Distinguishing errors in measurement from errors in optimization

Typical econometric production practices under duality ignore the source of disturbances. We show that, depending on the source, a different approach to estimation is required. The typical approach applies under errors in factor input measurement rather than errors in optimization. An approach to the identification of disturbance sources is suggested. We find credible evidence in U.S. agriculture of errors in optimization compared to errors of measurement, and thus reject the typical specification.