Quantification is not merely a strategy for describing the social and
natural worlds, but a means of reconfiguring them. It entails the impo
sition of new meanings and the disappearance of old ones. Often it is
allied to systems of experimental or administrative control, and in fa
ct considerable feats of human organization are generally required eve
n to create stable, reasonably standardized measures. This essay urges
that the uses of quantification in science, social science, and burea
ucratic social and economic policy are analogous in important ways to
accountancy.