Evaluating Changes in the Distribution of Income in the United States

Abstract

Insights gained from previous macroeconomic studies of the size distribution of U.S. incomes are incorporated into a richer model that considers the possibility of random walks and cointegration. The effects of macroeconomic, demographic, structural, and policy variables on quintile conditional mean incomes and Lorenz ordinates are estimated using more that four decades of data. The results are interpreted in terms of the powerful dominance method of evaluating entire income distributions. The model yields strong evidence of random walks and cointegration. Several findings from earlier studies are confirmed, and heretofore unrecognized influences on the size distribution of income are identified. Some surprising results emerge, especially as they relate to the effects of a more open economy on income inequality.