Picoeconomics and microeconomics (and macroeconomics)

"Ross (2005) argues that picoeconomics relates to microeconomics in the same way that post-war neoclassical microeconomics is related to macroeconomics - given the right construal of the latter relation, which is not the one found in standard histories of economic thought. Ever since macroeconomics was invented by Keynes and Hicks, economists have generally assumed that, at least in principle, it ought to reduce to microeconomics. It is not always obvious what they have in mind by this, or that they have in mind something clearly defined. Philosophers of science decompose the idea of reduction into several species... Sometimes reduction is mereological and explanatory: we seek to explain both the existence and the nature of composite (or ''higher-level'') kinds of things by showing how they are composed out of ''lower-level'' things whose existence and nature are taken for granted for the purposes of the explanation...

Almost all economists believe that macroeconomic objects and processes reduce to microeconomic ones in the first sense. Note that this is a meta-physical assumption, not something that economics shows. Nor is it something economists aim to show; if demonstrating the cogency of this sort of reductionism is anyone's job, it is one for philosophers. Many economists do, however, hope to reduce macroeconomic theory to microeconomic theory and some - rational expectations theorists - sometimes say they have pulled it off. This claim is misleading, however. Rational expectations macroeconomics borrows microeconomic modeling techniques, to be sure; but it does so by constructing theoretical fictions - infinitely lived agents and atomless measure spaces - that do not occur in any microeconomic generalizations (Ross 1999). This is not reduction as any philosopher of science would understand it. More to the point, it clearly has nothing to do with any sort of explanatory reduction. (In any case, few economists now expect rational expectations macroeconomics to be the whole or final account of macroeconomic phenomena, and many think it applies only to very special circumstances)...

Macroeconomics is explanatively basic relative to microeconomics, thus reversing the traditional picture of the relationship. Macroeconomic dynamics create microeconomic agents - roughly, whole people during the stretches of their biographies in which their behavior is consistent enough to satisfy the weak axioms of revealed preference, plus Houthakker's axiom. The interactions of these agents - their games - then feed back recursively into macroeconomic patterns...

This picture is apt to seem incredible to an economist trained in graduate school, like most, to think that any macroeconomic generalization is suspicious unless we can see at least in principle how it might have solid ''microfoundations.'' However, this combines a legitimate concern for unity of science - for concern with avoiding commitment to types of objects and processes, and to theoretical generalizations, that are ontologically ''stranded,'' and which thus must seem to be posited merely as ad hoc devices for writing down predictions - with a particular philosophical view about how unity must be achieved that does a poor, and increasingly poor, job of describing ontological and intertheoretic relationships found across the network of sciences."

--- The economics of the sub-personal, Don Ross