This report introduces a measure that describes the distribution of income among persons and presents data based on the new measure for the years 1964, 1969, 1974, 1979, 1984, and 1989. The data are from the March supplements to the Current Population Survey.

The measure, called relative income, is intended to supplement the information available from traditional measures such as median family income and the percent of the population below the poverty line. The latter measures have been and will continue to be of critical interest, but no single measure or small set of measures can fully describe all aspects of the distribution of income.

Median family income is the amount that divides the distribution of families in half. The measure is widely used to compare the economic status of population subgroups and to determine the direction and rate of change in overall economic well-being. The measure is very useful for such purposes, but differences between groups and over time must be interpreted with a certain amount of caution. The first point of caution is that the situation of unrelated individuals is not reflected in the measure. The second point of caution is that the size of families may differ between groups and over time. For example, if families in group A and group B have the same median income but differ in average size, it seems inappropriate to consider the two groups of families as having equivalent levels of economic well-being.

Median family income is not a measure of income inequality. That is, it does not provide information about the shape of the distribution below and above the median.

The poverty measure has some features that the median family income measure lacks. It is inclusive in that it covers both family members and unrelated individuals. Poverty status is determined by comparing the family income of a person (or the person's income if an unrelated individual) to a poverty threshold that varies by family size. The introduction of a family size variable means that the poverty measure, unlike the median family income measure, takes family size differences into account when comparing families at a point in time or over time.

By itself, the poverty measure is not a measure of income inequality. The poverty measure provides information on the number of persons below certain constant-dollar thresholds. Because the thresholds are expressed in constant dollars, changes in the poverty population need not be associated with changes in the distribution of income (poverty rates may change as a consequence of a shift in the entire distribution even when the shape of the distribution remains unchanged).

The two measures of income inequality most often used in Census Bureau reports are the Gini index of income concentration and the proportion of aggregate income received by households (or families) in the bottom and top income quintiles. The Gini index is a critical measure but it summarizes the extent of inequality into a single value and must be supplemented with additional information on the shape of the distribution. The second measure, the proportion of aggregate income received by households or families in the bottom and top income quintiles, is a very useful measure of overall inequality. Some caution is required in making comparisons over time because of possible changes in average family size.

The relative income measure that is introduced in this report shows the extent to which the income of a person (or group of persons) diverges from the middle income of the universe. At a point in time, it shows how the divergence pattern of one group compares with another. Over time, the measure shows whether the total extent of divergence is increasing or decreasing (whether income inequality is increasing or decreasing).

The relative income measure of an individual is the distance from the middle of an income distribution that contains adjustments for differences in family size. A person with a relative income of .25 has only one-fourth the income of a person in the middle of the distribution and a person with a relative income of 2.00 has twice the income of a person in the middle.

The calculation of relative income involves the following steps:

1. Assign an income value to each person equal to the income of the person's family (assign the income of the person if the person is an unrelated individual).
2. Use a set of equivalence factors to adjust incomes for differences in family size. The factors reduce the incomes assigned to members of large families relative to members of small families because the cost of needs increases as family size increases. The equivalence scale used in calculating relative income is one developed by Patricia Ruggles and described in her 1990 book, Drawing the Line. The scale is based entirely on family size and does not vary by individual characteristics such as age or sex. For additional information, see the section on "Relative Income" in appendix A.
3. Calculate the median level of equivalence-adjusted income using the entire universe of persons.
4. Assign a relative income value to each person equal to the ratio of his or her equivalence-adjusted income to median equivalence-adjusted income.

The measure is well-suited for studying the extent of income inequality and changes over time in inequality. Table A, for example, shows the percent of the population with specified levels of relative income over a 25-year span at 5-year intervals.

The measure is also well-suited for comparing the income situation of one population group with another. Median family income is inappropriate for this purpose when the interest is in persons rather than families, and the poverty rate is focused on a particular part of the income distribution. For example, table B shows distributions of relative income as well as median relative incomes for persons under 18 years, persons 18 to 64 years old, and persons 65 years old and over.