The Census Bureau does not have an official definition of the "middle class," but it does derive several measures related to the distribution of income and income inequality. Traditionally, the Census Bureau uses two of the more common measures of income inequality: the shares of aggregate income received by households (or other income recipient units such as families) and the Gini index (or index of income concentration). In the shares approach, we rank households from lowest to highest on the basis of income and then divide them into equal population groups, typically quintiles. We then divide the aggregate income of each group by the overall aggregate income to derive shares. The Gini index incorporates more detailed shares data into a single statistic which summarizes the dispersion of the income shares across the whole income distribution. The Gini index ranges from zero, indicating perfect equality (where everyone receives an equal share), to one, perfect inequality (where all the income is received by only one recipient).
Generally, the long-term trend has been toward increasing income inequality. Since 1969, the share of aggregate household income controlled by the lowest income quintile has decreased from 4.1 percent to 3.6 percent in 1997, while the share to the highest quintile increased from 43.0 percent to 49.4 percent. Most noticeably, the share of income controlled by the top 5 percent of households has increased from 16.6 percent to 21.7 percent. Over the same time period, the Gini index rose 17.4 percent to its 1997 level of .459.
Researchers believe that changes in the labor market and, to a certain extent, household composition affected the long-run increase in income inequality. The wage distribution has become considerably more unequal with workers at the top experiencing real wage gains and those at the bottom real wage losses. These changes reflect relative shifts in demand for labor differentiated on the basis of education and skill. At the same time, long-run changes in society's living arrangements have taken place also tending to exacerbate household income differences. For example, divorces, marital separations, births out of wedlock, and the increasing age at first marriage have led to a shift away from married-couple households to single-parent families and nonfamily households. Since nonmarried-couple households tend to have lower income and income that are less equally distributed than other types of households (partly because of the likelihood of fewer earners in them), changes in household composition have been associated with growing income inequality.
Comparing the shares of aggregate income received by quintile between 1993 and 1994 would suggest that the amount of inequality in the income distributions of households did not change. Between 1992 and 1993, however, the amount of inequality in the income distributions of households increased significantly, although part of the increase may have been due to a change in data collection methods. For more information on the 1992-93 change in inequality see the article by Paul Ryscavage, "A surge in Growing Income Inequality?", Monthly Labor Review, August 1995, pp. 52-62.