3,015 works, 150 years of economic thought. Each one summarized and searchable, with cited passages inside.
A numerical estimate can look precise before the economic relationship it measures has even been identified. In this 1953 journal article, Gerhard Tintner explains why econometrics needs more than statistical technique: theory must specify relationships that the available observations can distinguish. Demand for meat and the dynamics of corn markets give concrete form to problems of identification, omitted influences, and uncertainty. Tintner writes as a practitioner who sees numerical estimation as useful but insists on exposing its assumptions. His discussion of taxes and subsidies brings that discipline to policy: econometrics may estimate the consequences of alternative measures, but cannot choose their ethical or political ends. Readers can discover why a calculable result is not necessarily an economically warranted one.