Data on individual labor earnings are reported from two experimental economies where the primary factors responsible for income differences were differences in tastes for market income versus leisure and differences in abilities working manual job tasks. Measured income dispersion under these conditions was strikingly similar to that in the United states and other market economies, indicating that these two factors alone are sufficient to generate such income differences. Further, in tests of the functional form of the distributions of income, the hypothesis of lognormality fit better than the hypothesis of normality, just as it does in national data.

Selection dynamics are often used to distinguish stable and unstable equilibria. This is particularly useful when multiple equilibria prevent a priori comparative static analysis. This paper reports an experiment designed to compare the accuracy of the myopic best-response dynamic and an inertial selection dynamic. The inertial selection dynamic makes more accurate predictions about the observed mutual best-response outcomes.

Selection dynamics are often used to distinguish stable and unstable equilibria. This is particularly useful when multiple equilibria prevent a priori comparative static analysis. This paper reports an experiment designed to compare the accuracy of the myopic best-response dynamic and an inertial selection dynamic. The inertial selection dynamic makes more accurate predictions about the observed mutual best-response outcomes.

Starting from an initial (baseline) budget line, income-compensated price changes always resulted in substitution effects consistent with the Slutsky-Hicks theory. This behavior cannot be explained by a simple random-behavior model. Similar changes in relative prices that did not originate from the initial (baseline) budget line resulted in "undersubstitution effects": The composition of consumption changed in the expected direction, but the magnitude of change was not large enough to be consistent with the initial commodity bundle chosen. These undersubstitution effects are not explainable by shifting preference patterns or anchoring effects found in inconsistent choice sequences with human subjects.