PDF

oct09-bergerson.pdf

Much has been made of the extent to which mainstream economics is founded on analogies with equilibrium concepts in physics. I argue that more fruitful analogies can be found with current ideas in non-equilibrium statistical mechanics. I begin by discussing the extent to which aggregate quantities are meaningful and the assumption of neoclassical economics that assets can be substituted and related to a single one dimensional measure. Next I move on to the concept of utility and money in the context of state vs. process variables of thermodynamics. The rest of the paper is concerned with exploring differences between equilibrium and non-equilibrium statistical mechanics. The absence of extremum properties in the latter case gives rise to serious stability concerns, unlike the situation in equilibrium theory where extremum properties guarantee stability. I also explore the issue of “illusion of control” which arises in multiagent systems where agents have minds of their own, but has no analogy in simulations of physical systems of atoms and molecules. Finally I discuss the concepts of irreversibility which can be ignored in equilibrium and dissipative structures which appear far from equilibrium.