Profit Maximization, Returns to Scale, and Measurement Error.

A nonparametric analysis of agricultural production behavior was conducted for each of the contiguous forty-eight states for the period 1956-82 under the joint hypothesis of profit maximization, convex technology, and nonregressive technical change. Tests were conducted in each state for profit maximization and for constant returns to scale. Although considerable variability was observed among states, measurement errors of magnitudes common in secondary data yielded test results fully consistent with the profit-maximization hypothesis in all states with complete output and input data. Copyright 1992 by MIT Press.