A STOCHASTIC-DOMINANCE ANALYSIS OF GROWTH, RECESSIONS AND THE UNITED-STATES INCOME-DISTRIBUTION, 1967-1986

We apply standard first degree stochastic dominance (FSD) analysis and a modification of FSD to investigate changes in the US distribution of income over time. The modifications are of two sorts. The first relates to the reluctance of many observers to accept the postulate that an increase in income benefiting only the wealthy, leaving incomes in other positions unchanged, represents an improvement in overall welfare, a proposition quite consistent with the Pareto criterion. To give greater emphasis to the economic well-being of individuals at the bottom of the income distribution a related but more restrictive welfare criterion, which we term conditional FSD, is distinguished. The second modification involves truncating the income distributions and applying the FSD principle over segments to assess the effects of year-to-year changes on particular income classes. The power of FSD in applied studies of income distributions is an empirical question. Using statistical inference procedures to construct partial orders of income distribution functions, Bishop, Formby, and Thistle provide evidence that FSD has surprising power in cross sectional comparisons of income distributions. We find essentially the same results across time; the analysis provides new evidence on the distributional impacts of growth and recessions on the economic status of specific income groups, including the much discussed middle class. -from Authors