Other things equal: Samuelsonian Economics

Deirdre McClosky argues that we need to get beyond the Age of Samuelsonianism in economics and get back to theorizing and observing. Economics, especially mainstream American economics, for all its promise, is in very bad shape because it has fallen into a cargo-cult version of “science” in which qualitative theorem-making runs the “theory” and statistical significance without a loss function runs the “empirical work.” Consequently, none of the high-prestige “work” in the journals is to be taken seriously. Most (say 95 percent) of its alleged “results” have to be done all over again, by economic scientists using—in preference to the mumbo-jumbo that has passed for scientific method among economists since 1947— real scientific methods (such as serious simulation disciplined by the world’s facts; and functional-form math; and statistical significance, when relevant, with loss functions; and economic history; and inquiry into all the other human sciences we economists have been invited so long to ignore). A real science—or a real inquiry into anything about the actual world—should both think and watch, theorize and observe.