A Defense of Equilibrium Reasoning in Economics

[摘要] Critics both within and outside of philosophy have challenged economics wholesale as unscientific. In particular, economics seems unable to predict future events because it relies on assumptions like equilibrium conditions, which stipulate that the economy tends to stay in its current state absent external forces. The popular background view that gives rise to this criticism is that the job of science is to uncover laws of nature, by appeal to which we can determine (usually deductively) the future behavior of a dynamical system as it evolves. I argue that lawlike statements in economics have a very different role than this: they provide a means of understanding in terms of how efficient a particular system is. This account is more faithful to the history and the practice of economics. Perhaps surprisingly, it also accounts better for the explanatory power of some laws of physics.By reinterpreting ceteris paribus assumptions, and equilibrium assumptions more generally, as tools both for articulating constraints on a system as well as for identifying opportunities for a system, I am able to take into consideration the ways in which both engineers and policy makers aim to design or test such complex systems for stability. Macroscopic properties such as stability cannot be reduced to details about the individual atoms that make up bulk material or the individual agents that make up the economy. Yet, the behaviors of the micro-constituents and of the macroscopic aggregate are related. In order to address this lacuna between the micro and the macro, I explore the possibility of exploiting newer methods in the material sciences, namely multi-scale modeling, that have been useful in talking about these interesting and desirable macroscopic properties. These methods use not one but multiple models at different temporal and spatial scales to describe a system, without prioritizing any particular one. Given the substantial methodological and formal analogies between thermodynamics and economics, the success of the multi-scale framework in the former suggests it will be similarly useful for the latter.

[发布机构] the University of Pittsburgh