The origins and evolution of military Keynesianism in the United States

Abstract

From after World War II until late 1971, the U.S. economy demonstrated new capacities in terms of macroeconomic stability during the so-called Golden Age: recessions were brief and shallow while growth was strong, generating shared returns for the working class, a burgeoning middle class, as well as for capital-owners. Neither then nor up to now has there been sufficient recognition of the key role played by military expenditures in this dynamic process. These enormous outlays at times accounted for as much as 25 percent of gross domestic product—including their induced multiplier effects. This article examines the origins of what has been termed “military Keynesianism” during the Golden Age, emphasizing the little-known leading role played by Leon Keyserling. Keyserling exercised behind-the-scenes power in order to convert Keynesian theory into a macroeconomic policy that would be both consistent with the U.S. institutional context and sufficiently large to promote long-term conditions for full employment. With a degree of audacity, at a historically opportune moment given the socioeconomic resonance of the foundational document known as “NSC-68” promulgated by the “state within the state”—the National Security State—Keyserling successfully challenged the fallacious neoclassical “law” of guns or butter.

Key words:

• Guns or butter
• Keyserling
• military Keynesianism
• NSC-68

JEL classifications:

• B3
• E2
• E3
• E6
• H5

Notes

¹Scarcity became a requisite touchstone of “economics” as Lionel Robbins successfully sought in the 1930s to recast “economics” by expunging all remnants of a field of study that had long held that “the problem of historical specificity [could not be] legitimately ignored” (Hodgson, 2001, p. 28). In its stead, Robbins in particular maintained that economics was reducible to universal tenets—principal of which was “the science of choice” under conditions of scarcity (Hodgson, 2001, pp. 29, 208–209). By the 1940s the scarcity trope achieved hegemonic status within the corpus of neoclassical economics, as analyzed below.

Unsurprisingly, institutional economists, beginning at least with the work of John R. Commons (1862–1945), have long challenged the idea of scarcity as an underlying condition of modern industrial societies—such as the structure of monopoly (oligopoly) capitalism, as it first emerged in the United States in the 1880s. According to University of Tulsa economics professor William Dugger: “Scarcity is not a binding constraint, not an actual or theoretical problem unless all of an economy’s productive resources are being fully utilized and we do not know of a case in which that has been achieved on anything approaching a large scale or maintained for anything approaching an extended time. Instead, scarcity is a pervasive condition related directly to the failure to achieve universal employment: Universal employment is impossible, but moving toward it is not, so universal employment involves continually reducing measured unemployment, pushing it always closer to absolute zero even though absolute zero is never achieved” (personal communication, July 11, 2015, based on Dugger and Peach, 2009; Peach and Dugger, 2013).

According to University of Missouri-Kansas City economics professor John Henry: “During [World War II], economists were hired by the military to undertake analysis of optimal deployment of [military] resources in various operational scenarios … . The most important figure in this regard was Tjalling Koopmans and it is in this capacity that ‘the allocation of scarce resources among competing ends’ figured directly into real-world activity…. It is in this work that the production possibilities curve entered the neoclassical economist’s vocabulary in the context of the famous ‘guns versus butter’ trade-off … . What Koopmans and others did was to then generalize the mathematical operation [regarding the optimal usage of battleships, etc.— J.M.C.] to the economy as a whole where it did not apply—resources are not ‘given,’ there is no trade-off [at less than full-employment—J.M.C], various assumptions such as plastic capital cannot be satisfied” (personal communication July 12, 2015, based on earlier lecture notes; emphasis added).

²Long-wave analysis originated with N.D. Kondratiev’s (1892–1938) exploration of sequential, protracted, periods of relative economic expansion (perhaps twenty to thirty years) followed by equally enduring periods of relative economic stagnation, thus comprising one “Kondratiev Cycle” (Day, 1976). Kondratiev’s work regained attention in the 1970s as critical economists attempted to disentangle and order the multiple causes for a decade of recurrent economic crises and sustained weak macroeconomic performance. Three strands of analysis emerged:

First, we note a heterogeneous mix of formulations and theoretical attempts that were heavily influenced from the 1930s onward principally as a response to J. Schumpeter’s idea that major (product and process) innovations have tended to “bunch” in time—a formulation critically analyzed by N. Rosenberg (1994, pp. 47–84), and more broadly interpreted by C. Freeman (2009). Such innovations tend to generate multiple derivative innovations, thereby producing through diffusion “new technological systems,” or even initiating new “techno-economic paradigms,” wherein—gradually over a decade or more—the chain-reaction effect generating derivative innovations declines through a process of “maturation.” This formulation was advanced by both Mensch and (more recently) Pérez, and was well-documented by van Duijn (Mensch, 1979; Pérez, 2003, 2004; van Duijn, 1983). J. van Duijn, for example, notes the rapid occurrence of twelve major innovations in the period 1935–39, followed in only a two-year period, 1942–43, by eight major innovations, then eight more in the four years from 1945 to 1948, and thirteen in the four years 1950–54 (van Duijn, 1983, p. 178). After this, however, the pace slows dramatically, with “only” fifteen major innovations appearing in the 1955–71 period. It should be noted that a very significant number of the major innovations from the 1935–54 period can be traced to major military R&D efforts. Nonetheless, the idea of technological clustering undergirding a long wave of accumulation cannot be understood through the prism of “technological determinism.” Rather, it must be construed as a necessary but insufficient condition sustaining an extended expansion—which requires major institutional innovations, including policy innovations—such as military Keynesianism during the 1948–70 expansion in the United States.

