeyer and E. E. Witte, favored creating a full social security reserve fund (Witte Reference Witte1937). Witte argued that failure to set up a full reserve would only result in higher taxes later on. He wrote, “if the money collected from these [younger] workers is not in some fashion set up to their credit, there is the grave danger that the entire scheme will become one for taxing the working people for old age assistance extended to aged dependents of the present day” (Witte to Theresa McMahon, October 7, 1937, Witte Papers Box 34). On the other side of this argument were the “Keynesians,” including Currie, Eccles, Hansen, and Keyserling (Keyserling Reference Keyserling1972), who favored a pay as you go policy financed by progressive taxation. Others involved in the debate over social security, such as Abraham Epstein and Paul Douglas (Epstein Reference Epstein1936, pp. 793-806; Douglas Reference Douglas1936), were also highly critical of Witte's position. The revisions to the Social Security system made in 1939 moved away from exclusive reliance on payroll taxes, and the role of the reserves was reduced to that of a contingency fund.
IV. CONCLUSION
Institutionalists had a deep and abiding interest in the problems of unemployment and business cycles that ran back to the very beginnings of the movement immediately after World War I. They supported many types of intervention in the economic system under the rubric of “social control.” They were not believers in the adequacy of unregulated markets to achieve the maximum of social welfare, and had often advocated for policies such as counter-cyclical public expenditures, and unemployment insurance plans. The experience of the Great Depression led to a number of lines of institutionalist research, a variety of policy proposals, and a good deal of debate even within institutionalist ranks.
But because they did have established research agendas themselves, their reaction to the “new” economics of Currie, Keynes, and Hansen was far from uniformly enthusiastic. Some institutionalists did take on such Keynesian ideas. Keyserling was an early convert, and he and Ezekiel and Lubin came to support the Currie-Hansen position on expenditure policy. Clarence Ayres also adopted Keynesian ideas. Sumner Slichter, too, became increasingly close to Hansen in his views after 1950, although even then he never accepted Hansen's stagnationist thesis (Leeson Reference Leeson1997). J. M. Clark began to use the language of income-expenditure models, despite his doubts about their adequacy, and even Arthur Burns could applaud the fact that Keynesian economics had made the profession at large recognize the problem of unemployment was “the principal problem in economic theory” (Burns Reference Burns1946, p. 4).
Nevertheless, the leaders among the interwar generation of institutionalists remained skeptical about key elements of Keynesian economics. They did not like the degree of simplification in Keynesian models and were suspicious of the degree of aggregation and of the microeconomic detail left out. They were, in fact, highly doubtful about the entire “model building” approach to economics that was becoming dominant within the profession and displacing their own style of empirical economics. Their concerns, however, were not only methodological. The NBER business cycle program had shown a high degree of variation between cycles, and work on income and consumption was indicating that the Keynesian consumption function may not be as stable as assumed. They clearly did not accept Hansen's thesis of economic maturity with its implication of a need for ongoing government expenditure to make up for a lack of private investment. They were more concerned with reviving private investment than in finding a permanent substitute for it. Finally, most institutionalists were quite clearly more fiscally conservative than the Keynesians, and more concerned with maintaining price stability. They did not approve of ongoing government deficits, arguing that they would create inflation, damage business confidence, or impose long run burdens in the form of higher taxes.
In the context of the macroeconomic discussions in the immediate post World War II period, the policy concerns raised by the interwar generation of institutionalist critics of Keynesianism sometimes placed them on the more conservative side of the debate, warning of inflationary pressures and the need to consider microeconomic conditions.Footnote 25 With the benefit of hindsight, and experience with Keynesian econometric models, government deficits, inflation, and social security funding problems, their concerns seem to have been more than justified.
It is true neither that criticism of Keynes came only from more orthodox economists, nor that most of the interwar generation of institutionalists adopted Keynesian ideas. What did happen was that many of the leading interwar institutionalists remained quite suspicious of several key aspects of Keynesianism, but Keynesian economics appealed to the type of younger economists who in the early 1920s might have joined the institutionalist ranks. What institutionalism had promised then was a new approach to economics, one that related to the pressing problems of the day, and one that appeared to offer effective instruments for the improvement of economic performance. That is exactly what Keynesian economics held out to those beginning their careers in the 1940s. Also, Keynesian economics, with the linkages that Hansen gave it to a broad array of social programs, took over and expanded the social reform agenda that had been so central to inter-war institutionalism. The institutionalist movement declined after World War II, partly because of this, and the research programs they had initiated were displaced first by Keynesian modeling and then by the monetarist/Keynesian debates.Footnote 26
Those few institutionalists who did come to prominence in the post-World War II period generally adopted a Keynesian style of macroeconomics, stressing the use of fiscal policy to achieve full employment. Institutionalist markers continued to be visible, particularly in the advocacy of additional policy tools, including some form of planning or price controls. Perhaps the most obvious example of this is J. K. Galbraith, but the work of others, such as Allan Gruchy, fits the same mold.Footnote 27 Nevertheless, it can hardly be maintained that there were identifiable institutionalist research programs dealing with macroeconomic issues in the same sense that they existed in the 1930s.
