The Marshallian Roots of Keynes’s General Theory

Abstract

The aim of this paper is to elucidate Keynes’s Marshallian lineage. I argue that the result of bringing out the Marshallian antecedents of the General Theory highlights Keynes’s failure to achieve the theoretical project he was striving at, namely to demonstrate an involuntary unemployment result in the arising of which nominal wage rigidity would play no role.

In the first part of the paper, I reexamine Marshall’s theory of value. This section’s main conclusion is that no theory of unemployment is to be found in Marshall’s writings. In the second part of the paper, I study the literature spanning from Marshall to Keynes, focusing on Beveridge, Hicks and Pigou, in order to see whether the lacuna present in Marshall’s writings happened to be filled. Documenting the emergence of the notion of frictional unemployment, I come to the conclusion that its arising went along with little theoretical elaboration. The third and last part of the paper is a critical reflection on the General Theory. I start by making the point that Keynes’s theory of effective demand ought to be viewed as an extension of Marshall’s analysis of firms’ short-period production decisions. This enables me to bring out the decisive role played by the wage rigidity assumption in Keynes’s reasoning. I claim that, except for this assumption, the differences between “effective demand à la Marshall” and “effective demand à la Keynes” are minor. I close my analysis of Keynes’s reasoning by showing that no real removal of the nominal rigidity assumption is to be found in Chapter 19 of the General Theory.

A paper presented at the first Bi-annual Symposium “Perspectives on Keynesian Economics”, July 2009, Ben-Gurion University, Beer Sheva, Israel

Appendices

Marshall, Models, and Macroeconomics: Comments on Michel De Vroey’s The Marshallian Roots of Keynes’s General Theory

David Colander

Michel and I both agree that the correct interpretation of Keynes’s work involves seeing Keynes as a Marshallian. Where we differ is on what it means to be a Marshallian. Michel interprets Marshallian as “one who uses Marshallian models”; I interpret Marshallian as “one who uses the Marshallian method”. A Marshallian in my sense would use models in quite different ways than a Marshallian in Michel’s sense.

Because Michel interprets Marshall as someone who uses Marshallian models, much of his discussion is about Marshallian models, such as Marshall’s market clearing labor market. He judges Keynes’s work within this “model” interpretation of Marshallian, and finds that Keynes essentially added nothing to the Marshallian models in terms of an explanation of unemployment. He argues that Keynes did not extend Marshall’s theory, because Keynes did not develop a macroeconomic theory of unemployment based on Marshallian models. He ultimately concludes that seeing Keynes as a Marshallian does not reinforce the validity of Keynes’s argument, but rather shows its weaknesses and undermines Keynes’ contribution.

I fully agree that Keynes did not develop a macroeconomic theory of unemployment based on Marshallian models. However, I do not see that failure as undermining Keynes’ contribution. In fact, I suggest that since Keynes used the Marshallian method, his goal in presenting Marshallian models was not to develop a theory of unemployment or a macroeconomic theory using Marshallian models. Rather I see his goal as simply to explore how a vision of the economy grounded in his work on true uncertainty would fit in those models. If it had been possible to develop his ideas in Marshallian models in any straightforward manner, Keynes, having a high regard for Marshall’s abilities, would have fully expected that Marshall would have already done it.

