Economic Review — First Quarter 1993 19 Joseph H. Haslag Lori L. Taylor Senior Economist Senior Economist Federal Reserve Bank of Dallas Federal Reserve Bank of Dallas A Look at Long-Term Developments in the Distribution of Income and education profiles of the population, the gender composition of the labor force, and (mostly) inertia in income inequality, explain the lion’s share of the forecast error variance. Policy vari- ables, such as transfer payments and tax rates, account for only 15 percent of the variation in prediction errors. What has happened to the distribution of income? Our description of changes in the distribution of income proceeds in two parts. In the first part, we divide the population into five equal-sized subgroups, or quintiles, and examine each sub- group’s gains from income growth. 2 In the second S trong economic growth in the United States during the last half of the 1980s did not trans- late into economic gains for all income groups. Poverty rates, for example, remained higher than those observed in the 1970s. 1 To paraphrase the most common findings, the rich got substantially richer during the 1980s, while the poor may have gotten poorer. A trend toward greater income inequality can be cause for concern. Most Americans would not consider it desirable if, over time, all our society’s resources became concentrated in the hands of a small group of individuals. On the other hand, few Americans would desire a perfectly equal distribution of income because income equality implies, among other things, that people who are college educated earn exactly the same income as people who are high school dropouts. If everyone earned the same income, there would be little incentive for people to work harder, become better educated, or find better, more efficient methods of production. Thus, the reasons underlying a trend toward greater income inequality are at least as important for policy analysis as the level of income inequality. We set out to investigate how and why the distribution of income has changed over time. We find that the distribution of income has been becoming more unequal since the early 1950s, making what occurred in the 1980s a continuation of a longer-running trend. We also find that the distribution of income gains over the past dozen years is close to its historical average. Finally, we examine how rising income inequality relates to changes in the economy’s demographic, business- cycle, and policy characteristics. We find that factors outside of direct policy control, such as the age We wish to thank Zsolt Besci, Stephen P.A. Brown, Christo- pher Carroll, and Keith R. Phillips for their helpful comments and suggestions, and Anne E. King and Adrienne C. Slack for their research assistance. 1 A change in how poverty rates were calculated means that poverty rates before 1975 are not comparable to those since 1975. Poverty rates stayed below 10 percent during the period 1975–80. During the 1980s, poverty rates climbed and then fell, staying above the 10-percent threshold. While economic growth appears to have roughly coincided with the declines in poverty rates, growth failed to lift enough people out of poverty to reduce the poverty rates below 10 percent. 2 Another issue arises because we use tax returns as our data source. People do not have to file tax returns if their income levels are too low. Consequently, the sample we use is truncated in the sense that the lowest paid people are omitted.