Second, E. Mandel made a concerted attempt to refashion much of classical Marxism (which had found Kondratiev’s formulation to be anathema) in order to place long-wave analysis at the epicenter of a historical dynamic process based on “epochs” of accumulation (Cypher, 1977; Mandel, 1975). Analyzing the “third” epoch—from 1940/45 until the 1970s—he used the U.S. economy as his prime example. In this context he analyzed the role of military expenditures, coming to the conclusion that “in the long run the ‘permanent arms economy’ cannot resolve any of the basic contradictions of the capitalist mode of production” (as cited in Cypher, 1977, p. 81). In doing so, Mandel essentially elided military Keynesianism, while confusingly conceding that under conditions of less than full employment, military spending can raise profit, production, and employment levels (Cypher, 1977, p. 81). Ignoring the fact that policy initiatives are deployed in a series of short-run conjunctures, rather than “in the long run,” his muddled presentation eventually led only to “argument-by-assertion” based on theoretical abstractions since he did not offer a specific analysis of the concrete role of military expenditures within the context of the U.S. economy (Cypher, 1977, pp. 80–82).

Third, and also in the 1970s, another attempt was made to operationalize Kondratiev’s analysis—now renamed as “long cycles” or “long swings,” which defined the parameters of rising and declining social structures of accumulation, or SSAs (Gordon, 1980, p. 11; Bowles et al., 1990, p. 18). Exactly what constituted an SSA, however, alternated. Gordon (1980) proposed what was defined as a set of institutions that would compose the SSA, consisting of: (1) agents of accumulation (1 component), (2) motors of accumulation (two components), (3) systemic requirements for accumulation (2 components), and (4) requirements for individual capital accumulation (7 components). Subsequently this twelve-component analysis was collapsed into three (Gordon et al., 1987). In their 1990 version, military expenditures were very briefly analyzed, essentially from the standpoint of the crowding-out hypothesis advanced by Melman (as discussed later in this article): “Because the United States … [was] devoting [a] substantially higher proportion of aggregate domestic product to military spending, [it] had significantly less available to devote to productive capital formation” (Bowles et al., 1990, p. 64). This, of course, simply echoed the Robbins/Koopmans economics of scarcity argument, as briefly discussed above. They proceed, in their one-page analysis, to attribute the so-called dollar overhang (the amount by which U.S. dollars held abroad exceeded gold reserves) to military expenditures that entered the U.S. balance-of-payments account as “unilateral transfers” for various outlays including foreign military assistance. As a result, they claimed, “military spending eventually [sapped] the economy’s strength” (ibid.).

However, with regard to the unilateral flows in the U.S. balance of payments, specialists such as R. Parboni argued that from 1950 to 1970 the United States maintained “a current account surplus, in particular in its trade balance” (Parboni, 1981, p. 49). U.S. unilateral flows, particularly in the late 1940s through the late 1950s, were used “to finance the accumulation of reserves by the rest of the world” (ibid., pp. 49–50). During the 1960s, U.S. foreign direct investment was the main way in which the United States financed the build-up of these world reserves. In the first nine years of the 1970s, the United States incurred massive current account deficits, largely due to a cumulative trade deficit of $70 billion, but this must be “compared to a surplus of $40 [billion] over the preceding twenty years” (Parboni, 1981, p. 50). The reasons for this turnaround were multiple and complex; according to Parboni’s well-documented analysis U.S. unilateral military expenditures, from the late 1940s through the 1960s, were accorded no primary causal role (Parboni, 1981, pp. 29–98).

SSA analyses of the U.S. economy have continued to appear, but their focus has never been on offering a systematic understanding of (1) the role of military expenditures in terms of the underlying causal factors giving rise to alternating long waves of expansion and stagnation, or of (2) the long U.S. postwar expansion (e.g., Block, 2011a; Kotz, 2015; McDonough et al., 2006). R. Boyer recently hypothesized that the SSA approach, and other such attempts, have foundered because, while “it is possible to identify explicit partial and temporary regularities that are indexed upon a given institutional configuration of capitalism,” it is “quite unlikely” that economic analysts could ever locate the existence of “general quantitative regularities” due to evolutionary reconfigurations of the institutional structure (Boyer, 2011, p. 59).

Additional information

Notes on contributors

James M. Cypher

James M. Cypher is a research professor, Ph.D. Program in Development Studies (UAED) at the Universidad Autónoma de Zacatecas (Mexico) and research scholar at Binzagr Institute. This text is an expanded version of a work in progress; an entry on “Military Keynesianism” for volume 9 of Historisch-kritisches Wörterbuch des Marxismus (Berlin, 2016). The author is grateful to John Henry, Juha Koivisto, and Bill Dugger for comments on an earlier draft of this article.