Instead of seeing Keynes’s General Theory as an attempt to develop a theory of unemployment within Marshallian models, I see Keynes’ General Theory as an attempt to view the macroeconomy with a different vision than Marshall implicitly used as background for his models. Elsewhere (Colander 1996, 2006), I have called this vision a complexity vision, and have argued that the novel elements of the General Theory involved an exploration of how that complexity vision might help in explaining persistently and high unemployment and the depression that the economy was experiencing at the time. This complexity vision involved giving up Marshall’s “one thing at a time-hold everything else constant” partial equilibrium micro modeling method, and replacing it with a “everything is interrelated–all things at the same time” – macro modeling method. Doing so, Keynes saw that the assumption of a unique long-run stationary state that served as an implicit backdrop for Classical and neoclassical (both informally in Marshallian and formally in Walrasian) models, was problematic, and that in the short run, many different aggregate outcomes to the economy due to aggregate coordination failures were possible. In such a vision, Marshall’s micro models had no macro foundation. Given the highly complex interdependencies of agents’ decisions, nothing inherent in a model that included those interdependencies guaranteed the achievement of a desirable aggregate equilibrium in the short run.¹⁹ This meant that you couldn’t assume an aggregate full employment level of aggregate activity as a backdrop for models as Marshall had implicitly done, and as Walras had explicitly done. Instead, you had to develop an aggregate model that would determine the aggregate level of activity that would be achieved by the economy and embed micro models within the appropriate macro model. Classical economists had not done this, and thus did not have an acceptable theory.

Within this complexity vision an aggregate economy of interdependent agents would likely have serious coordination failures. Thus, unemployment of aggregate resource was a macro phenomenon that could have little to do with micro issues. Issues of unemployment could not be reasonable analyzed in ceteras paribus models. Whether in some abstract long run these coordination failures would be solved and a desirable aggregate equilibrium would be reached was irrelevant to Keynes, because society would never wait that long, and allow that particular long run to occur by the system.

Having arrived at that alternative complexity vision after a long struggle to escape the alternative unique equilibrium (stationary state) vision, Keynes faced the problem of how to convey his new vision to other economists. To do so, he naturally tried to relate it to models that they understood; much of the General Theory involves Keynes’ attempts to do so. But these attempts should be seen as attempts to explain his vision, not as an acceptance of the alternative vision implicit in the existing micro models, Keynes’s arguments did not rise or fall on his ability to translate his vision into existing micro models; that is why when the models and the vision came to different results, Keynes accepted the vision in his discussion. He could do so because he used the Marshallian method, which sees models as aids to intuition, not as the holders of truth. Consider Keynes’s description of the use of models.

Economics is a science of thinking in terms of models joined to the art of choosing models which are relevant to the contemporary world. It is compelled to be this, because, unlike the typical natural science, the material to which it is applied is, in too many respects, not homogeneous through time. The object of a model is to segregate the semi-permanent or relatively constant factors from those which are transitory or fluctuating so as to develop a logical way of thinking about the latter, and of understanding the time sequences to which they give rise in particular cases. Good economists are scarce because the gift for using “vigilant observation” to choose good models, although it does not require a highly specialized intellectual technique, appears to be a very rare one (Keynes 1938).

This description places intuitive judgment above models. As Michel, and many others, have pointed out, Keynes’ translation of his vision into Marshallian models is confusing and involved many mistakes. Unfortunately, later economists built on those models, and lost sight of the Marshallian method as they embedded Keynes’s ideas into models. Put simply, they choose the wrong models. As that happened, Keynesian economics lost its grounding in the Marshallian method, and was replaced with a neoKeynesian/Neoclassical synthesis that focused on models in place of judgment. As that happened Keynes’ complexity vision was lost and macro theory went astray.

Why didn’t Keynes and Keynesians object to this transformation? Some, such as GLS Shackle (1949) and Paul Davidson (1978) did, but most Keynesians were more interested in policy, not theory, and were willing to accept models that came to what they considered the “right” policy conclusion even though they were intellectually unsatisfying.

Conclusion

In conclusion, seeing Keynes as a follower of the Marshallian method is important both in understanding his contribution, and in understanding how later economists lost sight of it. Keynes’ contribution to Marshallian economics was twofold. First he developed a vision of the aggregate economy that saw it as possible that it could end up for long periods of time at highly undesirable equilibria. That was an enormous contribution to theory that was unfortunately lost. Second, he saw that it was likely that to get out of those undesirable equilibria in a politically socially accepted period of time, government action might be necessary. That was an enormous contribution to policy. True, he did not develop an acceptable scientific theory of how such a complex system would operate, nor did he have acceptable models to convey his vision to others. But for those who were willing to see the vision, it was there.