Classical political economy starts and ends with the liberal individual. In this story, when the rational homo economicus meets others of his ilk, his natural inclination to “truck, barter, and trade” compels him to engage in mutually beneficial exchange. Classical theory sees this moment as the organic birth of the market. Smith, Ricardo, et al. begin their theoretical work by analyzing how strictly rational agents navigate this exchange. This theory of individual decision-making bears the load of the entire classical edifice, which takes a market economy to be no more than the sum of these decentralized decisions. John Maynard Keynes, however, sees a fallacy of composition in this assumption. In his General Theory of Employment, Interest, and Money, Keynes argues that the economy as a whole has its own organic existence, irreducible to the individual agents making decisions within it. Thus, he inaugurates macroeconomics as the study of the whole economy, a discipline properly free from the classical microfoundations.
Rather than analyzing the implications of rational choices and markets for individual goods, macroeconomics works on aggregate quantities such as employment, national income, and effective demand. Keynes demonstrates how these data have an organic dynamical interaction that does not supervene on any theory of individual decision-making. Furthermore, they reflect the overall health of the economy in which people actually live. Despite the fact that it is irreducible to the analysis of rational choice, macroeconomics does have a fully developed theory of the individual psychology. Indeed, the way Keynes’ General Theory describes decision-making is far more faithful to subjective experience than the idealized rational agent depicted in mathematical neoclassical theory. The Keynesian individual lives, works, and invests in a fundamentally uncertain world, and the nature of his expectations reflects that indeterminacy. In making predictions about the future, he relies on his recent experience, not double integrals. His decisions depend more on confidence than on rationality.
This essay proposes a novel reading of Keynes’ General Theory, tracing its intellectual roots back to nineteenth century physics in order to understand the complex relationship between individuals and macroeconomics’ aggregate quantities. During the 1800s, statistical mechanics and thermodynamics emerged as complementary theories of matter on different scales. While the former uses robust statistical methods to characterize the interactions of vast numbers of particles and the latter describes the behavior of matter in bulk, the two theories are irreducible to each other in just the same way Keynes’ macroeconomics is irreducible to his account of individual psychology. This reading is buoyed by Keynes’ earlier Treatise on Probability, which develops a subtle epistemology of probabilistic and statistical laws, working on the exact physical theories in question. The Treatise shows how statistical laws link the particulate-level mechanics of a complex system to an emergent dynamics, logically independent from any theory of those fine-grained interactions. As a result, neoclassical attempts to produce a macroeconomics with microfoundations are inherently contradictory.
Of course, neoclassical econometricians are much enamored of performing linear regressions and risk analysis, but this reading raises the question of whether they properly understand how to use and interpret statistics. Recent work suggests they do not. Nicholas Nassim Taleb has shown how most economic and financial models mistakenly assume that aggregation wipes out “fat tails,” allowing orthodox economists to build models using well- behaved normal probability distributions. While financiers assume this transforms uncertainty into manageable risk, unsophisticated use of statistics causes this modeling process to fall into a tautological trap. When this circular logic is used to justify highly leveraged investment strategies, it yields a financial system shockingly vulnerable to rare but consequential “Black Swan events.”
For more, download the PDF.